chapter i: introduction - Reduplication



CHAPTER I: INTRODUCTION This dissertation is a study of reduplication in general and a description of the types of reduplication in Bikol (bcl)1, a Philippine language, in particular. Reduplication has always attracted the attention of linguists, perhaps especially because speakers of Indo-European languages are not familiar with the phenomenon, but they can find it in almost every nonIndo-European language. Sapir noted that "Nothing is more natural than the prevalence of reduplication, in other words, the repetition of all or part of the radical element" (Sapir 1921: 76). Among the many languages which make use of reduplication, the Philippine languages are known to do so to an outstanding degree. Blake even argues that nowhere "perhaps is this linguistic principle more productive than in the Philippine languages [...]" (Blake 1917: 425). I do not consider this statement sustainable, but it reflects very well the first impression that one gets by looking at the morphology of Philippine languages, namely that reduplication can exercise almost all functions and that the operation can be applied nearly without any restrictions. This might be seen as one of the many challenges that Philippine languages pose to the universality of grammar (cf. Himmelmann 1991). Although the extensive use of reduplication with its wide range of different forms and meanings is often cited to be a characteristic feature of Austronesian languages, and although there are some compilations of reduplication types for some of these languages2, grammars of Austronesian languages usually do not provide any detailed analysis or explanation of the reduplication system. What Sperlich (2001) notes on Niuean (niu), an Eastern-Polynesian language, is true for Austronesian languages in general: Previous studies "... have noted the importance of reduplication but have not analysed the phenomenon in depth" (Sperlich 2001: 280). Grammars usually list the different forms and their respective meanings (at best, different types are separated from each other), but no further information on morphology, syntax and semantics of the reduplication types is provided. In order to be able to obtain deeper insights to reduplication in Austronesian languages, different systems must be described and compared. In a broader perspective, this is then also relevant for larger typological research on the topic. 1 2

The small letter codes refer to the Ethnologue, 15th edition ( For example Blake (1917) and Naylor (1986) for Tagalog (tgl), Gonda (1950) for Indonesian languages, Finer (1986/87) for Palauan (pau), or Kiyomi (1995) for Malayo-Polynesian.



The "Graz Database on Reduplication"3 is dedicated to such a cross-linguistic typological study of the phenomenon, collecting a broad array of data from very different and typologically sampled languages. Naturally in such broad comparative work, many details have to be simplified for structural reasons. This dissertation on reduplication in Bikol aims to exemplify the other side of the same coin – a detailed study of one specific reduplication system of one language. To this end it makes use of insights gained from cross-linguistic research, and it pays back its dues, since the much more detailed insights into a specific system undoubtedly enrich the typological results and structural knowledge of reduplication in general. This is expressed by the citation of Trendelenburg, chosen by Pott as a motto for his work (1862: ii): "Wo das Einzelne scharf beobachtet wird, offenbart es an sich die Züge des Allgemeinen." Due to its high iconicity and to its association with child (directed) speech, reduplication was and is often considered to be a "primitive" means of word formation. Wundt (1900) analyzes: "Der einfachste Fall einer Verbindung articulirter Laute zu einem Ganzen [...] ist die Lautwiederholung. Sie lässt sich einerseits als die primitivste Form der Wortbildung überhaupt auffassen, als eine Form, die eben erst an der Grenze liegt, wo der articulirte Laut in das Wort übergeht, und die mit den einfachsten Mitteln zu Stande kommt. Andererseits gehört aber doch auch dieser Vorgang schon den erscheinungen der Wortbildung durch Zusammensetzung an, [...]" (Wundt 1900: 578-579). This view is probably enhanced by the fact that reduplication exists rarely as a morphological procedure in Indo-European languages in adult speech4, but is almost exclusively known as a phonological phenomenon in child language. Therefore it was ascribed to “more primitive” stages of languages. Gonda (1950) introduces his article on reduplication in Indonesian with the following statement: "In 'more advanced stages of civilization', among groups and classes which have, generally speaking, lost contact with 'primitiveness', although it is always apt to come to the surface, in circles where the so-called intellectual or modern mental structure is dominating, reduplicating and iterative devices are a rather unimportant part of language" (Gonda 1950: 170). 3


Located at the Institute of Linguistics, University of Graz, funded by the ÖNB (Österreichische Nationalbank) and the FWF (Fonds für Wissenschaft und Forschung). This dissertation was partly developed within the scope of the project. Maybe this view has to be modified with more careful examination of non-standard varieties of IndoEuropean languages. For example Harris and Halle (2005) "detected" a reduplicative plural inflection in Spanish dialects.



Such a view is long since antiquated, and the following description of the complex system of reduplication types in Bikol exemplarily shows that reduplication can be (and in most languages is) a full-fledged part of grammar, and is far from being simple or primitive. This work is based on two fieldwork stays, ten weeks altogether, in the Bikol region in Southern Luzon, The Philippines.5 Data were collected with various consultants of different age and social background by making recordings of spontaneous speech and story telling, and through elicitation tasks. The most important source of data on lexical reduplication was the analysis of the reduplication types found in the Bikol-English dictionary by Mintz and Del Rosario Britanico (1985).6 The dissertation is organized as follows: In Chapter II, a definition of reduplication is provided and the phenomenon in general is described, from the formal and the functional points of view, and with respect to its occurrence in the languages of the world. That is followed by an overview of reduplication in linguistic research, pointing out the most important problems and unresolved questions in the theoretical approaches to reduplication. Chapter III is a grammar sketch on Bikol, which is not at all exhaustive, but focuses on the aspects which are considered to be important to the understanding of the following description of the reduplication system of Bikol in Chapter IV. Chapter IV is the “core chapter” of this work, the analysis of the types of reduplication in Bikol. It is divided into two main parts: The first part is dedicated to lexical reduplication, the second part to productive reduplication. The systematic description is supplemented by several excursuses which go into detail with respect to specific problems of reduplication, illuminating them with methods of resolutions from different theoretical perspectives. The excursuses are intended to provide some additional information and discussion on selected theoretical issues for those readers who are interested, but they can be skipped for a continuous lecture of the description of the Bikol reduplication system. Finally, Chapter V briefly summarizes the results of the work.

5 6

Cf. map 2, appendix A. For more information on the language and the collected data see III.1.



CHAPTER II: REDUPLICATION 1 Definition of Reduplication Reduplication is a linguistic form which contains systematic non-recursive repetition of phonological material for morphological or lexical purposes. This is a very general definition of reduplication and, as such, a result of my in-depth investigation of the phenomenon. It intentionally mentions neither the morphological property, nor any restriction on productivity or on the number of copied units. Previous definitions which contained such more precise specification were abandoned on the basis of the analysis of the various reduplication types in Bikol and comparable data from other languages and language families. Hurch (2002)7 and Kouwenberg and LaCharité (2001) for example, restrict the definition of reduplication to only those constructions which are systematically and productively applied in a specific grammar. After my extensive lexical analysis of Bikol and the comparison with other unrelated languages, I abandoned this restriction due to the observation that lexical reduplication in Bikol, as well as crosslinguistically, is systematically structured to an outstanding degree (cf. IV.3, also Mattes and Vollmann 2006). Wilbur (1973: 5) emphasizes the morphological nature of reduplication, although it superficially resembles a phonological rule in the classical sense of the SPE 8 framework (Chomsky and Halle 1968). The reason why I do not integrate this fact in my definition is that there is a large amount of lexical reduplications in Bikol, as well as in other languages, which indeed partly have their origin in productive full reduplication, but in a great measure have to be assumed to have arisen as reduplications without a morphological rule, i.e. "extragrammatical reduplications". Some languages without productive morphological reduplication often possess extragrammatical ones, as for example Portuguese (por) lemba~lemba and chique~chique (both denoting plants) etc. (cf. Kröll 1991). Bikol clearly 7


A reduplicative construction is a set of at least two linguistic forms F and F' in a paradigmatic, i.e., nonsuppletive morphological relation in which F' contains a segment or a sequence of segments which is derived from a non-recursive repetition of a part of F. Reduplication is given, if a specific grammar makes systematically use of reduplicative constructions (Hurch 2002: 56). The Sound Pattern of English (Chomsky and Halle 1968).


Definition of Reduplication

has all kinds of reduplication types, i.e. lexical and productive, grammatical and extragrammatical. And I assume that this combination of types in one language is not exceptional. The following typology will show that these types have much more commonalities than differences. Therefore I claim that all must have their proper place in a complete typological study and in a universally valid definition of reduplication. There are some phenomena which are superficially similar to reduplication, and which are labeled as such by some authors, but which definitely do not fulfill the criteria of the above definition of reduplication. Some languages possess purely phonological/prosodic doubling. It seems to aim at producing context-sensitively more highly preferred prosodic patterns out of less preferred ones. But - and this is the crucial difference to reduplication - it does not at all affect the semantics or the grammatical status of the word. For example in Tarahumara (tar), in certain prosodic contexts a finally accented bisyllabic word may add a final syllable consisting of the consonant /k/ plus the vowel of the last syllable, e.g. chopé = chopé~ke 'pine firewood' or muní = muni~ki 'bean(s)'. This phonological reduplication has clear preferences with respect to the position of the word within the intonational phrase. 9 Another example of reduplication for prosodic well-formedness comes from Movima (mzp), where certain noun roots can be reduplicated in order to meet the phonological requirement to occur as an independent noun (e.g. lo-ba:~ba 'body', or du~du:~du' 'back' (Haude 2006: 94-95). Such phonological doubling does not exist in Bikol. Much more frequent in the languages of the world are two grammatical operations that are often wrongly considered to be reduplication. The crucial difference is however, that these phenomena do not copy a certain unit of the base, but apply a morphological procedure several times (cf. also Hurch 2003, Hurch and Mattes, to appear). One example is recursive affixation, e.g. the German (deu) ur-prefixation, like Ur-ur-großmutter 'great-great-grandmother' suffixation. In Bikol it is very common to recursively apply the affixation of intensifying -on, as in Pagal-on-on ako.10 'I am very, very tired' (cf. III.2.2.3). Another example is syntactic repetition, e.g. English (eng) very, very hot. or German. Also in Tagalog, intensification can be expressed by the repetition of the respective word, connected by a linker11, as in Pagod na pagod ako. 'I am very tired.' 9 For more details see Hurch (2002). 10 tired-INT-INT 1SG.AF 11 In Tagalog, as in Bikol, content words cannot stand in juxtaposition. They are always “separated” by a function word (cf. III.2.4).


Definition of Reduplication

(Schachter and Otanes: 231).12

2 Criteria for the classification of reduplication types Special cognitive properties of reduplication differentiate it from other additive and modificatory morphological procedures. I.e. the morpheme is not specified with respect to its segmental form, but its phonological form depends on the form of the base. Despite the preliminary formal character of the definition (cf. II.1), reduplication is clearly a word formation procedure and is not phonologically motivated. Because of this particular characteristic, there are two relevant levels for the categorization of the different types of reduplication – be it a cross-linguistic or a language internal study, i.e. the formal and the functional level. With some languages or language comparisons it might be useful to compose a list of the functional reduplication types, especially if there is only one formal type, as for example full reduplication in Afrikaans (afr) (cf. Botha 1984). But generally it is more convenient to start from the formal level. Whereas the different formal types usually can be listed unambiguously, the functional types often overlap or can be described only vaguely. 2.1 Formal classification The basic formal distinction between reduplication types is that of full versus partial. For productive reduplication, full means that a morphological constituent (the simplex form) is copied as a whole.13 E.g. a root, a stem, a word, or, much less frequent, an affix. Examples are Afrikaans laag 'layer' --> laag~laag 'one layer after the other' (Botha 1984: 102), Papiamentu (pap) ketu 'quiet' --> ketu~ketu 'very quiet' (Kouwenberg 2003: 162), Bikol ma-gutom 'hungry' --> ma-gutom~gutom 'somewhat hungry'. Full lexical reduplication looks identical, but it is not (synchronically) related to a simplex form. This means that it can simply be defined as one lexeme which consists of two or more segmentally identical parts. For example Brazilian Portuguese (por) quem~quem 'ant' (Kröll 1991: 35), Kwaza dynã~dynã 'to nod no' (van der Voort 2003: 78), Bikol sing~sing 'fingerring' or sari~sari 'various, miscellaneous'. Partial reduplication means that a portion of the simplex form, smaller than the whole, is copied, or, in the case of lexical reduplication, that the lexeme contains a certain segmental string twice or more. The reduplicated “portion” can be a segmentally or a prosodically defined unit, i.e. a 12 Such a repetitive construction for intensification of properties or states does not occur in Bikol, however. 13 In Pott's terms: "Gemination, d.h. Wiederholung im Ganzen" (Pott 1862: 16).


Criteria for the classification of reduplication types

phoneme sequence, a syllable, a foot, etc.14 Examples are French (fra) fille 'girl' --> fi~fille 'little girl' (Rainer 1998: 279), Kwaza dury 'roll' --> dury~ry 'is rolling' (van der Voort 2003: 75), or rilo~lo-da-ki 'I am staggering' (ibid.: 79), Bikol tulo 'three' --> tu~tulo-he 'exactly three' or bula~lakaw 'shooting star'. Several authors have argued in favor of the hypothesis that the existence of partial reduplication implies the existence of full reduplication (cf. Moravcsik 1978, Rubino 2005a, b). Reduplication can be further differentiated as to the position, i.e. the position of the part of the simplex form which is reduplicated (the "reduplicating base"). It can be initial, as in French fi~fille 'little girl', final, as in Kwaza dury~ry 'is rolling' or internal, as in Bikol dar~akula 'big,


(for details cf. IV.4.1.4) or bula~lakaw 'shooting star'. In the case of full

reduplication, the position is irrelevant. In addition, reduplication can differ with respect to the direction in which the reduplicated unit (referred to as "reduplicant" in the following) is copied, i.e. left, as in Bikol tu~tulo-he 'exactly three' or d-ar~akula 'big,


right, as in

Kwaza dury~ry 'is rolling' or undefined, when there is no simplex form as in Bikol bula~lakaw 'shooting star', or in cases of full reduplication, as Papiamentu ketu~ketu 'very quiet'. However, if a language has a clear preference with respect to the positioning of its affixes, then for cases of full reduplication a direction is assumed that is congruent with this preference, at least for descriptive purposes. Bikol has a strong preference for prefixation and it has only leftwards partial reduplication. Therefore full reduplication is also glossed as "left" (cf. for example (26)). Reduplication can also be described with respect to adjacency, i.e. the reduplicated material is adjacent, if it is attached directly to its base, whereas in other cases the reduplicated material and its base may be separated by a part of the stem, as in Woleaian (woe) liugiuw 'expect it' --> liugiu~liug 'expect,


(Sohn 1975: 130) or by other material,

as for reduplicants with fixed segmentism (for example Bikol muru~malisioso 'somewhat malicious' (for details cf. IV.4.2.3)), or as in Alamblak (amp), where full reduplication is always separated by -ba- (hingna-mara-ba~mara-më-r15 'he worked very well', Bruce 1984: 165). Reduplication is classified as contiguous if a contiguous succession of segments from the base is copied contiguously (for example Afrikaans laag~laag 'one layer after the other' or Woleaian liugiu~liug 'expect,


It is non-contiguous, if the reduplicant is

14 In Pott's terms: "verkürzte und nur zum Theil, also bloß andeutungsweise vollzogene Wiederholung: Reduplication" (Pott 1862: 16). 15 work-RED-ba-straight-REMOTE.PAST-3SG.MASC


Criteria for the classification of reduplication types

interrupted by additional material, for example by an infix as in Bikol b-in-a~bakal 'is buying, UG'

(for details cf. example (68)). Reduplication can be exact or non-exact. In the first case the

reduplicant is identical to the corresponding material in the base, in the latter case the reduplicant differs in some respect from the base by exchange of a feature or a segment, as for example in echo-words, like English party~shmarty, or Bikol harap-hasap 'rough' (for details cf. IV.3.5), or by addition (e.g. Bikol muru~malisioso 'somewhat malicious') or deletion (e.g. Bikol nag-ta~trabaho 'is working', for details cf. IV. of segments, etc.16 Furthermore reduplication can be obligatorily or optionally combined with other affixes. Obligatory emergence of reduplication in company of a certain affix is sometimes called "automatic reduplication". For example in Ilokano (ilo) the pretense expressing prefix aginautomatically triggers reduplication: singpet 'behave' --> agin-si~singpet 'pretend to behave' (Rubino 2005: 18). At last, in a few languages reduplication can systematically include more than one copy, e.g. in Mokilese (mkj) which differentiates between "duplication" for progressive and "triplication" for continuative: roar 'give a shudder' --> roar~roar 'be shuddering' --> roar~roar~roar 'continue to shudder' (cf. Harrison 1973: 426). Moravcsik (1978: 304) claims that there is no reason to expect any restrictions on the possible number of copies in reduplication. Nonetheless, the grammatical use of three copies is rare, and I am not aware of any language that has regular "quadruplication". This fact together with the observation that in Bikol full exact reduplication is phonologically blocked for bases with two identical syllables (which would result in four identical syllables), I assume that in most languages the sequence of two identical units are acceptable, while more than two cause a "horror aequi" and are rejected. 2.2 Functional classification As can be seen from the preceding section, reduplication can have many different phonological appearances. But as reduplication is a morphological or lexical phenomenon, the important question is, whether these different forms are restricted to expressing specific meanings, or whether they can express any semantic category. The term "functional" in this respect refers to categorial semantics. There is no reason to assume a priori that reduplication is restricted to certain functions (cf. Moravcsik 1978: 316). And indeed, cross-linguistically 16 These features are all defined in the Graz Database on reduplication (cf. Hurch and Mattes, to appear; also Moravcsik 1978).


Criteria for the classification of reduplication types

the categories which are expressed by reduplication are fairly diverse. They range from plural to diminution (for example in verbs, as Emerillon plural zeka 'break' --> ze~zeka 'split at different points' (Rose 2005: 354), or Afrikaans diminutive vat 'touch' --> vat~vat 'touch tentatively, several times' (Botha 1984: 10), in nouns, as Ilokano plural kailián 'townmate' --> ka~kailián 'townmates' (Rubino 2005: 12), or Lillooet (lil) diminutive s-yáqca 'woman' --> s-y~yqca 'girl' (Shaw 2005: 177), in adjectives, as Nahua (nhn) weka 'far away' --> we:~weka 'far away,


(Peralta Ramirez 1991: 68), or in Ndyuka (njt) lontu 'round' -->

lontu~lontu 'roundish' (Huttar and Huttar 1997: 408)), from possessive marking to object defocussing (e.g. Movima possessive wa:ka 'cow' --> wa:~wa:ka 'to own cows' (Haude 2006: 86), or Fiji e buli 'it was shaped' --> e buli~buli '(s)he was shaping' (Schütz 1985) etc. However, from a cross-linguistic comparative view, "typical" functions of reduplication can be found: The procedure is most frequently associated with the broader categories of plurality, diminution and intensity. In general, these can refer to all word classes and all major lexical classes, although often there are often language specific constraints. In verbs or lexemes expressing events, it is often tense-aspect categories such as continuity, imperfectivity etc. which are marked by reduplication. In nouns or lexemes referring to entities reduplication often marks plural. Furthermore, the procedure is frequently used to derive word classes, mainly adjectives, or to derive new lexemes. For example the denominal adjectives in Fiji (fij), e.g. rere 'fear' --> re:~rere 'fearful' (Schütz 1985), or lexical enrichment in Tok Pisin (tpi), e.g. wil 'wheel' --> wil~wil 'bicycle' (Rubino 2005: 19). So, in some languages, reduplication types can be defined with respect to the word classes with which they appear or which they can create. But in other languages such a classification might cause problems, because of their peculiarities with regard to word classes (as it is the case for Bikol, cf. III.2.6). Similarly, reduplication can be classified as inflection or derivation in many cases, but in some languages and with some examples such distinction is very difficult, if not impossible.17 Most languages, probably all of the ones which have productive reduplication, also possess a certain amount of lexical reduplications, i.e. reduplicated word forms, which have either no corresponding simplex form, or which have only a loose or an arbitrary semantic relation to 17 I do not go into detail for the distinction between inflection and derivation, as this is a general problem of morphological description and not specifically related to reduplication.


Criteria for the classification of reduplication types

its (former) base. Some languages exclusively have lexical reduplication, but no productive reduplication rule. The semantic regularities of lexical reduplication have not received much attention so far, but chapter 5.3. will go into detail on this point (cf. also Mattes and Vollmann 2006). As Abraham points out, notwithstanding the numerous studies on reduplication, the "reduplication-meaning heterogeneity problem" is still unresolved (Abraham 2005: 565). One aim of this study is to provide a contribution to a better understanding of this problem (especially 5.4.2). 2.3 Correspondence between form and function There are multitudinous possibilities for formal as well as for functional types of reduplication and languages make use of these options in very different ways and to very different extents. What is interesting and sometimes puzzling is the way in which the forms and the functions are matched. In the "ideal"18 case of one-to-one correspondence in terms of distinctiveness, i.e. isomorphism, one form would express exactly one meaning and one meaning would be expressed by exactly one form. For example Lampung (ljp) has a clear distinction between intensification by full reduplication and diminution by partial reduplication (e.g. balak~balak 'very large' – xa~xabay 'somewhat afraid', cf. Walker 1976, cited Rubino 2005b: 20). However, most languages do not exhibit such a clear correspondence. Very often reduplicative forms are polysemous or even homonymous. For example, Afrikaans has only full reduplication, but with several different functions such as plural marking, diminutive marking, adjective derivation, etc. (cf. Botha 1984). Tagalog has some different formal types, but nevertheless, they do not refer to one meaning each. One type copies the first two syllables of the base, and this can mark distributive (e.g. dala~dalawa 'by twos') as well as attenuation (ma-tali~talino 'rather intelligent') and reciprocity (mag-kita~kita 'see one another') (cf. Schachter and Otanes 1972).19 Bikol full reduplication is another example of outstanding polysemy, which will be discussed in detail in 5.4.2.

18 "... the old principle that the natural condition of a language is to preserve one form for one meaning, and one meaning for one form" Bolinger (1977: x). 19 The examples suggest that the semantic categories of the reduplicated words depend on the semantic categories of the simplex forms. This is analyzed in detail for Bikol, cf. IV.


Criteria for the classification of reduplication types

Conversely, there are also polymorphies, i.e. various formal alternates for only one function in complementary distribution, as for example CV- and -Vr- reduplication for pluractional in Bikol (cf. IV. My suggestion is, that not only in Bikol, but in general, the variation of forms expressing one function is more systematically organized by clear conditions or restrictions than the variation of meanings expressed by one form, which often can only be disambiguated by the context (cf. IV. And furthermore, as a preliminary result of the typological research in the Graz reduplication project I assume that the one-toone correspondence is rather an exception than the rule.

3 Distribution Reduplication is a very widespread phenomenon, and much more systematically used in language than one might assume from the West-European point of view. Reduplicative morphology is found in genetically completely different languages (cf. Rubino's map 2005a). But in some language families and even in some linguistic areas reduplication appears to an especially great extent. On the Indian subcontinent reduplication seems to be an areal phenomenon. It can be found there in many unrelated languages, even in the Indo-European ones. The Austronesian family in South-East Asia and the Salish family in North America, which are genetically unrelated and geographically distant, but strikingly similar in some respects, are both "typical" reduplicating language families. So are some other American families (e.g. Algonquian, Uto-Aztecan), or African families (e.g. Nilo-Saharan, AfroAsiatic), to name but a few. Furthermore creole languages though developed on the basis of "reduplication-free" Indo-European languages usually have productive reduplication systems.20

20 One explanation for the numerous appearance of reduplication in creoles it has been taken over from the substrate language. However, as reduplication is so common cross-linguistically, an independent development is just as probable as transfer from the lexifier language (cf. Bakker and Parkvall 2005: 516). Moreover, the origin and development of reduplication in creoles underlies additional dynamics, which are not yet fully understood (more detailed see Bakker and Parkvall 2005 and Kouwenberg (ed.) 2003).


Brief history of the research on reduplication

4 Brief history of the research on reduplication 4.1 Typological approaches to reduplication The oldest compilation of data on reduplication in a typological perspective was the noteworthy study by Pott (1862). The author provides a huge collection of data from many various languages of different families, and he divides his book in two major parts: The formal types of reduplication and the functional types ("Verschiedener intellectueller Werth der Doppelung"). Pott's "Doppelung" refers equally to sentences, words, syllables, and individual sounds, as well as to both grammatical and extragrammatical word formation. The amount of examples from American, African and Asian languages is remarkable. Pott's compilation can be considered the most important typological database on reduplication. Nevertheless, it was largely ignored and is rarely cited in studies on reduplication. Some greater attention was given to a much smaller, but comparable typological study on reduplications by Brandstetter (1917). It is also divided into formal and functional parts. In the period between Pott's (1862) and Brandstetter's (1917) cross-linguistic studies, numerous articles on reduplication in Indo-European languages were published. The reduplicated preterit received especially great attention (e.g. Grein 1862, Wood 1895, Bezzenberger 1908, Brugmann 1912/13, Karstien 1921). In her dissertation Kocher (1921), inspired by Pott's and Brandstetter's work, provides a remarkable collection of reduplicated lexemes in languages and dialects of France and Italy. The first study that goes into a detailed description in a non-European language family, namely Indonesian, but also referring to other languages, is Gonda (1950). Like Pott and Brandstetter, the author tries to set up a systematization of the most characteristic meanings of reduplication, namely different variants of plural, intensive and diminution. Furthermore, all the three works point out that reduplicative word structures are often used for names of plants, animals, body parts etc. (cf. also 5.3, and Mattes and Vollmann 2006). Examples of language specific typologies on reduplication include Hestermann (1915) on Serer Sine (srr), Haeberlin (1918) on Salish dialects, Blake (1917) and Lopez (1950) on Tagalog, Haas (1942) on Thai (tha), Anagbogu (1955) on Igbo (ibo), Ansre (1963) on Ewe (ewe), and many more.


Brief history of the research on reduplication

The seminal works on reduplication within the framework of "modern" typology were Moravcsik (1978)21 and Wilbur (1973), the second with a focus on the phonology of the phenomenon. However, with few exceptions, typological studies on reduplication followed only about thirty years later, because in the 80's and 90's the research on reduplication was completely dominated by the new Generative Linguistics (cf. II.4.2). Only in the last few years, with the emergence of Cognitive Linguistics and an newly increased interest in Typological Linguistics, among other things, a new interest in the functional properties of reduplication can be observed, attested by recently published volumes, such as those edited by Kouwenberg (2003) and Hurch (2005). Compared to the restricted data available to linguists from 100 or 150 years ago, new typological studies can now access a much broader database, due to an increasing amount of comprehensive grammars of poorly known languages, in the tradition of descriptive linguistics. 4.2 Reduplication in Generative Grammar Shortly after the publication of Moravcsik it was Marantz' "Re Reduplication" (1982) with with which heralded a very new approach to the topic, with the main focus on its formal properties. The growth of interest in reduplication that followed was motivated by the development of Generative Grammar, for which the phonology of reduplicative processes provides challenging problems. Reduplication was and is a popular object of investigation in Autosegmental Theories, in Prosodic Morphology, and, still very effectively, in Optimality Theory (OT). Although the forerunner of these models, i.e. the famous dissertation on the phonology of reduplication by Wilbur (1973), emphasizes that reduplication is a morphological means and does not belong to the phonology of the language, all these studies treat reduplication exclusively as a phonological phenomenon, without any reference to its functions. This is indeed not surprising coming from within the Generative framework. Beginning with Chomsky and Halle's "Sound Pattern of English", morphology was no longer accepted as a linguistic module of its own. It was either considered to be part of the lexicon or of syntax (split-morphology hypothesis). Morpho-phonological processes, on the other hand, are usually considered to be purely phonological.

21 A more tentative and less perceived typologically oriented paper on reduplication is Key (1965).


Brief history of the research on reduplication

In the following few sections I very shortly summarize some models of reduplication developed in the tradition of Generative Grammar.22 4.2.1 Wilbur (1973) Wilbur, in her dissertation on the Phonology of Reduplication, attends to the "irregular" phonological behavior of reduplication. She not only criticizes the traditional generative approaches to the problem (for example the assumption of a change of the underlying representation, or a change of the ordering of rules), but also offers a new explanation. She mainly describes two phenomena which can be observed in reduplicative phonology, causing "paradoxical" results if a precedence of morphology over phonology is assumed, as is widely accepted since Chomsky and Halle (1968). Wilbur calls these phenomena "overapplication" and "underapplication" of phonological rules. In the first case, a rule applies in an environment in which it should not apply, and in the second case, a rule does not apply although the environment would require it to apply. The innovative idea within a generative framework is that her explanation lies outside the domain of phonology. Wilbur recognizes that the "exceptional" phonological behavior of reduplication is based on a morphological structuring principle of language, i.e. the tendency towards the identity of the reduplicant and its corresponding base. She introduces the term "Identity Constraint" for this tendency (Wilbur 1973: 57-59). Later, Optimality Theory (cf. II.4.2.4) adopts this principle as its major element. For more details on the Identity Constraint, cf. Excursus II. To Wilbur's remarkable merit she is very clear about the morphological nature of reduplication. Unfortunately, most other generative models do not bear this matter in mind, but treat reduplication as a phonological rule. 4.2.2 The copy-and-association model Marantz (1982) provides a new view on reduplication, setting it against the background of the Autosegmental Theories (Goldsmith 1976, McCarthy 1979 and 1981). Central to Marantz' approach is the assumption that reduplication is just a special type of affixation. The peculiarity of reduplication is that instead of a fully phonologically specified morpheme, a skeletal morpheme, i.e. a segmentally empty template, is attached to the stem. This skeleton, 22 I do not go into much detail for these models, because they are already repeatedly summarized in other dissertations on reduplication, e.g. Niepokuj (1997), Spaelti (1997), Raimy (1999).


Brief history of the research on reduplication

which can be defined by C and V-slots, or by syllabic shapes, is then filled ("associated") with melodic content which is copied as a string of segments from the base. The filling of the skeleton is explained by autosegmental spreading and templatic phonology. This means that after the affixation of the CV-skeleton the melody of the base is associated to this skeleton and, if necessary, all non-associated elements are deleted, as in the following example: takki




= CVC+



Figure 1. Marantz' description of Agta initial CVC-reduplication (Marantz 1982: 445)

This is the basic model, and naturally Marantz has to establish additional conditions which govern the association of the melody to the skeleton. (For more detailed summaries of the approach cf. Steriade 1988: 85-93, Niepokuj 1997: 54-57; Raimy 1999: 20-25, Wiltshire and Marantz 2000: 561-562). Marantz delivers a model which permits a formalization of reduplication, but, as Raimy (1999: 21) points out, it does not add anything to a better understanding of reduplication as a morphological process and its phonological properties. Furthermore, an essential and often criticized weakness of the model is, that it can easily explain types of reduplication which are very "unnatural" and indeed are not attested in language. The basic assumption of the Marantz model (and other "copy-and-association" models, e.g. McCarthy and Prince (1986)) is that a reduplicant is stored as a skeletal morpheme in the lexicon of the speaker and is available for word formation. Its shape is independent of the base it is affixed to. Since such a model assumes reduplication to be the attachment of an independent affix, a big problem, among others, is to explain the abundant instances of reduplications which cannot be well described with morpheme-based rules, but only with word-based rules. Niepokuj (1997: 56-57) mentions an example from Tarok (yer), which is "prefixal" in some cases, but "suffixal" in others23 (e.g. a-wó 'hand' --> a-wú~wo 'his/her hand', but a-ríjìyá 'spring' --> a-ríjìyá~jiya 'his/her spring'). An example from Bikol is imperfective reduplication. Originally it was always the root initial syllable which was reduplicated. In modern speech however, more often than not it is the second syllable of the derived word which is reduplicated, irrespective of whether this syllable belongs to the root or 23 Prefixal for monosyllabic roots, but suffixal for polysyllabic roots.


Brief history of the research on reduplication

to the prefix, e.g. naka-bo~bohe = naka~ka-bohe 'managing to escape' (example (62) or nagpa~i-isog = nag-pa~pa-isog 'making brave' (example 63), (cf. IV. For a non-linear phonological study of Tagalog reduplication and other morphological procedures on the basis of Marantz (1982) and McCarthy and Prince (1986), cf. French (1988). 4.2.3 The full-copying model Based mainly on McCarthy and Prince's approach (1986), Steriade (1988) develops a "syllabic transfer model" for reduplication. She assumes that exact full reduplication underlies every kind of reduplication. The material which is not needed in types of partial reduplication ("disallowed units", Steriade 1988: 81) is eliminated. Fixed segments result from the insertion of prespecified material into the reduplicant. The basic unit in Steriade's model is the syllable. Any modifications to the underlying full reduplication are described as operations of syllabic adjustment, guided by two kinds of parameters: the weight parameters and the syllabic markedness parameters. I.e. the reason for truncation of full reduplication is the reduction of certain marked structures. One example therefore is the reduction of the reduplicant to the core syllable (CV), which is indeed a widely documented phenomenon (for details cf. Excursus III). In Sanskrit, for instance, the reduplicant kan- in kan~i-krand- 'cry out intensely' combines a simplification of the marked onset and rhyme of the base syllable krand-, which furthermore results in a reduction of the syllable weight (Steriade 1988: 81). The more differences there are between the base and the reduplicant, the more parameters have to be set. In this sense, full exact reduplication is predicted by the model to be the most common pattern because it requires no parameters. One problem of this model is that by setting up a hierarchy of parameters the resulting hierarchy of preferences of reduplication types does not mirror the distribution of actually occurring ones (i.e. the very common pattern CV- requires three constraints: weight, complex onset, and coda, cf. Raimy 1999: 30). Both, the copy-and-association model and the full-copying model, do not offer any explanation for interesting phenomena which are particular to reduplication, namely overapplication and underapplication (cf. above, Wilbur 1973). The reason is that these models are rule based and as such have a conventional view of rule ordering. And this is exactly what is severely challenged by the mentioned phenomena. A solution for this problem 16

Brief history of the research on reduplication

was offered by the Optimality Theory that emerged in the 90's which abandons the assumption of rules. (For details on overapplication and underapplication in OT and elsewhere cf. Excursus II). 4.2.4 Reduplication in Optimality Theory Opposed to rule based theories, OT (founded by Prince and Smolensky 1993) is a theory based on constraint interactions. It does not assume any rules for the generation of phonological forms. Basically, it assumes that all morphological and phonological parameters are available simultaneously. Rather, a number of "candidates" for the output of a word formation process are generated and evaluated with the help of hierarchically ordered constraints. The "optimal" candidate is the one which violates the fewest and the lowest ranked constraints, and is therefore selected as the output. Some important constraints with respect to reduplication are the "Identity constraint", the "Faithfulness Constraint", or the "Base-Reduplicant Correspondence Constraint" (cf. Excursuses II and III). The ranking of the constraints is language specific. OT is an especially attractive model for reduplication, because it offers a solution to problems which the phenomenon causes in other theories, as for example rule ordering. But rather than explaining the exceptional behavior of reduplication, OT provides an alternative method of formalizing it. The constraint ranking is deduced from the existing output form (which is compared to fictive output "rivals"), and as such is circular. I do not go into more detail here, because some specific contributions of OT to the theory of reduplication are discussed more elaborate below in Excursus II, Excursus III and Excursus VII. 4.2.5 Raimy's Readjustment Rule Rejecting the assumptions of OT, Raimy (1999, 2000) proposes an alternative theory of reduplication as resulting from a readjustment rule, triggered by a zero-affixation. He maintains that morphology completely precedes phonology by assuming a zero-affixation as the morphological part of reduplication and the readjustment rule as the phonological one. Raimy proposes a so-called "loop representation", an illustration of rules for the realization of the segments sequence in the word-formation process. His basic idea is that, as a result of the readjustment rule triggered by reduplication, a word final segment at the same time precedes


Brief history of the research on reduplication

the word initial segment. This is illustrated as following by the author: #--> k --> æ --> t --> % = [k æ t k æ t]

Figure 2. Raimy's loop representation of kætkæt (Raimy 2000: 1)

The phoneme loop is followed by a linearization process, which takes place during the transfer of a phonological representation to the articulatory system. The result is a serial sequence of base and reduplicant. For serial based theories the examples of overapplication or underapplication where the reduplicant affects the phonological appearance of the base are the biggest problem (cf. for example McCarthy and Prince 1995: 41). In Raimy's model these cases do not cause problems, since in the loop representation the base has the same phonological environment as the reduplicant, which regularly triggers a phonological change (or not) in both units. Therefore Raimy rejects the necessity of concepts such as "overapplication" and "underapplication" being not necessary in a theory of reduplication. In his view, the phenomenon that is described by these terms is simply an effect of opacity, resulting from the linearization process of the loop (for details cf. Raimy 2000: 16-23). Raimy (2000) tries to find a solution to the problems which are not satisfactorily solved by other theories of reduplication. However, for this aim he has to interpret reduplication as a phonological readjustment rule, following a zero-affixation. There is no evidence which would justify the assumption of such a zero-affixation, but this is only a tool, introduced by Raimy, in order to be able to treat the morphological means of reduplication on a purely phonological level. As already mentioned, the purely phonological view on reduplication is the big weakness of all of the mentioned generative approaches. As an answer to this, Inkelas and Zoll (2005) have developed an alternative model which makes allowance for the morphological nature of reduplication. 4.2.6 An alternative approach: The Morphological Doubling Theory (MDT) Inkelas and Zoll (2005) view reduplication as a purely morphological construction whose "daughters" are constrained to be morphosemantically identical.24 The two daughters are both 24 Inkelas and Zoll's "daughters" are usually labeled "reduplicant" and "base" in other theories. The "mother"


Brief history of the research on reduplication

generated by the morphology, embodying semantic and phonological information. The daughters are each subject to individual phonological changes. But additionally, they are linked by a shared "cophonology". I.e. the phonological shapes of the daughters are concatenated and these are again subject to a third "cophonology" at the mother node that produces the surface form. The key assumption of MDT is that the two units are treated as independent from each other, i.e. the reduplicant is not generated from the base. The daughters are primarily semantically identical. A phonological identity is secondarily created by the "cophonology". In partial reduplication one daughter is assumed to be phonologically truncated. Mother (meaning = some function of the meaning of the daughters; phonology = some function of the phonology of the daughters) Daughter #1 (meaning = that of Daughter #2; may be subject to special phonology)

Daughter #2 (meaning = that of Daughter #1; may be subject to special phonology)

Figure 3. Inkelas and Zoll's basic schema of the Morphological Doubling Theory (cf. Inkelas 2005: 65)

The attractiveness of this model definitely lies in its focus on morphology. Opposed to all preceding generative theories, MDT treats reduplication as what it definitely is, i.e. a morphological procedure. The theory has no need to explain phonological deviations of the two reduplicative constituents, because they do constitute two basically independent morphological units. Although this model is very enriching for the concept of reduplication, I doubt its full adequacy on the grounds that I consider the formal dependency of reduplicant and base as essential to the phenomenon. Opposed to the descriptive compilations of reduplication types which preceded the mentioned theories, these models try to formalize of the highly complex morphological phenomenon of reduplication which sometimes has complicated phonological properties. The question is however, if these models really provide more insight into the nature of reduplication or if they corresponds to the simplex form.


Brief history of the research on reduplication

are in fact only models of formalization of the phenomenon? Each model reflects interesting ideas of how to describe the puzzling phenomenon of reduplication. And, to different degrees, each of them creates a large apparatus of rules, tables, etc., which finally can correctly generate the forms which are indeed observable in language. However, these formalizations with their abundant additional parameters have no explanatory value. For this reason I did not generally adopt any of these theories in the following description of the reduplication types of Bikol. Basically, for reasons that will be argued later in connection with relevant examples, I do not consider reduplication to be just a specific case of affixation, and I do assume the reduplicant to be dependent of the base in sound shape. But I reject templates or full copying and truncation as models that reflect language processing in humans. In my opinion, these are models which indeed enable us to formalize the reduplicative phenomenon, but they do not explain or describe it in a cognitively realistic way. 4.3 Central questions/problems in the research on reduplication There are some very central questions inherent to reduplication, which all of the mentioned theoretical approaches try to solve, albeit in very different ways. Already mentioned several times was the question of the morphological status of reduplication: Can it be described as a special kind of affixation, or must it be conceived as a morphological procedure in its own right? And if reduplication is seen as a kind of affixation, how can one argue for an affix without a constant phonological form? The proposed solution for this "lack" of specification, offered by McCarthy (1981) and Marantz (1982) and widely accepted since then, is an affix which is only partly specified (prosodically, for example) and then filled with segmental material by a copying process ("copy-and-association model"). This model can describe some of the existing reduplication types, but has severe problems with others, such as discontinuous reduplication, full reduplication without prosodic restrictions, etc. Another central question that is relevant in both a synchronic and a diachronic treatment, is the relation between full and partial reduplication. Is it really a universal implication that all languages which have partial reduplication also have full reduplication? And, related to that, another question of whether full reduplication always underlies partial reduplication in the 20

Brief history of the research on reduplication

historical development, as assumed by Bybee et al. (1994) and Niepokuj (1997)? With respect to the synchronic language processing: Is partial reduplication a truncation of full reduplication, as proposed by Steriade (1988) in the "full-copying model"? Based on the available data, I strongly suggest that the hypotheses on the relationship between full and partial reduplication need critical testing, and that, especially for diachronic dimension, alternative explanations are needed. A further issue, which is often not directly discussed but always implicitly referred to, is the iconicity of reduplication. Reduplication is considered to be one of the most iconic procedures in language, because more of the same form compared to the base expresses, in most cases, more of the same meaning of the base. However, this convincing principle, which is assumed to be a constitutive one in reduplication, can lead to confusion when deviations are observed. For example when reduplication expresses less of the meaning of the base instead of more, or when the formal transparency is severely reduced. The important questions in this respect are first, in what exactly lies the "iconicity" of reduplication, and second, if it is indeed reasonable or necessary to expect reduplication to be more iconic than other morphological principles? "Reduplicative plurals illustrate iconicity in an obvious way" (Newmeyer 1992: 763). Virtually every work on reduplication that in some way gives attention to the functional level of the phenomenon implicitly or explicitly says something about iconicity and plurality. But it turns out that both terms which are used mostly as a matter of course actually need more clarification. The iconicity of the most frequently observed plural reduplication leads to the question of the semantics of reduplication in general. I.e. are the meanings expressed by reduplication arbitrary or can they be subsumed under one general concept? In other words, is there one general semantic aspect which is universally related to reduplication, and all specific meanings are somehow derived from that superordinate category? The research on reduplication based on empirical data has increased considerably in the last decades. However the area of diachrony is quite unexplored with respect to reduplication. The reason for this is obvious: Reduplication is predominantly a grammatical feature of non-IndoEuropean languages, many of which have only oral tradition. This means that the data on these languages are in most cases only recently collected, and with few exceptions, there are 21

Brief history of the research on reduplication

neither any data from former language stages nor enough comparative data available, which would be necessary for insights into the diachronic developments as well as for reduplication in a synchronic view. Having said that, Austronesian is relatively well documented and there has already been some reconstruction of the lexicon and the morphology done on this family (cf. for example Blust 1998, 2001, 2003). Fortunately, in the last years, the database for such research has started to expand due to the growing field of language description and grammar writing and, connected with this development, the growing interest in typological questions. With respect to reduplication some of the unresolved typological questions are: –

Are there any universal constraints on possible bases for reduplication, e.g. word classes, semantic categories, etc.?

Do certain reduplication types – formal as well as functional – appear more often than others, i.e. are there any universal preferences for certain types?

Are there any correlations between reduplication types and other typological properties of the language (e.g. word order, morphological type, etc.)?

Are there any implications with respect to the semantics of reduplication (e.g. "If a language has reduplication, one of the meanings is 'change of quantity'", or "If a language has diminutive reduplication, it has also augmentative reduplication", as proposed among others by Uspensky (1972: 69-70)?

Such questions can only be answered by a comparison of many different reduplication systems. The problem with such comparative studies is that the information on reduplication that can be found in reference grammars is usually very superficial. This causes severe problems for a specific study on reduplication which includes a larger pool of languages. This dissertation addresses the open research questions, to the extent possible in a language specific study.



CHAPTER III: BIKOL 1 The language and the data Bikol is spoken in the south of Luzon Island of the Philippines, in the provinces Albay, Camarines Norte and Sur, Catanduanes, and Sorsogon25. Bikol is a Central-Philippine language, and as such belongs to the Western-Malayo-Polynesian branch of the Austronesian family and is closely related to other Central Philippine languages, such as Tagalog, Cebuano (ceb) and Hiligaynon (hil)26. Historical linguists have reconstructed different sub-branches of Proto-Austronesian (PAN)27 (Proto-Extra Formosan (PEF), Proto-Malayo-Polynesian (PMP), etc.) as hypothetical antecedents of the Philippine languages (cf. for example Blust 2003, Dahl 1976; on the controversy of the Austronesian genealogy cf. Fincke 2002: 29-31). Although the relatedness of these Central languages is undoubted, the precise genetic affiliations are not clear, i.e. Austronesian linguists differ considerably in their evaluation of whether Bikol or Tagalog is closer to the Bisayan group than to each other or if the two languages are closer to each other than to any other Bisayan language, etc. But for this study it is sufficient to rely on the grouping of the four languages to the Central Philippine group, as opposed to the Northern Luzon languages such as Ilokano and Agta (agt)28 which separated earlier from the PEF languages. A summary of the sub-groupings of some major Philippine languages which are referred to in the following sections, is presented in figure 4, based on Fincke (2002: 31).

25 Cf. map 2, appendix A. 26 Cf. map 1, appendix A. 27 The capital letter codes are the acronym which will be used in the following text for the different earlier language stages. 28 Cf. map 1, appendix A.


The language and the data

Proto-Austronesian Proto-Extra Formosan Proto-Central Philippine Proto-Bisayan Ilokano






Figure 4. Genetic affiliation of some Philippine languages

The oldest available document of the Bikol language is a dictionary composed by Marcos de Lisboa, a Spanish Franciscan missionary who settled in the Philippines. He collected the language material in the middle of the seventeenth century, but the dictionary was first published in 1754 in Manila. The first grammar notes on Bikol as well as an updated dictionary were compiled by Mintz (1971), and Mintz and Britanico (1985). The first study of the dialects of the Bikol area is McFarland's dissertation (1974). The first dissertation with a focus on conversational Bikol is Fincke (2002). Current research on Bikol, and especially the collection of texts, is carried out by Lobel (cf. Lobel and Tria 2000; Lobel 2004) and Mintz (cf. 2004, 2005). Today, Bikol has about 3,5 millions of speakers (Ethnologue, census 1990). The language is loosing domains rapidly, predominantly because of the strong influence of the two national languages Tagalog and English (or "Taglish", the mixture of both, respectively), which are the only languages used in official (i.e. also educational) contexts and in mass media. Because of its low prestige, children are properly "prevented" from learning Bikol in order to improve their chance of success, especially in the social middle class. But fortunately, some groups consciously make an effort to keep Bikol "alive". There is one radio station in the city of Legaspi broadcasting partly in Bikol, and there is one weekly regional newspaper, "Bicol Mail", printed in the city of Naga, each issue of which contains at least one page written in Bikol, alongside Tagalog and English, the usual languages used in mass media. Furthermore there are some poets writing in their regional dialect, and in fact, one printed collection of 24

The language and the data

Rawit-Dawits29 from 2005 and some hand written poems in Bikol are part of my corpus. Bikol shows a considerable diversity of dialects. The main dialects are Naga and Legaspi (which is considered as "Standard Bikol"), corresponding to the two most important commercial towns in the Bikol region30. McFarland in his dissertation (1974) has documented the diversity of Bikol dialects, taking into account lexico-statistical and grammatical criteria. Figure 5 reflects his findings and is taken from Fincke (2002: 33).

N. Catanduanes

S. Catanduanes

Naga Legaspi Daraga





Figure 5. McFarland's subgrouping of nine Bikol dialects

The corpus which I am using for my analysis is a collection of recordings of spontaneous speech (natural dialogs as well as story telling), elicitations and some written contemporary poems, plus one academic text on the Bikol language by Jose Maria Carpio (2000). Additionally, I took some examples from the written corpus composed by Jason Lobel (Bikol CD-Rom Reference Set). The data was recorded during two fieldwork trips, ten weeks altogether, to the Bikol region in 2005 and 2006, in the city of Legaspi, the province of Albay, and in the district of Pilar (Calongay, Dao), the province of Sorsogon, where the "Legaspi dialect" is spoken. Additional data come from elicitation sessions with a native speaker of Legaspi dialect in Graz. The recordings contain speech from people whose age ranges approximately between 30 and 75 years. The Rawits-Dawits are written predominantly by students in their twenties. In addition to the data which I collected by myself, I extracted a long list of reduplications from Mintz and Del Rosário Britanico's dictionary (1985), which is of great relevance for the following analysis of reduplication.

29 Rawit-Dawit is the name for poems or verses. 30 My own data is from Legaspi dialect, the data in Mintz and Del Rosario Britanico 1985 and Mintz 2004, is the Naga variety. The written text corpus (i.e. the Rawit-Dawits and the collection of Lobel) contain different Bikol varieties.


A short grammar sketch of Bikol

2 A short grammar sketch of Bikol In the following sections some important features of the language will be described briefly. They are considered necessary to get a basic knowledge of Bikol and to understand the subsequent argumentation and especially the cited example sentences. Within this overview, the issue of word classes is touched upon since in the common literature there is much disagreement on the topic within the description of the Austronesian languages. In order to argue for the terminology which I am using it is useful to go into somewhat greater detail with this respect. I do not provide an exhaustive grammar sketch of Bikol here, mainly because there already exist some (even if incomplete) grammatical surveys of the language (Lobel and Tria 2000, McFarland 1974, Mintz 1971 and Mintz 2004). 2.1 Phonology and Morphonology 2.1.1 The Phoneme inventory The native Bikol phoneme inventory consists of three vowels, four diphthongs, fourteen consonants and two semivowels. As table 3 shows, Bikol originally had no labio-dental and no palatal consonants at all. With the borrowing of Spanish and English words, nine additional consonants entered into the Bikol sound system. For example fricatives like /f/ or //, or the affricates /t/ or /d/ were introduced into the language, although some speakers do not pronounce them as such but substitute them with "native" phonemes. 31 Therefore they are marked with brackets in the following charts of phonemes. The original sound system of Bikol contained three vowels: /i/ - /a/ - /u/. But via the Spanish loanwords, a five vowel system was introduced into the language, although it should be noted that [e] and [i] as well as [o] and [u] in spoken language cannot be considered as phonemically distinct. The five vowels are represented in the spelling, but inconsistently (see below). In the following charts the “foreign” phonemes are indicated in brackets.

31 The pronunciation of the borrowed phonemes depends strongly on the speaker's level of education and the associated level of knowledge of English as a foreign language.


A short grammar sketch of Bikol









(o) a

low Table 1: Chart of vowels

front high





mid low

ai au

Table 2: Chart of diphthongs32







dental) stop





affricate (f)






















? h



Table 3: Chart of consonants

For a more detailed description of the Bikol sound system and its history cf. Lobel and Tria (2000: 7-11) and Mintz (2004: 9-17). 2.1.2 Nasal assimilation With prefixes ending in the nasal //, assimilation with the base initial consonant takes place depending on the place of articulation. For example the mang-/pang- prefix series (for generalization of an action) is realized as following: When attached to a base with an initial 32 All diphthongs are falling.


A short grammar sketch of Bikol

bilabial consonant, // and the consonant fuse to /m/. bakal 'buy': mang + bakal --> mamakal 'to go shopping'. When attached to a base with an initial /s/, // and /s/ fuse to /n/. sublí' 'borrow': mang + sublí' --> manublí' 'to go around borrowing'. When attached to a base with an initial /k/ or /h/, // and the initial consonant fusion to //. ha'bón 'to steal': mang + ha'bón --> manga'bón 'to go around stealing'. When attached to any other consonant, no assimilation takes place: ta'ó 'to give' --> mang + ta'ó --> mangta'ó 'to give away things generally' (cf. Mintz 2004: 35-36). However, these phonological processes apparently are subject to some variation. For example in my corpus I found the form nangubod 'to generally believe', from tubod 'believe'. Following the rule given by Mintz (2004), it should have the form nangtubod. Regional variation might be a reason for these inconsistencies, but also factors such as speech tempo or speech style. 2.1.3 /h/-epenthesis If a vowel initial suffix is attached to a vowel final base, /h/ is inserted between base and suffix (in some Bikol dialects a glottal stop instead of the fricative is inserted). For example totoo 'true' + nominalizing ka- -an --> ka-totoo-han 'truth' (cf. example (61)), ma-ogma 'happy' + intensifying -on--> ma-ogma-hon 'very happy' (cf. example (67)), or duwa 'two' + limitative CV- -e --> du~duwa-he 'exactly two' (cf. example (81)). 2.1.4 /r/ and /l/ /r/ and /l/ are distinct phonemes of Bikol. There are for example the minimal pairs: ragos 'to lack steps (a ladder or stair)' vs. lagos 'to tie', dalas 'quick' vs. daras 'piece of fish/slice of meat'. Both phonemes are reconstructed to as early as PAN by Dahl (1976: 101). However, /r/ and /l/ in the Philippine languages undergo widespread variation. There are many lexemes in related languages and dialects which only differ by the selection of /r/ and /l/. For example kalabasa (Tag.) vs. karabasa (Bik.) 'pumpkin', dalaga (Tag.) vs. daraga (Bik.) 'maiden, young lady', etc. But also within one language /r/ and /l/ are exchanged in certain circumstances. One example is the metathesis of /r/ and /l/ in case of -Vr- plural infixation33. Whenever the plural -Vr-infix is applied to a /l/-initial word, the metathesis takes place. E.g. luto ‘cook’ (--> *l-ur~uto) --> r-ul~uto ‘cook, pl.’, layog 'fly' (--> *l-ar~ayog) --> r-al~ayog 33 For explanation of the -Vr-plural infix see IV.4.1.4.


A short grammar sketch of Bikol

'fly, pl.' (cf. example (88)). The result of a surface analysis of my data (including Mintz' dictionary 2004) is that even in underived Bikol native words /r/ always precedes /l/, never the other way round.34 The scope of this phonotactic constraint is surprisingly large: It seems to concern the entire word. For example baralagat 'describing a road crossed by many trails', rungkal 'weigh down to keep it steady', tarawal 'a bar consisting of a pole or length of a bamboo placed across a window or entry porch', while no native words with the reverse combination l...r were found. This finding, together with the observation that in the case of -Vr- infixation metathesis of /r/ and /l/ takes place, allow to formulate the constraint that in Bikol words the alveolar flap /r/ can never follow the phonetically very similar alveolar lateral /l/, even at the interval of several segments and syllables35. 2.2 Prosody 2.2.1 Syllable Structure The Bikol syllable structure is basically CV(C), where V can also stand for a diphthong. Tautosyllabic consonant clusters only appear in loanwords and foreign words, for example trabaho 'work', prutas 'fruit', green 'green', etc. The strong tendency toward the CV-pattern can be seen in the case of reduplication. When a syllable with a consonant cluster is reduplicated it is obligatorily reduced to the core syllable CV. For example trabaho 'work' --> nag-ta~trabaho 'is working', treno 'train' --> nag-te~treno 'going by train' (cf. IV. 2.2.2 Stress Bikol is a syllable-timed language. This means that we do not find vowel reduction processes. The vowels are fully pronounced in both stressed and unstressed positions. Bikol has lexical stress, which can distinguish meanings. Thus, there are minimal pairs differentiated only by stress. For example: dá.pog 'seed bed for rice' and da.póg 'hearth, kitchen'; lí.pat 'to move, to 34 Some comparative data suggest a general preference for /r/ over /l/ in Bikol. The language has many lexemes and morphemes containing an /r/ where related languages have an /l/. E.g. 'girl': Bik. daraga – Tag. dalaga; 'pumpkin': Bik. karabasa – Tag. kalabasa; plural reduplication for bases with more than two syllables: Bik. Curu- - Ceb. Culu-, etc. 35 I am fully aware that the mere non-existence of a phenomenon is not a proof of its impossibility. However, considered together with the metathesis rule, such a constraint is at least very probable. All exceptions to this constraint which were found are loan words, predominantly of Spanish origin, for example altar < Sp. altar, bulgar 'coarse' < Sp. vulgar; laringhitis < Sp. laringitis; also: daldalero 'a gossip' < Bik. daldal 'talkative' + Sp. agentive ending -ero.


A short grammar sketch of Bikol

transfer' and li.pát 'to hit s.o. without warning', tá.ta 'title for father, grandfather, uncle' and ta.tá 'door' (cf. Mintz 2004). With one prosodic form however the word stress is predictable: If a lexeme consists of two closed syllables, stress is always on the second syllable. For example tul.sók 'to poke'; singsíng 'ring', gub.tík 'to jump', hi'.bóg36 'thick' etc. 2.2.3 Stress shift Stress shift plays a role in various derivation processes37. First of all, there exists one type of derivation, namely derivation of properties and states, which is realized exclusively by stress shift, as for example búhay 'life' --> buháy 'alive' (cf. III.2.5.4). In additive derivational operations, the affixations cause a movement of the stress of the base: When a voice-suffix is added to the base, stress shifts one syllable to the right. For example: hapót 'ask a question' – hapot-ón 'ask a question-UG', háli' 'remove' – halí'-on 'remove-UG', or túkaw 'sit' – tukáw-an 'sit-UG'. Compared to voice suffixes, the intensifying suffix -on38 behaves differently. Irrespective of the stress pattern of the unaffixed root, word forms with the intensive suffix -on always have stress on the penultima. This also holds true when the suffix appears more than once39: ma-sirám 'delicious' --> ma-sirám-on 'very delicious' --> ma-siram-ón-on 'very very delicious'; ma-lípot 'cold' --> ma-lipót-on 'very cold' --> ma-lipot-ón-on 'very very cold'. When a prefix is added to a base or when the initial CV-sequence is reduplicated for imperfective aspect, stress is assigned to every second syllable to the left, starting from the stress of the base. For example ta.ó ‘give’ --> tá-ta.ó ‘IMPFV-give’ --> i-tá~ta.ó ‘CV-IMPFV~give’ --> í-t{i.n}á.~ta.ó ‘CV-{BEG.UG}IMPFV~give’. This means that the reduplicant does not receive stress when the stress of the base is on the initial syllable, as in há.li' ‘remove’ --> ha~há.li' ‘IMPFV~remove’. When there is a suffix in addition to the prefix or the reduplication, the original lexical stress shifts one syllable to the right, but without causing any change to the preceding stress. For example a.pód 'call' --> á~a.pód 'IMPFV~call’ --> á~a.pod-ón 'IMPFV~callUG’;

ha~há.li 'IMPFV~remove’ --> ha~ha.lí-on 'IMPFV~remove-UG’ (cf. Mintz 2004: 18).40

36 37 38 39

The apostrophe < ' > stands for a glottal stop, cf. III.2.3. On derivation and inflection in Bikol see III.2.5. Intensive suffix -on is homonymous to the undergoer suffix. The recursive application of intensifying -on is theoretically possible without limitation. In spoken speech it frequently occurs with three or even four repetitions. 40 These stress rules were not checked by myself, I only refer to Mintz (1971 and 2004). Especially the last rule, where the stress shift “stops” at the root initial syllable, seems somewhat awkward and demands for explanation. A careful prosodic study of Bikol (as of the Philippine languages in general) is missing so far


A short grammar sketch of Bikol

2.3 Spelling There is no standard orthography for Bikol, but it is usually written based on the Tagalog orthography. Before the arrival of the Spaniards41 in the Philippines there existed syllabaries. The original Bikol syllabary was called Baybayin or Alibata. It consisted of eighteen symbols for syllables (cf. Lobel and Tria 2000: 1):

The Spaniards introduced the Latin alphabet, which is the only script used today. In addition to the common symbols, there is the glottal stop which is sometimes represented by < ' >, but only in syllable final position; or by a hyphen in internal position. For example , [ba?.gó] 'new' or [?á.ki?] 'child'. A glottal stop between two vowels is often not represented in the orthography. I.e. two adjacent vowels in spelling are separated by a glottal stop in pronunciation. For example: 'eat' is pronounced regularly as [ká.?on], 'take' as [ku.?á]). Because of the lack of a standardized orthography, there is much inconsistency in writing, especially of vowels and diphthongs. The alphabet provides five vowels, but, as already mentioned above, there is no phonemic distinction between [e] and [i], nor between [o] and [u]. Therefore, there is noticeable variation in this respect. For example the reduplicant Curu(cf. IV.4.2.3) can sometimes be found as . I will use throughout this work, except for examples taken from written language, i.e. the printed and the hand written RawitDawits42. The /au/-diphthong is usually spelled , for example 'sun, day' or 'water buffalo', the /ai/-diphthong is spelled , as in 'dance' or 'crazy'. The /ui/-diphthong is spelled or , as for example in and would doubtlessly be very illuminating in several respects. 41 The Philippines were discovered by Ferdinand Magellan in 1521, but the Spanish settlement started with Miguel Lopez de Legaspi in 1565. 42 My corpus contains some contemporary poems from different authors (see above). Although all these authors are well-trained, their orthography differs considerably. The reason for this is that Bikol is not a taught subject at any stage of education.


A short grammar sketch of Bikol

'swim' or 'shiver, tremble' and the /iu/-diphthong is written when it occurs in an open syllable, as in 'snare, spring trap'. In a word-final closed syllable it is spelled , as in 'bend (a limb)', and in a word-medial closed syllable , as in 'honeybee' (cf. Mintz 2004: 11). No strict orthographical regulations exist for the affricates either. I use the spelling for // as in [a] 'he/she' or [am] 'nine' and for /t/ as in [kote] 'car', following the most common way, based on the Tagalog orthography. /N/ is spelled , as for example in [unyan] 'now, today' or [bagi] 'evening, night'. Full reduplication is often written with a hyphen (for example in "Basa-basa an buhok mo; ..."43 'Your hair is soaking wet ...' or the reduplication in "... pinangat na maharangharang ..."44 '... the Pinangat45 which is a little bit pungent ...'). This is also the way I represent full reduplication in this work. But there is also much of variation in this respect. Sometimes reduplicated forms are represented as one word (for example "An magayagayang aldaw ...."46 'The very pleasant day ...'), sometimes as two words (for example "...kinonot na naglalana lana"47 'in very creamy coconut milk'). Different spellings are found even within one text by one writer. Some writers also separate other affixes by a hyphen, others do not use the symbol at all. Partial reduplication and Curu-reduplication are never marked with a hyphen in my written corpus. For a more detailed description and an explanation of the historical changes of orthography cf. Lobel and Tria (2000: 3-5), Mintz (2004: 4-9). 2.4 Syntax At a syntactic level different types of phrases in Bikol can be easily distinguished from each other because they can be identified unambiguously by their specific markers and/or their position: predicate (function of predication), predicate base and other arguments or adjuncts 43 44 45 46 47

From corpus [agom]. From corpus [r-d: kaon]. Pinangat is a typical Bikol vegetable dish ("Taro leaf wraps"), prepared with coconut milk. From corpus [pakipagmootan]. From corpus [r-d: kaon].


A short grammar sketch of Bikol

(function of reference) and modifier (function of attribution). The terminology used for what I here call "predicate base" and "argument" varies considerably in the literature. What Himmelmann (1987 and following) introduces as "predicate base" is traditionally referred to as "subject", and what he labels an "argument", is usually named "object". If these traditional terms are used48, the writer and the reader have to be aware that the Philippine morphosyntactic system differs to a high degree from the Indo-European system for which these terms have been created. Himmelmann (1987) delivers a very convincing new analysis of a Philippine language (Tagalog), which I have largely adopted here for Bikol. In my opinion it describes and explains the underlying system much more appropriately than the "classical" concepts of phrases and word classes. The concept of "subject" is very strongly associated with agent and nominative, as well as with topic, what does not apply to the same extent to languages such as Bikol or Tagalog. For example the agent is expressed by the predicate base just as often as for example a patient or an action/event. In order to point out the differences, some authors refuse to use the term "subject". The reasons for the choice of the term "predicate base" will become clearer with the concrete examples. It reflects the strong relationship to the predicate: The semantic role of the predicate base is indicated in the predicate by the "voice"-affixes (cf. below in detail). The order of the phrases within the sentence is relatively constant, but not at all rigid and obligatory. Reordering is especially relevant for negation and topicalization (cf. for example Fincke (2002: 48-49)). In the unmarked case, the predicate has sentence initial position, followed by (clitic) pronouns and particles. The next position is usually taken by the predicate base (also referred to as ang-Phrase)49. It is followed by other arguments or subordinated phrases. In spoken language, arguments are often not expressed. Even predicate bases can be omitted. Bikol can be categorized as an agglutinating language, but it has some inflectional features like syncretism or fusion (cf. Bossong 2001). Many of the morphemes are either cumulative or multifunctional and additionally there are several homonymous affixes (for example -on 48 Himmelmann (to appear) himself goes back to using the term "subject" in recent publications, because it is much more commonly used in Austronesian linguistics. 49 If the predicate base consists of a pronoun, this appears in a specific form (glossed as 'ang'-form ( AF) in the examples here but often glossed as nominative in the literature). See for example (19): 2SG is attached as a clitic to the predicate; the 1SG in the 'ang'-form ako is found as a free morpheme in the position of the predicate base.


A short grammar sketch of Bikol

for undergoer voice and for intensive marking, -an for undergoer voice and reflexive marking, etc.). Bikol has no copula. It has no agreement marking and no grammatical gender, and, as Tagalog (cf. among others Naylor 1980), its sentences have equational character, as the notation of the literal translation of sentences (1) and (2) below shows. Propositions as "The water is very cold." (1) and "My father bought a big house." (2) are expressed by the same sentence structure, and, as can be seen from example (2a) and (2b), the referents of the predicate and the predicate base can be simply exchanged.50 Generally speaking, in Bikol there is an alternating sequence of content words and function words. Each argument of the sentence is preceded by a phrase marker, each modifier is connected with its head by a linker. I.e. juxtapositions are not possible. The following examples should serve as an illustration for the marking of phrases: 1) [Ma-lipot-on] pred [ang tubig!] pred base ST-cold-INT PB water 'The water is very cold!' lit.: very cold = the water Here the sentence initial malipoton 'very cold' is the predicate, ang tubig 'the water' the predicate base, marked as such by ang. 2a) [Maestro] pred [ang teacher PB 'My father is a teacher.' lit.: teacher = my father

ama=ko] pred base father=1SG

2b) [Nag-bakal] pred [ang ama=ko] pred base [ki dakula-ng harong.] arg BEG.AG-buy PB father=1SG ARG big-LK house 'My father bought a big house.' lit.: (s.o. who) bought = my father, a big house In this example nag-bakal 'bought' can be identified as the predicate again, because of its sentence initial position. What follows is the predicate base (ang ama ko 'my father'), marked as such by ang. The second argument of the sentence, marked as such by ki, is ki dakulang harong 'a big house'. Within this phrase the modifier dakula 'big' is linked to its head harong 'house' by the linker -ng.

50 For more detailed discussion of equational sentence structure and its possible development from a former verbal sentence structure see Himmelmann (1991).


A short grammar sketch of Bikol

2c) [Ama=ko] pred [ang nag-bakal] pred base [ki dakula-ng harong.] arg father=1SG PB BEG.AG-buy ARG big-LK house 'My father bought a big house.' lit.: my father = (s.o. who) bought, a big house Here ama ko 'my father' is the predicate of the sentence. It stands in initial position. The phrase ang nagbakal ('the one who has bought') is the predicate base. In the examples (2) it becomes already clear that the lexemes in Bikol are not categorized for specific syntactic positions. Maestro 'teacher' appears in predicative function (2a), ama 'father' appears in referential function (2b) as well as in predicative function (2c), likewise for nagbakal 'bought'. Table 4 lists the paradigm of the phrase markers, which are all prepositional: predicate base



argument argument

locative argument




nin (ki)





kan (ki)








Table 4: Phrase markers

Modifiers are connected to their heads by the linker na (following a consonant) or -ng (following a vowel). There is no obligatory ordering of head and modifier. Their particular functions are only identifiable via semantic criteria. 3) ma-gayon na daraga ST-beauty LK girl 'beautiful girl'

= daraga-ng ma-gayon = girl-LK ST-beauty

4) lighter na zippo = zippo lighter LK zippo = zippo 'pocket lighter' (“zippo"-lighter)

(Lobel and Tria 2000: 47)

na52 lighter LK lighter [Paul's stories]

51 In some varieties of Bikol all undergoer arguments are marked by ki. 52 Here na is used instead of the postvocalic linker -ng. Here, this is obviously due to the extreme lento speech of the speaker in this example. The substitution also occurs sometimes for the purpose of emphasis.


A short grammar sketch of Bikol

Subordinated clauses are linked to their head phrases by the same linkers. Structurally, they are treated like phrase internal modifiers: 5) Asin na-hiling man niya na igwa duman ki nag-tubó' na mga berdi-ng kahoy. and ST-see also 3SG LK EXIST DEM.DIST.LOC ARG BEG.AV-grow LK PL green-LK tree [enot na tawo] 'And she saw also that there were green trees growing.' 6) dai niya ma-sabot-an kun ano talaga ito-ng na-hiling=mi. NEG 3SG ST-understand-REFL if what really DEM.DIST.AV-LK ST.BEG=1PL.EXCL [merr_asuwang] 'he could not understand what it really was that we saw' Bikol has two existential forms, may and igwa, corresponding to 'there is' and one form that is its negative equivalent mayo' corresponding to 'there is no(t)'. Most authors list the sentences containing these forms as a separate type, labeling it 'existential sentences/clauses' (for example Lobel and Tria 2000: 32, Mintz 2004: 29). May always precedes the constituent whose existence is expressed. I would call the combination of may and the following constituent a special type of predicate, and analyze the clause like all the other clauses, with equational structure (cf. above). If there is reference to a possessor, it appears in the form of the predicate base (mostly personal pronouns in ang-form). 7) [May prutas.] pred EXIST fruit 'There is fruit.'

8) [May kwarta] pred EXIST money 'I have money.'

[ako.] pred base 1SG.AF

9) [may gi~gibo-n] pred [siya] pred base [na ma-importante] rel clause EXIST IMPFV~do-UG 3SG.AF LK ST-important 'he will have important business' lit.: there is ( to be done = by him, that is very important It is obvious that igwa is used differently. But unfortunately the literature (Lobel and Tria 2000, Mintz 2004) does not point out any syntactic difference between the two existential expressions. And as this was not in the focus of my field work, I can only draw some conclusions from the few examples which I have in my corpus. Igwa might be preferably used with sentence modifiers and/or demonstrative pronouns. The existential expression is used predicatively, whereas the objects referred to are in argument position. 10) dai pa nin dagá', igwa saná ki lángit budá dágat NEG still ARG land EXIST only ARG sky and sea 'there was no land, there was only the sky and the sea'


[enot na tawo]

A short grammar sketch of Bikol

11) Daw ... na-lingwa-n=mi na igwa pa lang duman ki cemeterio. HEY ... ST.BEG-forget-UG=2PL LK EXIST still just DEM.DIST.LOC ARG cemetery [valentine's day] 'Hey ... we had forgotten that there was still the cemetery there.'

2.5 Morphology Preliminarily the Bikol lexicon can be divided into two groups: in content words (an open class) and in function words (a closed class). The function words can be further divided into subgroups. There are markers and linkers, all monosyllabic. Other subgroups are the pronouns (personal pronouns and demonstrative pronouns) and a somewhat larger group of conjunctions, prepositions, sentence modifying particles, etc. These particles often have functions which are fulfilled by adverbials in other languages (for example garo 'seemingly', gayod 'probably', daa 'is reported' etc.)53. The content words all have at least two syllables, except for loanwords and borrowed words which can also be monosyllabic (for example ref 'refrigerator', bol 'ball', tren 'train', etc.).54 Bikol makes extensive use of morphology. There are a large number of affixes, and these can be combined with each other in a multiplicative manner. Furthermore Bikol has a sophisticated reduplication system and to a small extent a mechanism of morphological accent change (cf. above). A detailed list of morphological procedures in Bikol can be found in Lobel and Tria (2000), and a short survey in Mintz (2001). 2.5.1 Voice- and TAM-affixes The core characteristic of Bikol morphosyntax and a typical feature of the Philippine languages is "voice marking"55. In the predicate, it indicates which semantic role is taken by the referent of the predicate base. Most of these voice-affixes are portmanteau morphemes 53 A detailed list and description of Bikol particles is found in Lobel and Tria (2000: 59-71). Cf. also Fincke (2002: 102-109). 54 In first instant these are of Spanish and English origin, both languages of colonial powers of the Philippines. (Spanish: 1565-1898; America: 1898-1946). To a much smaller extent words are also borrowed from Chinese. From the 10th century up until now, the Philippines have had intensive trade relationships with China, and today, more than 20 percent of the Philippine population is Chinese. 55 The terminology for this marking varies very much from one author to the next, and one finds for example „trigger“-system, „focus“-system or „case“-system. Himmelmann (1987), referring to Lehmann's suggestion (1984: 355) introduced the name Ausrichtungsmarkierung 'orientation' for Tagalog, which reflects the function of the marker most precisely. I am following here the term common in Austronesian linguistics, voice-marking.


A short grammar sketch of Bikol

and also express tense-aspect-mood56. In example (12a) the "actor-voice" prefix nag- in nagbakal 'bought' indicates, that the referent of the predicate base (si Mamay 'Mama') is the agent. 12a) Nag-bakal si Mamay ki BEG.AV-buy PB.PERS mom ARG.UG 'Mom bought a box.'

saro-ng kahon. one-LK box

In (12b) the "undergoer-voice" prefix pig- in the predicate indicates that the referent of the predicate base (ang sarong kahon 'a box') takes the role of the undergoer (the object). The agent is expressed in an (optional) argument, marked by ni. 12b)Pig-bakal ang saro-ng kahon ni Mamay. BEG.UG-buy PB one-LK box ARG.PERS mom 'Mom bought a box.' (lit.: (It is) bought, a box, by mom.) In (12c) the "conveyance-voice" prefix i- in the predicate indicates, that the referent of the predicate base (ang kwarta 'the money') takes the role of the instrumental (i- indicates that the undergoer changes its position). 12c) I-bakal ang kwarta ni Mamay ki saro-ng kahon. CV-buy PB money ARG.PERS mom ARG.UG one-LK box 'Mom bought a box with the money.' (lit.: (It is) bought with, the money, by mom, a box.) These affixes (prefixes, infixes and suffixes) are very transparent and regularly applied. The same is true for imperfective reduplication (i.e. root initial CV-reduplication), which can be combined with the voice- and TAM-affixes. Every root can take voice- and TAM-affixes as well as imperfective reduplication. These affixes are paradigmatically organized and therefore they superficially have an inflecting character. But while discussing in depth the issue of derivation and inflection, it turns out that there are strong arguments to assume no inflection but only derivation for the Philippine languages.57 The main argument is that the morphological operations are not syntactically conditioned, i.e. every morphological form can 56 In the following I will use the abbreviation TAM. There is an ongoing discussion if these affixes express time and/or aspect and/or mood, or something else. Himmelmann (1987) for example interprets it as a realis-/irrealis-distinction. Fincke (2002) on the other hand argues for distinguishing between begun and not begun actions or events. I adopt Fincke's interpretation here. 57 See also below. There are reasons to assume that all Bikol roots are "nominal" and that all other word forms are derivations from these nominal roots (cf. III.2.6).


A short grammar sketch of Bikol

be used in almost every syntactic function. All morphological changes mark in first instance the semantic changes. Therefore Himmelmann (among others) opts also for these regular affixations as being derivative (cf. Himmelmann 1987: 95, 129-139; 1991: 26-43, 2006: 497498; to appear: 1). The following table gives a survey of the meanings and functions of the basic voice- and TAM-affixes58: affix


function / use


mag- / -um-59

not begun

proposition, desire, imperative, infinitive



not begun

future, intention




past (and present)



begun and imperfective

(past and) present (progressive)


Table 5: Actor-voice affixes

Examples: 13) S{um}akay / Mag-sakay=ka sa {NBEG.AV}ride / NBEG.AV-ride=2sG.AF LOC 'Go with a 'padsya' (bicycle taxi)!'

padsya. padsya.

14) Nag-(a~)adal ako ki Bikol. BEG.AV-(IMPFV~)study 1SG.AF ARG.UG Bikol 'I learnt Bikol / I am learning Bikol.' 15) Sa aga ma:-laba ako. LOC morning FUT.AV-wash 1sG.AF 'Tomorrow I will do the laundry.'

58 I do not mention here other forms which are also paradigmatically organized like the abilitative, the causative, the repetitive, the social, the occasional, etc. (cf. for example Lobel and Tria 2000: 77). 59 mag- and -um- comprised two contrasting and lexeme-class establishing paradigms in "Old Bikol" (17th century). On the loss of the -um-paradigm see Lobel (2004). 60 CV- reduplication expresses imperfective aspect, cf. IV.4.1.2.


A short grammar sketch of Bikol



function / use


pig- / -in-


past (and present)


pig-CV- / C-in-V

begun and imperfective (past and) present (progressiv)


-on / -an




neutral (for undergoers past and present which move)


(i-)CV- (-on/-an)

neutral and imperfective future, intention


past and present

Table 6: Undergoer-voice affixes

If there is no imperfective reduplication, the action or the event is interpreted as either terminated (past) or general, as in example (16). In the predicate (pigkaon 'eat') the prefix pigindicates that the referent of the predicate base (ang dahon '(a) leave') is in the role of the undergoer. The agent niya (3sG) is expressed as an optional argument (in the form of a clitic attached to the predicate). 16) Pig-kaon=niya ang dahon. BEG.UG-eat=3sG PB leaf 'It eats leaves.' The infix -in- in the predicate of the following example (17) indicates the undergoer role of the referent of the predicate base (ang kwarta na papel 'paper money') as well as the aspect "begun action". There is no imperfective reduplication, so the action is interpreted as terminated. 17) B{in}ayad=ko ang kwarta na papel. {BEG.UG}pay=1sG PB money LK paper 'I paid with paper money.' The -on suffixation indicates the role of the undergoer of the referent of the predicate base (ika (2SG.AF), here as a component of the contracted form for taka 'I - you'). The imperfective reduplication and the absence of the -in-infix for begun actions result in an interpretation of a future/intended action. 18) Pi~pirit-on=taka mag-taram. IMPV~force-UG=I-you NBEG.AV-speak 'I will force you to speak.'


A short grammar sketch of Bikol

I- in the predicate indicates the undergoer which moves (therefore the term "conveyance voice"). This undergoer is expressed in the ang-form in the predicate base by ako (1SG.AF). The agent is expressed with an enclitic pronoun (mo, 2SG) attached to the predicate. 19) I-baba=mo ako tábi sa Bicol College. CV-down=2sG 1SG.AF POL LOC B.C. 'Please let me get off at the Bicol College!' The voice and TAM morphemes usually appear in the predicate (cf. above), but it is not obligatorily. They can also appear in the predicate base or in other arguments (cf. for example (2b), (5)), and vice versa, there are predicates without these affixes (cf. for example (2b), (21)). As already mentioned, every root can take these TAM-affixes, independent of its semantic features. This is also true of those lexemes which have a clearly nominal character from an Indo-European perspective, as for example machine-gun: 20)na-sabat-an sinda kan patrolya kan hapon, m{in}achine-gun sinda, ... ST-meet-UG 3PL.AF ARG.SPEC patrouille ARG.SPEC Japan {BEG.UG}machine-gun 3PL.AF [paul's stories] 'They met a Japanese patrol, they were shot by machine-guns, ... .' In spoken language the voice- and TAM-markers are often omitted, when the context disambiguates the role of the referent of the predicate base (for example in imperative predications): 21) Tapus=ka na po? finish=2sG.AF already POL 'Are you ready?' 22) Dai

ako NEG 1sG.AF 'I don't do anything!'

[pig-gi~] 61gibo! [BEG.UG-IMPFV~]do

23) Para tabi diyan stop POL DEM.MED.LOC 'Please stop where the dog is!'





ayam! dog

2.5.2 Pronouns Bikol has three demonstrative pronouns, for expressing proximal, medial and distal, each of which comes in three forms. Cf. also Fincke (2002: 87-89), Lobel and Tria (2000: 46) and Mintz (2004: 23-24). 61 Voice- and TAM-marking is optional.


A short grammar sketch of Bikol
















Table 7: Demonstrative pronouns

The personal pronouns in Bikol indicate person, number and voice. The first person plural has an inclusive/exclusive distinction: ang-form































Table 8: Personal pronouns

Even if these forms are all perceived as monomorphemic, it is obvious that all locative forms contain the locative marker sa. And except for the 2nd person singular, all ang-forms have obviously resulted from the prefixed undergoer forms. The undergoer forms are enclitics, immediately following the predicate. The only clitic ang-form exists for the second person singular (ka is a clitic allomorph of the free form ika). There is one special pronoun, taka, which combines 1SG and 2SG and has the meaning 'I to you' (cf. example (18)). This grammaticalized form originates in a combination of 1PL.INCL ta and 2SG.AF (i)ka. For the personal pronouns see also Fincke (2002: 81-84), Lobel and Tria (2000: 44-45) and Mintz (2004: 21-23). Some authors, for example McFarland (1974: 102), use the terms "nominative", "genitive" and "oblique" in place of "ang-form", "undergoer-form" and "locative-form". The former terms are quite conventional in Austronesian linguistics and as long as their users are aware that they do not refer to case in the common sense, these terms are absolutely useful. However, as a consequence of the decision to referring to "voice marking", I chose the latter terms here.


A short grammar sketch of Bikol

2.5.3 Number "Number" in morphology refers to a grammatical category which, very generally speaking, marks singularity or plurality. Number is usually classified as a clear nominal category, which only appears with verbs for agreement purposes. However, in various languages outside the Indo-European family, number marking also appears with verbs or other word classes, although this is far less examined than nominal number. For more details see Excursus VI. In Bikol, the only numeral category with grammatically regular expressions, is plural. It can appear with lexemes denoting entities, states and events/actions (i.e. semantic nouns, verbs and adjectives. For the problem of word classes in Bikol see III.2.6). There are two plural markers, -Vr- (an infix which is inserted between the root initial consonant and the following vowel, for example dakula 'big' - d-ar~akula 'big, pl.') and mga (/maNa/) which can be inserted at different places within the phrase (cf. for examples (24) and (25)). -Vrindicates plural actors or undergoers of an event or an action or plural referents of states or properties. -Vr- cannot appear with certain derivations. It has a suppletive alternate, i.e. CVreduplication, which applies to ma-derived statives, for example ma-ga~gayon 'beautiful, pl.' (*ma-g-ar~ayon). Mga can appear with every word form, but usually does not with TAMmarked words (cf. III.2.6). Number is not an obligatory category, though there is a form of agreement in cases where it is used. If there is a -Vr-infix in the predicate, the referent of the predicate-base has to be pluralized by mga. But plural actors of an argument or the predicate base do not have to agree obligatorily with an -Vr-infix in the predicate. 24)Nag-k{ar~}arigos kami. = Nag-karigos BEG.AV-{PL~}bath 1PL.EXCL.AF = BEG.AV-bath 'We took a bath.'

kami. 1PL.EXCL.AF

25)D{ar~}akulaang mga harong. = Dakula ang mga harong. {PL~}big PB PL house big PB PL house 'The houses are big.' but *D-ar~akula ang harong. For details about -Vr-plural see IV.4.1.4. The plurality of the action or the event itself is expressed by full reduplication, but this has a much more aspectual character than -Vr-plural and is treated in detail in 5.4.2.


A short grammar sketch of Bikol

26) Kan nag-hapot~hapot ako ... when BEG.AV-PL~ask 1SG.AF 'When I asked here and there ...'


2.5.4 Morphological marking of properties and states Lexemes which are marked as properties or states at the morphological level can operate in an attributive as well as in a predicative and even in a referential function. 62 The examples illustrate that there are rare semantic constraints with respect to the productivity of deriving properties and states. Himmelmann (2004) also attests this high productivity in Tagalog: "Stative (...) affixations are not restricted to a specific class of lexical bases but (...) may, in principle, occur on any lexical base, provided the resulting form makes semantic and pragmatic sense" (Himmelmann 2004: 113). There is no reason to separate the property and state derived forms from the other voice-and TAM-affixations. Their morphosyntactic behavior is exactly the same (cf. also Himmelmann 1987: 160-161 for Tagalog). In Bikol, properties and states are derived in the following four ways: 1. ma-prefix 27) ma-siram na prutas. ST-delicious LK fruit 'delicious fruit'

28)ma-bulod ang Bikol ST-mountain PB Bikol 'In the Bikol region there are a lot of mountains.'

2. ha-prefix (“property”-prefix for roots that express spacial dimension) 29) ha-baba' ang tukaw-an ST.SP-low PB sit-LOC 'The chair is low.' 3. accent change

30)ha-langkaw na lalaki ST.SP-tall LK man 'tall man'

31) búhay 'life' – buháy 'alive' ma-gayon ang búhay ST-beauty PB life 'Life is beautiful.'


buháy na manok alive LK chicken 'alive chicken'

62 Himmelmann (in detail in 2004, 2006) assumes two different ma-prefixes for Tagalog. He calls them stative and potentive. This distinction establishes a paradigm. Himmelmann's argumentation is very convincing and I assumed that in Bikol we would find something very similar. An initial check through my data however, leads to the conclusion that the differences between Tagalog and Bikol in this respect are considerable, so that I am not able to fully analyze it for Bikol at the moment.


A short grammar sketch of Bikol

32) gútom 'hunger' - gutóm 'hungry' ang

gútom na dakula hunger LK big 'ravenous hunger'



gutóm na ako hungry LK 1SG.AF 'I am hungry.'

4. Some roots have "inherent" meaning of properties or states and need not to be derived. 33) dipisil ang Bikol difficult PB Bikol 'Bikol is difficult.'

34) dakul na beer much PB Bikol 'a lot of beer'

To express inchoative aspect of properties or states, the word is marked for actor voice and the aspect 'begun': 35a) Ma-lipot ang tubig. ST-cold PB water 'The water is cold.'


35b) Nag-li~lipot. ang tubig BEG.AV-IMPFV~cold PB water 'The water is getting cold.'

Adding ma- to a root can yield different meanings, depending on the basic meaning of the lexeme. For lexemes expressing a physical or emotional feeling, there are usually two possibilities of affixation: It makes a difference, if the person to whom the feeling is associated experiences it (na-63), or if the person causes the feeling (ma-). The m-/nalternation is a TAM-contrast (begun/not-begun or irrealis/realis in Himmelmann (2004)), corresponding to mag- and nag- (cf. III.2.5.1). The prefix na- can be combined with imperfective reduplication. The ma-/na- word forms cannot be said to indicate "actor" voice, but they indicate the role of an "experiencer" (na-)64 or of a passive possessor of a property (semantic "absolutus"?) (ma-). Therefore they also fulfill the function of a voice affix. 36) Na-gu~gutom na ST.BEG-IMPFV~hunger already 'We are already hungry.'

kami. (*ma-gutom, *nag-gu~gutom) 1PL.EXCL.AF [bisita]

The following examples illustrate the difference between ma- and na-(CV)-: 37a) Ma-pagal ika. ST-tired 2SG.AF


37b) [Na-pa~]pagal ika. [ST.BEG-IMPV~]tired 2SG.AF

63 See also example (36) where the prefix na- has the additional meaning of "unvolitional/uncontrolled action", as opposed to nag- which expresses "volitional and controlled" action. 64 Therefore the na- construction is only possible with animates. For example na-li~lipot ako 'I am feeling cold' but *na-li~lipot ang tubig 'the water is cold' (but: ma-lipot ang tubig).


A short grammar sketch of Bikol

'You are tiresome.' 38a) Ma-mundo' ang pelikula. ST-sad PB film 'The film is sad.'

'You are tired.' vs.

38b) Na-mundo' ako sa pag-hali=mo. ST.BEG-sad 1SG.AF LOC DERIV-leave=2SG 'I was sad because you left.'

*Na-mu~mundo' ang pelikula.

*Ma-mundo' ako sa pag-hali=mo.

If ma- or na- is used together with the reflexive suffix -an, the affectedness of the experiencer is focused. For example Na-lipot ako. simply means 'I felt cold.' but Na-lipot-an ako. expresses that 'I was affected by the cold.' Some bases can never be used without the reflexive suffix when derived for state. For example intindí 'understand' (Na-intindi-han65 ika? 'Do you understand?') or aram 'know' (Na-aram-an ako. 'I knew.'). Cf. also example (6), (76), (117). 2.5.5 Further derivational affixes As already mentioned, Bikol has a large number of derivational affixes, besides the paradigms of the voice- and TAM-affixes. Only some of them are mentioned here: –

pag- (for example gibo 'do' - pag-gibo 'handcraft' or kaon 'eat' - pag-kaon 'the act of eating')

ka- -an (for example turog 'sleep-' - ka-turog-an 'bed' or tapus 'finish' - ka-tapus-an 'end')

(CV-) -an66 (balyo 'cross, traverse' - balyo-han 'bridge' or tukaw 'sit' - tukaw-an 'chair'; lakaw 'go' - la~lakaw-an 'way' or saki 'ride' - sa~saki-an 'vehicle').

In addition to these "nominal affixes"67, frequent derivations by the regular undergoer-voice and TAM-affixes -in- and (CV-)-on can be found which are lexicalized to various degrees.68 For example t-in-apay 'bread' < tapay 'dough' (this derivation is fully lexicalized and not transparent to the speakers), s-in-anglag 'fried rice' < sanglag 'fry' (this derivation is transparent, but sinanglag refers unambiguously to a specific rice dish), kaon 'eat' - kaon-on 'the food' (already eaten), gibo 'do' – gi~gibo-hon 'business' (that has to be done). From the point of view of semantics all these word forms clearly refer to entities, concrete as well as 65 66 67 68

/h/-epenthesis: cf. III.2.1.3. -an is a locative suffix. "Nominal" is to be understood only in a semantic sense. Comparable to the German or English past participle.


A short grammar sketch of Bikol

abstract. They are therefore called "nouns" by linguists (cf. for example Lobel and Tria (2000: 33-34)) and by the native speakers69. It is problematic, however, to speak of nouns and verbs in Bikol, since it is highly questionable whether the language has any parts-of-speech distinction at all. In order to understand the following argumentation on reduplication it will be necessary to understand my notion of word classes in Bikol and the terminology I am using. Therefore I will highlight the question of parts of speech in Bikol in the following section (cf. also Mattes 2005 and 2006b). 2.6 Parts-of-Speech 2.6.1 Functional elasticity What is striking as soon as one begins to concern oneself with a language like Bikol is its high "functional elasticity" (Himmelmann, to appear) of its content lexemes. I.e. regardless of its basic semantic meaning, every lexical root can undergo almost any kind of derivation and appear in any syntactic function. To illustrate: in (20) machine-gun is combined with a TAMand voice affix and used predicatively. In (39a) lana 'oil' is combined with a TAM- and undergoer-voice affix and is used referentially. In (39b) it is also combined with a TAM-AVvoice-affix but used predicatively. In (39c) the same lexeme has the function of reference, without any affixation. A: (39a) Dakul-ón pa an l{in}ána=mo. (39b) Ma:-lána man sana ini-ng ano. much-INT still PB {BEG.UG}oil=2SG FUT.AV-oil also just DEM.PROX.AF-LK SUBST 'You have put too much oil.' 'The stuff produces it's own oil, anyway.' B: (39c) Ali-on=ko an iba-ng lána? away-UG=1SG PB some-LK oil 'Should I remove some of the oil?'

[bikol express]

The crucial point concerning the definition of parts of speech, at least in the classical sense, is the correspondence of the behavior of lexemes at the various levels of language, i.e. the semantic, the syntactic and the morphological (cf. Hengeveld 1992, Sasse 1993: 648-652). It is exactly this correspondence between the grammatical levels which is very weak in 69 Such statements by the native speakers I talked to have to be taken with a grain of salt: They all speak English or German and they all studied the grammar of these second languages and of Tagalog on the basis of the traditional Latin grammar.


A short grammar sketch of Bikol

languages of the Philippine type. 2.6.2 Derivations Every root of the language can appear with almost any derivation. This means that there are no restrictions and no precategorizations with this respect. Cf. example (39) above, and the following lists of derivations: (40) haróng 'house' mag-harong BEG.AV-house 'build a house'

ka-harong-an ma- harong pag- harong etc. NMLZ-house-NMLZ ST-house NMLZ-house 'living room' 'full of houses' 'the building of a house'

(41) babá' 'low' mag-babá' ha-babá' BEG.AV-low ST-low 'lower, get off, etc.' 'low'

ka-ba'b-án pag-babá' NMLZ-low-NMLZ NMLZ-run 'valley' 'lowness'


ma-dalagan ST-run 'fast',


(42) dalágan 'speed' mag-dalágan BEG.AV-run 'run'

para-dalágan REP-run 'runner'

pag-dalágan NMLZ-run 'the running'

Furthermore, no (derived) word form is specified for certain syntactic positions, at least for predicate and argument. (43) Dakul-ón na pag-kaón an h{in}andá' ni much-INT LK NMLZ-eat PB {BEG.UG}prepare ARG.PERS 'Very much food was prepared by the mayor.'

Máyor. mayor [merr_asuwang]

In this sentence, a pag-derivation, which has clear nominal semantics, is found in predicate position, while a TAM-derived word, with clear verbal semantics, takes a referential function. (44) May sirá' na nag-la~láyog. BEG.AV-IMPFV-fly EXIST fish LK 'There was flying fish.'

[paul's stories]

Here a TAM-derivation in used in modifier position (attributive function). In the following sentence, the TAM-affixation can be found in the predicate (in predicative 48

A short grammar sketch of Bikol

function) and in the argument (in referential function). (45) Ngunian nag-hanap siya ki ma-tu~tugdon-an, now BEG.AV-look.for 3SG.AF ARG.UG ST-PL~perch-LOC 'Now she looked for somewhere to perch, ...'

... [enot na tawo]

Nevertheless, with regard to the syntactic criteria of attributive usage, we can establish two categories of word forms: the first class contains word forms which can appear in every syntactic position, i.e. predicate, argument and modifier. These are the TAM-affixed words. The second class contains word forms which cannot be used as modifiers. Preliminarily I call them “anti-adjectives”. These are derivations like pag-abot 'arrival', pag-tubod 'faith', kaharong-an 'living room', ka-putí'-an 'whiteness'. Because of the clear nominal semantics of the second group I consider it justified to talk of a nominal and verbal word form classes. 2.6.3 Roots A classification of word forms into two categories is not sufficient to justify the existence of different parts-of-speech in a language. If we take into consideration Hengeveld's criteria proposed in his Functional Grammar Approach, the question is, do the two word form classes correspond somehow to the lexical root level. "A verbal predicate is a predicate which, without further measures being taken, has a predicative use only. A nominal predicate is a predicate which, without further measures being taken, can be used as the head of a term." etc. (Hengeveld 1992: 58). As it is true for all derived word forms, all roots can also be used predicatively as well as referentially "without further measures". Even if the most frequent use of unaffixed roots is the referential one, all roots can also appear unaffixed in predicative function, at least in colloquial speech. The reason of the high flexibility of roots and derived word forms with respect to the predicative and the referential use seems to be the dominant equational character of the Philippine syntax (see above) which renders the predicate and the predicate base to be easily interchangeable. Again, what is restricted, often at the root level, is the attributive function. Many roots can be used attributively without further measures (denoting properties, states): (46) dakula 'big' Dakula an harong. (pred.) 'The house is big.', an dakula-ng harong (attr.) 'the big house.', Ba-bakal-on=ko an dakula. (ref.) 'I buy the big one.' 49

A short grammar sketch of Bikol

(47) turog 'sleep' Turog na ika? (pred.) 'Are you already asleep?', an turog na aki (attr.) 'the sleeping child', An turog kaipohan mo! (ref.) 'What you need is to sleep!' (48) tubod 'believe' Tubod ako sa Diyos. (pred.) 'I believe in God.', an tubod na lalaki (attr.) 'the religious man', An tubod mapakusog mo. (ref.) 'The faith will make you strong.' Some other roots cannot be used attributively without derivation, for example ayam 'dog', lalaki 'man', harong 'house', gayon 'beauty' (objects, animates, abstracts, ...) and e.g. ha'bas 'steal', dalágan 'run/speed', etc. (dynamic actions). But as soon as they are affixed with TAMmarkers, they can appear in any syntactic position: (49) ma-bulod na Bikol ST-hill LK Bikol 'hilly Bikol region'

but * bulod na Bikol hill LK Bikol

(50) ma-gayon na babae ST-beauty LK woman 'beautiful woman'

but * gayon na babae beauty LK woman

(51) ha'bas 'steal' (51a) an ma-ha'bas na lalaki PB ST-steal LK man 'the stealing man' (51b) an h{in}a'bas-an na relo PB {BEG.UG}steal-UG LK watch 'the stolen watch' (51c) *


ha'bas na relo / lalaki but: an ha'bas kan relo steal LK watch PB steal GEN watch intended: 'the stolen watch' / 'the stealing man' 'the theft of the watch' PB


A short grammar sketch of Bikol

(52) dalágan 'speed / run' (52a) an nag-da~dalagan na auto PB BEG.AV-IMPFV~run LK car 'the fast (running) car' (52b) an ma-dalagan na auto PB st-run LK car 'the fast car' (general) (52c) * an

dalagan na auto but: an dalagan kan auto PB run LK car PB run GEN car intended: 'the fast car' 'the speed of the car' Again, also at the root level, we find one group of roots which can do "anything" (syntactically speaking), and one group of roots which cannot be used attributively without further measures (“anti-adjectives”). However, one cannot draw a parallel between these two groups and cannot be parallelized with the "nominal" and "verbal" derivation classes. Unlike the “nominal” word form class, the lexical (the "anti-adjectives") class is not a semantically homogeneous group, i.e. it contains objects as well as actions. 2.6.4 Conclusion At the lexical level, even if two word classes can be established with respect to the syntactic behavior, it is not possible to clearly classify them as nouns or verbs or even adjectives. There can be identified one "all-round class" and one "anti-adjective class". At the word form level, one big class of "verbal derivations" can be classified (the term is justified via semantic and syntactic criteria) against a smaller class of "nominal derivations". But the levels do not correspond to each other, which makes a traditional parts-of-speech categorization impossible. Even if it does not hold true that in Bikol everything is possible with every root and every word form, and even if there are some roots and derivations that have a nominal or a verbal character, I state that it is impossible to establish a clear noun-verb distinction in Bikol (and probably in many other Philippine languages). The syntactic restrictions which were found concern only the attributive function, and they are exclusively semantically/logically conditioned. In general, semantic categorizations play a dominant role in the grammar of Bikol, as can also be seen with the highly polysemous full reduplication. The exact meaning 51

A short grammar sketch of Bikol

of this derivational means depends mainly on the semantic category of the base (cf. IV. It has been stated of Tagalog (Himmelmann (to appear, 2001: 17ff ), Starosta, Pawley and Reid 1982, Naylor 2006), that it basically consists of nouns on the lexical level, and that every other word form is derived to a certain degree. Of course one can doubt, if it makes sense to identify only one basic category and call it "noun" (I will not go into detail with this respect), but indeed, the roots in Bikol have nominal properties, and in this sense the most frequent and unmarked function of underived roots is that of reference. So it might be reasonable to associate a "nominal character" to Bikol and possibly to languages of the Philippine type in general.


The reduplication system of Bikol

CHAPTER IV: THE REDUPLICATION SYSTEM OF BIKOL 1 Reduplication in the Austronesian language family Reduplication is a morphological phenomenon which appears in many languages of the world all over the globe. The Austronesian languages are known to use reduplication abundantly, but to varying degrees (cf. Klamer 2002: 938). To start in the west, most Indonesian languages for instance make extensive use of full reduplication but only rarely have partial reduplication.70 In the east, i.e. the Oceanic area, on the other hand, both full and partial reduplication are very widespread in most languages71 (for example Mokilese, Kiribati (gil), etc.), but there are also several languages which completely lost reduplication or never had it as a productive morphological procedure (for example some New Caledonian languages72). The Philippine languages mostly have several reduplication types, and Bikol undoubtedly belongs to those which have a rather sophisticated reduplication system, i.e. it has many different formal types with many different functions, which can further combined with each other in multiple ways. The reduplication systems in the Philippine languages have some common features, but there are also striking differences between the individual languages. For the purpose of comparisons in this study I will only take into consideration the Central Philippine neighbor languages Tagalog, Cebuano and Hiligaynon, and sometimes refer to Northern Luzon languages (Agta, Ilokano), which are more distantly related, but which provide some interesting comparative data. Compared to most other related languages, Bikol has not only exact reduplication, but also two types of reduplication with fixed segments (-Vr- and Curu-). It shares these types with Cebuano and Hiligaynon (-Vl-, Culu-) for example, but not with Tagalog. In Bikol there are several productive or highly regular reduplications which are used in all text types and all registers. Additionally there are a numerous lexical reduplications which are becoming more and more rare. Firstly because they denote things and concepts which are out of use in modern life (for example a lot of mythological terms or names for old tools etc.), and secondly (and consequently) because they are considered as rural and out-dated. For example there is a fascinating array of 70 For the very different usages of reduplication in the various Indonesian languages see Gonda 1950. 71 „Reduplication is almost used universally in Oceanic verbal morphology, as well as in noun derivation“ (Lynch et al. 2002: 44). 72 Isabelle Bril, pers. comm.


Reduplication in the Austronesian language family

different and very specific reduplicated words for various sounds, but they are also out of use for the younger speakers especially in the urban environments. I cannot really judge the role of the superstrate languages Spanish and English in this respect, but they might be responsible to a certain degree for the reduction of reduplication in the Philippine languages. In Northern Luzon languages, for example in Ilokano the comparative is formed by reduplicating C 1V1C2of the root (dakes 'bad' – dak~dakes 'worse', na-kuttong 'thin' – na-kut~kuttong 'thinner'; cf. Rubino 2000: lvi-lvii). In Bikol and in its neighbor languages the comparative is formed by the prefix mas-, a borrowing from Spanish (Bik.: dakul 'much' – mas-dakul 'more', ma-gayon 'beautiful' – mas-ma-gayon 'more beautiful'). Probably the comparative was originally formed by reduplication in these languages, too, but was substituted by prefixation based on the Spanish model. This isolated example is a case of direct supersession of reduplication by an alternative procedure, introduced by the superstrate. But the main reason for a reduction in the use of reduplication is possibly an indirect one, i.e. via the speakers' conceptions about their own language. Compared to English, many Philippine speakers consider the words in their language as much too long.73 Especially fully reduplicated words are therefore perceived as "too long" and "too complicated to pronounce" and consequently as "old fashioned". But even if lexical reduplications are often either avoided or at least shortened, the productive use of reduplication is still very frequent.

2 A survey of the reduplication types in Bikol In Bikol there can be drawn the distinction between lexical reduplications (without corresponding simplex form) on the one hand, and productive reduplications (which can be related to a simplex form) on the other hand. With respect to productivity, the many types of Bikol reduplication range from highly productive and regular at one end on the productivity scale to completely fossilized at the other end, and types which are located in between these two extremes. The distinction between lexical and productive reduplication is the main basis for the organization of the following chapters. Based on the criteria established in 2.3, I will begin the analysis of the reduplication system of Bikol with a survey of the occurring types. 73 I heard often comments like "We prefer to use the English names, because they are much shorter". For example shop for tindahan or bread for tinapay or hi for kumusta. That is an interesting issue, because Bikol originally does not allow monosyllabic content words. Generally, the preference of Philippine people for using abbreviations, short forms, and acronyms is noticeable and would be a valuable topic of research. A systematic tendency for abbreviation in word formation is observed in Tibetan by Vollmann (2006).


A survey of the reduplication types in Bikol

As argued in 2.3, I start from the formal point of view and from there proceed to the functional. Bikol has full reduplication and several types of partial reduplication. In the first category, namely lexical reduplication, four formal types can be defined. First, bisyllabic roots in Bikol which consist of two identical syllables, i.e. C1V1(C2)C1V1(C2). The development of this type has to be considered with respect to the Bikol syllable structure of lexical bases which all consist of at least of two syllables. For some of these reduplicated bases there can be found a corresponding monosyllabic root, reconstructed for PAN, which was obviously reduplicated in order to achieve the required syllable structure. But furthermore, these bases composed of two identical syllables seem to have no arbitrary semantics. Almost all of these lexemes refer to concepts which are cross-linguistically typical for reduplication (in detail cf. IV.3.2). The second type is partial reduplication (with different structures, cf. IV.3.3), i.e. there is a partial reduplicative structure within the basic lexeme. In most cases these bases consist of three syllables. The third type of reduplication without simplex form has the structure of two identical bisyllabic units. Some of these reduplicated forms can alternatively be used in their "unreduplicated version", without any change in meaning, others become ungrammatical without their reduplicative structure. Finally there are lexical echo-words:

their initial or medial consonant alternates in the two copies. The

alternation of the consonants follows a certain rule and the semantics of these echo-words is restricted to few domains, such as like movements, sounds, and shapes (forms / surfaces). All these different types of lexical reduplication have meanings which are the typical meanings for reduplicative constructions (productive and unproductive) in many languages, and the meanings of all four types without simplex form overlap in a great measure (cf. IV.3).


A survey of the reduplication types in Bikol

form bisyllabic roots



hypocoristics, taboo-words; sapsap (kind of fish) sounds


movements; kadkad 'dig a hole'

animals, plants and co. partial

dapdap (kind of tree)

sounds; plurality; animals, bulalakaw 'shooting star' mythology and co.




pagatpat 'rice bird'

movements; kimot-kimot 'mumbling'

plurality echo-word


kiling-kiling 'shake the head' and

movements; karog-kadog 'shaking sound'

surface properties

harap-hasap 'rough'

Table 9: The lexical reduplication types of Bikol

Besides these numerous lexical reduplications, which are transparent in many cases, Bikol has five productive reduplication types. First of all, full reduplication, with its variant Curu-, in complementary distribution. This type is multiply polysemous (plurality, imitation, attenuation, diminution, etc.) and requires probably the most elaborated and detailed analysis (cf. IV.4.2). There is a second type of full reduplication for intensification, which has a very restricted domain however. It can only be applied to bisyllabic bases with the structure C1V1C2V2(C3). Finally there are three types of partial reduplication: The completely regular CV- reduplication for imperfective aspect, the infixed -Vr- reduplication for plural (with alternative CV-reduplication for ma-derivations), and CV-reduplication which only appears with numerals and has a limitative function. Interestingly, all these reduplication types can be multiply combined with each other and with other affixations. From the functional point of view reduplication in Bikol can be summarized as being used to express plurality (in very many different nuances as will become clear in the following analysis), diminution (with extensions like pragmatic attenuation), intensity, and imperfective aspect. This means that reduplication in Bikol has very concrete and highly iconic functions as well as more abstract and purely grammatical functions. An interesting function of full reduplication in Bikol (and perhaps in other languages) is the pragmatic usage, i.e. as a means of expressing politeness in spoken language. This


A survey of the reduplication types in Bikol

function is achieved through the diminutive meaning of reduplication. The use of diminution for politeness strategies is common in many languages (cf. Dressler and Merlini Barbaresi 1994). For more details cf. IV.4.2.2. Form full (I)/Curu-



plurality (distributive, iterative, balik~balik 'come and go' repetitive, etc.) and diminution huru~harong 'small house' (attenuative, imitative, etc.)

full (II)


mahal~mahal 'very expensive'


imperfective aspect

nag-tu~turog siya 'he is sleeping'


plural actors

nag-k-ar~aon sinda 'they ate' ma-ga~gayon sinda 'they are beautiful'

Table 10: The morphological reduplication types of Bikol

As already mentioned, Bikol and Hiligaynon have very similar reduplication system, with respect to both forms and functions. Compared to Tagalog, there are some differences, which might not be visible at first glance. Tagalog full reduplication for example denotes distributive and imitative, as in Bikol, but intensive, plural and diminution in Tagalog is expressed by bisyllabic reduplication which does not exist in Bikol. Unlike Bikol, Tagalog does not have reduplication with fixed segments. Furthermore, plurality in Tagalog can also be expressed by CV- reduplication, which is not possible in Bikol. Northern Luzon languages like Ilokano or Agta on the other hand, have reduplications which cannot be found in languages spoken in Southern Luzon, as for example the plural of human nouns (cf. Reid to appear). Typically, these Northern languages have CVC- reduplication, whereas the Southern languages use CV- reduplication, for example for aspect. The diachronic development is briefly described in 4.1.2.


Lexical Reduplication

3 Lexical Reduplication 3.1 Introduction Lexical reduplication is defined as one lexeme which consists of two or more segmentally identical parts (cf. II.1). It is not related to a simplex form. I consciously do not use the term “lexicalized reduplication”, because a part of these forms I am referring to might be a result of lexicalization of a former productive reduplication rule, but another part has probably come into the language without such a rule (cf. IV.3.2.1). The research on reduplication mainly focuses on productive reduplication, whereas lexical reduplication does not receive much attention. However, many languages have lexical reduplication to a much greater extent than generally noticed. They are especially interesting, because of the cross-linguistic regularities that can be observed in this respect (cf. also Mattes and Vollmann 2006). The reduplications without a corresponding simplex form are not synchronically produced by a morphological rule, but are stored as such in the lexicon. Consequently many of them are listed in the dictionaries and vocabulary lists which are available for Bikol. And in fact, this is the main database for my analysis of lexical reduplications. I collected all reduplicated entries of every type in the dictionary of Mintz and Britanico from 1985 74 which also incorporates the items of de Lisboa's Vocabulario from 1754. Unfortunately, Mintz and Britanico do not provide any information on the up-to-dateness of the lexemes which they incorporated from de Lisboa. I tested all the reduplications (from 1985 and 1754) with native speakers, so I got an idea about the actual use, at least for the speaker community I am working with. Mintz and Britanico's dictionary (1985) contains approximately 20.700 entries. A surprisingly high percentage, 5 percent (approx. 1.040), of theses entries are reduplications. 3.2 Bisyllabic reduplicated roots 3.2.1 Origin and development The first type of reduplication I am describing is very old and completely fossilized. It appears synchronically as a monomorphemic bisyllabic root consisting of two identical 74 The newly edited dictionary of Mintz (2004) was not available to me at the time I started with my research. But spot checks showed, that the changes from the 1985 edition to the 2004 edition are minor and do not require a reanalysis of my results.


Lexical Reduplication

syllables, i.e. C1V1(C2)C1V1(C2). As mentioned earlier (cf. II.2.5), all native content words of Bikol consist of two or more syllables. The avoidance of monosyllabics (or a tendency towards disyllabics respectively, cf. Blust 2001: 15-16) is a characteristic feature of Austronesian languages. For PAN many of monosyllabic roots were reconstructed, which have been affixed or reduplicated in many daughter languages in order to create bisyllabic lexemes. Lobel and Tria (2000: 13-15) list several examples for PAN roots and their derived words in Bikol, like PAN *luk75 > Bik. luklok76 'hidden away', luklob 'to crouch', lukon 'to conceal/hide' (Lobel and Tria 2000: 14), and further lukon-lukon 'the curvature of the hollow of the knee'. Dempwolff's hypothesis (1934) is, that these words of the form C1V1(C2).C1V1(C2), like luklok, emerged by reduplication of monosyllabics (“iterierte Wortwurzeln”, Dempwolff 1934: 109).77 Blust (1976) discusses a different possible development, in opposite direction. He assumes that these bisyllabic reduplicated roots originate in full reduplications of the form C1V1C2V1(C3)~C1V1C2V1(C3), with C2 being a laryngeal, which by reduction achieved the actual form. E.g. *buqu~buqu > Tagalog and Bikol bubo 'fish trap' (cf. Blust 1976). But besides these lexemes, many of the bisyllabic reduplicated word forms can be categorized as onomatopoetica, and for those the origin is not as clear. Gonda (1940), points out that "Such words, like onomatopoetica, may be adopted in a language at any time in its history, and they often resist the normal sound changes. Some of them may thus belong to the proto-language, others are innovations, and the difficulty is to distinguish between the two.” (Gonda 1940, cited Dahl 1976: 105)78 The question of which of the C1V1(C2)C1V1(C2)-roots can be traced back to PAN and which of them have entered the language as highly iconic words at a later stage cannot be answered within the frame of my study, because of the lack of appropriate data for an in-depth diachronic investigation. However, I will attempt a semantic categorization of the bisyllabic reduplicated lexemes as I found them in the dictionary from 1985. At least, what I could test 75 The meaning of *luk is not mentioned by the authors. But the meanings of the derivations suggest that it refers to the general concept of „hiding“. 76 As already explained in 4.2.3, /u/ and /o/ are often distinguished only in orthography (usually the final syllable is written with , other syllables with . But phonemically they both have the value /u/. I.e. is an exact reduplication and is pronounced as [lukluk]. 77 For roots with a C1V1C1V1C2- structure, 78 On bisyllabic reduplicative word structure see also Wundt (1900: 581).


Lexical Reduplication

was the actual use of these words. A great amount (about one third) of all the bisyllabic reduplicated roots found in the dictionary (including the dictionary from 1754), were unknown to the speakers I interviewed, or they were only passively known, but out of use nowadays, at least in urban environments. For example the term gamgam for a bird, found in Mintz (1985), was known by the speakers, but only passively. They would call the bird maya, whereas they consider gamgam as belonging to a purely poetic style. Many words of course are simply out of use today because they refer to actions or objects or to a type of knowledge which are not relevant in modern life. This applies especially to the terms referring to old mythological rituals or spirits, rural techniques or animals which are unknown to urban people, for example pupo' 'un rito, que tenian antiguamente que decian, que una phantasma ponia la mano à los ninos sobre la cabeza, y por esto no ciecian' (de Lisboa 1754: 533), subsob 'arroz traspuerto en mucha agua' (de Lisboa 1754: 642) or ngakngak79 'un genero de ave de rapina grande' (de Lisboa 1754: 274). For some words, in the time span from de Lisboa's recording till today, different semantic changes, i.e. a broadening of meaning can be observed. For example ladlad which is translated by de Lisboa (1754: 380) as 'cosa descogida [sic], ò tendida', means today 'make it know, outcome', i.e. the meaning has changed from concrete to abstract. Another example is kalkal80, which is translated in de Lisboa (1754: 157) as 'quitar la postilla de las llagas, ò heridas'. It still has the meaning of 'scratch the scab off a sore or wound', but it also can be used figuratively as 'digging or investigating'. Or gu'go' which in Mintz and Britanico (1985: 296) is translated as 'shake/raffle off', but which has since obtained the metaphoric meaning 'manage everything by oneself / without help'. sungsong81 is translated in de Lisboa (1754: 651) as 'agua, ò viento contrario al que navega' and in Mintz and Britanico as 'go against the current or into the wind'. My consultants translated the word as 'wade in the water' and figuratively as 'meddle with s.o. else's affairs'. 3.2.2 Child language expressions Some of the bisyllabic reduplicated lexemes originate in child (directed) speech, most of them are hypocoristics, which appear with reduplicative structures in many languages, even in 79 de Lisboa (1754: 274): gnacgnac. 80 de Lisboa (1754: 157): calcal. 81 de Lisboa (1754: 651): songsong.


Lexical Reduplication

those which do not have grammatical reduplication (cf. for example Comrie (1989) for Brazilian Portuguese or Rainer (1998) for French). Examples for hypocoristics in Bikol are tuto for an intimate friend, tita for a respected female person, which is older than the speaker, nonoy for the youngest son of a family and yaya for a baby sitter. Some of the hypocoristics are clearly borrowed from Spanish, like tata for father or another respected male person, or nene for the youngest child of a family. So it cannot be judged whether the reduplicated hypocoristic forms entered the language via Spanish or if they emerged already in the preSpanish period. At least, hypocoristics do not appear in de Lisboa's dictionary (1754). Dressler et al. (2005) argue for the origin of hypocoristic reduplications in general in the frequent occurrence of reduplicative structures in child language and even more in child directed speech. The authors point out that adult hypocoristic meaning is probably a reanalysis of child language expressions. I.e. the frequent occurrence of reduplication and diminutives in child language and baby talk has led to a reinterpretation of these word forms as possessing primary emotional meanings. This might universally lead to the association of reduplication "with hypocoristic, diminutive, attenuative meanings" (Dressler et al. 2005: 467; for more details on hypocoristics and diminutives and their connection to emotional pragmatic use see Dressler and Barbaresi (1994)). Not surprisingly, there are also several reduplications in Bikol child directed speech referring predominantly to intimate organs, as is the case in many other languages (for example German Popo 'bottom', or Pipi 'urine'), like dodo 'breast', mimi 'female breast', pipi 'female genitals', titi 'male genitals', pu'po' 'wash the genitals'. 3.2.3 Sounds and Movements A lot of the bisyllabic reduplicated lexemes in Bikol are onomatopoetica, i.e. they are referring to sounds. For example kikik 'squeak', ngongo82 'speak through the nose', tangtang 'the clanging of a bell', tubtob 'thumping sound', tuktok 'knock', tugtog 'play a music instrument', sa'sa' 'crushing metal', etc. Like hypocoristics, onomatopoetica are very often reduplications cross-linguistically, even in languages where reduplication does not play a major role (cf. Key 1965: 99-100). Arbitrarily chosen examples from other languages are 82 ngongo contains the submorphemic unit ng- which appears with lexemes denoting body parts or actions connected with mouth, nose and throat, like for example ngaros 'gums', ngata 'chew', ngilo 'kind of tooth pain', ngipon 'tooth', ngirit 'smile', nguso' 'mouth' etc.


Lexical Reduplication

Tibetan (bod), zer zer 'rumor', rlung phing phing 'breath heavily', hur hur 'sound of water, wind, fire' (Mattes and Vollmann 2006), from Kwaza83 dyra/-dyra/ 'rattle (small objects)', tsitsi/txi-txi 'burn' (van der Voort 2003: 79) and Portuguese cri~cri 'chirp', fru~fru 'rustle' (Kröll 1991: 26, 30). Like all the reduplicated lexemes without a corresponding simplex form, the onomatopoetic expressions do not belong to the productive morphological procedures. Dressler et al. (2005: 456) call these forms which are syntagmatically, but not paradigmatically, iconic "extra-grammatical" reduplications. Besides the clear onomatopoetica which denote sounds, there are many lexemes which in the first instance denote specific movements. On the one hand these movements semantically all include inherent plurality, like for example kadkad 'dig a hole with the hand or a stick', where the actor has to repeat one movement several times in order to accomplish the task. This iconicity of a reduplicative structure for plural actions or events (cf. Excursus I) would be one possible explanation for this form - meaning association. But furthermore in most cases there are also sounds associated with the respective movements (for example a scratching sound with kadkad, or a scraping sound with sidsid 'shuffle the feet'). Therefore it would also be reasonable to subsume all these lexemes under the category "onomatopoetica". Thus, for the bisyllabic reduplicated lexemes which denote movements like “swaying, waving, scraping, shaking, mincing, squeezing, pushing, etc.”, there are actually two iconic motivations which explain the respective form: inherent plurality and/or intensity of the movement and of the sound. Some further examples are gitgit 'squeeze through', gutgot 'cut by drawing a knife back and forth like a saw', ki'ki' 'jump with one leg, hop', kudkod 'grate coconut', ngatngat 'mince with the teeth', paypay 'fan, wave', paspas 'threshing rice', etc. 3.2.4 Animals, plants and Co. Besides the lexemes which are clearly iconically motivated, there is a considerable number of bisyllabic reduplicated lexemes which cannot easily be explained with iconicity, at least not in any obvious way, but which are commonly expressed by reduplication (of different types) in other languages, especially in other Austronesian languages. These lexemes can be assigned to the semantic fields "animals" (predominantly insects, birds and fish), "plants", 83 Kwaza is an unclassified Amazonian language.


Lexical Reduplication

"diseases", "body parts" and "tools and techniques". Additionally there is a field denoting variants of "food, cooking, eating and drinking". Here are some examples for the mentioned categories: - "animals": nuknok 'gnat(s)', wikwik 'bird of prey' (and: sound produced by this bird 84), sapsap (a small, silver-colored, disk-shaped, saltwater fish) - "diseases": gisgis 'dog's skin disease', hadhad 'skin disease, abnormal pigmentation' - "plants": gugo', dapdap (trees), wagwag (a variety of rice) - "tools and techniques": kagkag 'rake', paspas 'a method of threshing rice', pungpong 'close off the mouth of a canal', putpot 'cutting of firewood', taktak 'spread out nets for fishing or hunting' - "body part": ku'ko' 'chin', kudkod 'finger- and toenails, hoofs' - "cooking, food, eating, drinking": gawgaw 'starch (made of arrowroot or cassava)', luglog 'noodle dish', puspas 'rice porridge with chicken', sa'sa' (prepare fish by rubbing salt into the flesh in order to soften it), sabsab / sibsib 'eating grass by the animals', sumsom (crackers or nuts which are served with drinks), ung'ong 'drink from its container, instead of a glass or cup' Almost all these semantic fields contain also roots with a partial reduplicative structure besides the C1V1(C2)C1V1(C2)-roots, as will be described below. In other Austronesian languages, the same word families contain reduplicated lexemes as well. Cebuano, Fiji, Mokilese and Woleiean (woe) have names for animals, plants and diseases with reduplicated lexemes of different forms. For example there are abundant names for fish species in Mokilese, like pwei-pwei, moan-moan, mwel-mwel, etc., for diseases, like kho-ko, kar-kar, and for plants, like ping-ping, lam-lam or kam-kam (cf. Harrison and Albert 1977). Reduplicated names for animals, plants and movements are also documented for languages of other families, even for Indo-European. For example Portuguese (chas-chas, tim-tim (bird names), chu-chu, mio-mio (plant names), cf. Kröll 1991), Arabic (arb) (qa'qa' 'raven', 'ut'ut 'goatling', tahtah 'young horse', hašhaš 'poppy', dardār 'elm tree', cf. Procházka 1995: 58). Based on these comparative data I hypothesize that there is a universal tendency to associate animals, plants, diseases and food with reduplicative structures, but at the moment the cross84 Obviously the bird name is derived from the onomatopoetic expression for the sound produced by the bird.


Lexical Reduplication

linguistic database available to me is still too small to verify this assumption.85 In any case, I opt for the following explanation of the widespread association of reduplicative structures with these semantic fields: Animals and plants etc. can very often be associated with plurality and/or sound and/or movement. The most obvious case is that of birds, which are associated with the sounds they produce. Fish, especially small fish, usually appear in flocks, plants characteristically have a big amount of leaves, blossoms or stingers, which could lead to the association with plurality. Diseases, especially skin diseases (which are the majority in my corpus), are often characterized by distributed spots or blains, or diffuse and/or continuous pains. Here again an association with plurality can explain the reduplicated word form.86 However, one may not overlook the fact that all these word families have a lot of unreduplicated word forms. So, the interpretation is only valid in one direction: Most of the lexemes with reduplicative structure belong to the mentioned semantic fields in several Austronesian languages. But this does not mean conversely that all or most lexemes belonging to the mentioned word families have reduplicative structures. 3.3 Lexical partial reduplication 3.3.1 Forms and origin Besides the roots which consist of two identical syllables, a certain amount of roots with three or more syllables in Bikol contain partial reduplicative structures. These can have different forms. In most cases we find adjacent syllable reduplication. Either the initial syllable (i.e. C1V1~C1V1C2V2(C3)



ba~baga' as







second the


syllable syllable

(i.e. (i.e.

C1V1C2V2(C3)~C2V2(C3) as alinaw~naw 'pupil') is reduplicated. In some cases we find nonadjacent reduplications, i.e. the initial CVC-sequence and the final CVC-sequence are identical, but are separated by a vowel (i.e. the pattern is C1V1C2V2~C1V1C2 as gisá~gis (a 85 It is one research objective of the Graz Database on Reduplication to answer this question by collecting typologically balanced data. 86 Diseases in German often have (obligatory) plural marking. For example Maser-n 'measles', Rötel-n 'rubella', Halsschmerz-en 'sore throat', etc.


Lexical Reduplication

skin disease))87. One type is non-adjacent and non-contiguous, which is a cross-linguistically is a rare type. The initial consonant plus the final vowel and consonant, attached on the right side (i.e. C1V1C2V2C3~C1V2C3 as haging~hing (a whizzing sound)). Productively, in Bikol there exists only one of these formal patterns, namely initial CV- reduplication (cf. IV.4.1.2). Compared to the bisyllabic reduplicated roots, the origin of the different partial reduplicative structures within the root is less obvious. But their semantics is very similar to the bisyllabic type. The meanings are either "iconic" (plurality or sound) or "typical" (at least within the Austronesian language family). Some of the forms correspond quite clearly to the specific semantic fields: 3.3.2 Diseases and plurality The C1V1C2V2~C1V1C2 sequences denote diseases or pain on the one hand (giságis (a itchy skin disease of dogs), h{ar~}abá~hab 'stomach acidity', h{ar~}ití~hit88 'pain caused by new wounds), and actions with inherent plurality on the other hand (for example sayasay 'stagger when walking, walk unsteadily', wasiwas 'wave, swing'). 3.3.3 Sounds The non-contiguous, non-adjacent type, i.e. the C1V1C2V2C3~C1V2C3 sequence, denotes in most cases sounds (for example kagul~kol 'coughing sound', taging~ting 'sound of bamboo etc.') and bird names (tarik~tik 'woodpecker'), in which it is obvious that the bird names are derivations/transfers of the sound which is produced by the bird. In addition to the reduplicative structure, sound symbolism with respect to the vowels can be observed in these onomatopoetica: Words with back vowels (/a/ and/or /u/) denote dark, loud, or dull sounds (hagung~hong 'buzzing sound', pagukpok 'knocking sound'). Words with the front vowel /i/ refer to light, clarion sounds (haging~hing 'whizzing sound', sagiw~siw 'sound of reeds in the wind'). Key (1965) notes correctly that "sound imitation can be secured by other devices than by reduplication. So very often when R occurs in an imitative form, R is not the sole or the 87 Rubino (2001: 317-319) gives the same phonological word structure for sound denoting onomatopoetica and many other meanings in Ilokano. 88 In these words the PAN infix -ar- (cf. Blust 2003: 471) appears in addition to reduplication, which can be observed in several lexicalized forms of different types. Cf. also Rubino (2001: 314-316). Today plural is productively formed by the reduplicative infix -Vr- which might have been developed from -ar- (cf. IV.


Lexical Reduplication

primary device used" (Key 1965: 100). Rubino (2001: 319) states that in the Philippine languages in general sound symbolism and iconic patterns of word formation play an important role, and an even more detailed analyses might reveal that the relation between meaning and form is much less arbitrary than widely assumed. Although these symbolic and reduplicated terms for sounds resemble ideophones in some respects, they cannot be classified as such, because handoff some important differences between the two: First, unlike ideophones in other languages, these Philippine89 words do not have any unusual phonological characteristics. They only contain sounds or syllable structures which are found elsewhere in the language. Second, these lexemes behave exactly as all other lexemes of the language. They can be regularly and fully productively derived for all categories (cf. Rubino 2001). 3.3.4 Animals, mythological figures and Co. The roots with the second reduplicated syllable (i.e. C1V1C2V2~C2V2C3V3(C4)) denote primarily








(bula~langaw 'rainbow', bula~lakaw 'shooting star') and mythological figures (ana~nanggal 'witch', andu~duno 'vampire'). The roots with the final reduplicated syllable and the few examples with the initial reduplicated syllable cannot be classified unitarily. They refer for example to body parts (alinaw~naw 'pupil', bu~butkan 'wrist'), diseases or anomalies (bu~bu'a 'a puckered swelling occurring on the genitals of some women', bu~bungaw 'hernia, rupture'), plants (bagán~gan 'kind of grass'), perceptions (ramis~mís 'tasteless, sweet'), fish (butí~ti 'poisonous, small fish species') or mythological figures (ga~gambán 'hell', tambalus~los 'mythological forest') and can have plural or distributive meanings, also in the broadest sense like subay~bay 'continuous attendance to, for examples a series in TV', lupay~pay 'desperation', bikang~kang 'having the legs spread apart' or buhag~hag 'spread widely'. 3.3.5 General remarks on lexical partial reduplication As with all lexical reduplication types, there are some instances which seem (at least without etymological knowledge) not to fit in any of the categories listed above (e.g. halayhay 89 This can at least be stated for Ilokano (cf. Rubino 2001) and Bikol.


Lexical Reduplication

‘hanging the laundry; kaluskos ‘flange the skirt/trousers’). But in most cases it is not difficult to figure out a possible plural or distributive origin. The analysis of the different formal types of lexical partial reduplications resulted in no significant correspondence between certain forms and certain meanings. But nevertheless it is apparent that all lexical reduplicative structures on the whole have certain typical meanings. The comparative data from other languages lead to the assumption that the association of reduplicative structures with the mentioned semantic fields is a common feature of the Austronesian language family and probably also of other languages. Although in the individual language this holds true to very different degrees in the individual languages. Interestingly, the word families which at first glance are not related to one another, but which share the feature of having many members with reduplicative structures (of whatever form) in many Austronesian languages, are the same word families which Blust (2001) associates with the PAN prefix *qali-/kali-. He assumes an Austronesian tendency to lengthen with this prefix words denoting „creepy-crawly“ creatures, natural processes, „whirling“ shapes, words of motion, body parts, animals and mythological concepts. This leads to the assumption, that reduplication is probably, besides or in addition to the *qali-/kali-prefixation, a tool for acquiring the desired length of a lexeme. It is not clear whether the lexical reduplicative structures are consciously associated with the respective semantics by the speakers of today. If this is the case, it can only be the reduplication pattern itself which is associated with certain meanings, independent of the specific reduplication type. A lot of the terms containing reduplication are considered to be characteristic of the rural environments or of the former times (especially those which belong to the mythological domain) and in several instances they increasingly get substituted by English terms (e.g. bulalangaw > rainbow, bulalakaw > shooting star, amamatak > scorpion, aninipot > firefly, alimbu~buyog > bee, etc.). 3.4 Lexical full reduplication 3.4.1 Form The third group of lexical reduplication in Bikol consists of two per two syllables, i.e. C1V1C2V2(C3)~C1V1C2V2(C3), for example hibot~hibot 'pain of inflammation'. As with 67

Lexical Reduplication

productive full reduplication (cf. IV., there are no lexical reduplications with four identical syllables (*C1V1C1V1~C1V1C2V2), nor reduplications with more than four (i.e. two per two) syllables. 3.4.2 Reversative movements The range of meanings of lexicalized full reduplications is very wide, as it is the case for the partial reduplications described so far. And again, there is much overlapping with the meanings of the other types. Nevertheless there is one semantic field which seems to be exclusively associated with lexicalized full reduplication, namely the field of reversative90 movements, usually of loose, unstable things, e.g. a swaying back and forth, and/or sounds associated to them. For example gí'áy-gí'ay 'hobble', kawál-kawál or kiwál-kiwál 'hang loosely and move back and forth (as a loose trouser leg when s.o. walks)', lawí-láwi 'wave' or tangó'-tangó' 'shake (the head, due to age, illness)', idól-ídol 'sound of shaking', kadál-kadál 'rattling or clanking sound of something broken (e.g. the motor)' or nguráb-nguráb 'mumble, talk to oneself, babble so as not to be understood (as when talking in one's sleep)'. The second large group of this reduplication type denotes entities or actions or properties with the salient semantic feature "plural" or "distributive", for example bandyíng-bandyíng 'gallivant, roam around with nothing to do', or sari-sari (store) '(store which sells) various things'. Like the other lexical reduplication types, a certain amount of the full reduplications also denotes animals, plants and diseases/pain. For example gulíng-gulíng 'kind of snail', walówálo 'a sea animal', lubí-lubí 'plant (leaves are eaten as a vegetable)', labó'-labó' 'disease (swelling of the face and body)'. 3.5 Echo-words 3.5.1 Form Bikol echo-words consist of two per two syllables, and contain two alternating consonants. 91 In Bikol it is in most cases the medial consonant which alternates between the two 90 A movement that comes back to its starting point, i.e. „back and forth“, „in and out“, etc. The term is used by Dressler (1968: 65). 91 In other languages the alternating segment is often a vowel, like in Tibetan or Tamil (tam).


Lexical Reduplication

constituents, for example bulak~bugak 'gurgle', in some cases it is the initial consonant, for example ra'án~da'án 'hold a grudge against s.o.'. The formal analysis of the echo-words of my corpus resulted in the following regularity92: a) Alternation of the initial consonant: If the first constituent has initial /r/, this alternates with /d/ in the second constituent, for example rapák~dapák 'sound of running footsteps', riwág~díwag 'move from side to side'. ([rX-dX]) b) Alternation of the medial consonant: If the initial consonant is not /r/, then it is the internal consonant which alternates: - /g/ as internal consonant in the second constituent corresponds with /l/ in the first constituent, for example kalíng~kagíng 'jingling sound', kalá'~kagá' 'boil', saláy~sagáy 'describing a woman's voice when she yells in an argument'. - /d/t/s/ as internal consonant in the second constituent corresponds with /r/ in the first constituent, for example guró'~gudó' 'tremolo, gobble', piríng~pitíng 'shake the head (as when indicating negation), hurók~husók 'sound of falling water'. ([C1V1/l/V2(C3)-C1V1/g/V2(C3)] or [C1V1/r/V2(C3)-C1V1[+alveol]V2(C3)] if C1 ≠ /r/) This regular alternation is only valid for the echo-words as a closed group. It does not mean that every full reduplication changes the respective consonants in order to achieve the mentioned alternations. I.e. there are no echo-words with initial /r/ or /d/ which have medial consonantal alternation, but there are abundant exact reduplications with initial /r/ as for example rugong~rugong 'quarrel' or runggaw~runggaw 'a little bit topsy-turvy', and there are many exact reduplications with internal /g/, /l/ and /d/, for example lagaw~lagaw 'tramp, vagabond', palit~palit 'dangle' or radap~radap 'sound of fruits falling off the tree'. Although reduplication is a word formation which is obviously governed by a preference for equal structures (cf. “Identity Constraint”, Excursus II), it is noticeable that in echo-words a contrary "force" plays an important role, namely dissimilation. Interestingly, of all consonant pairs in Bikol echo-words, one segment of the consonant pair is a liquid. The sound class of liquids is cross-linguistically typically involved in dissimilatory phenomena (cf. Hurch 2005). From my data it cannot be concluded whether the echo-words have their origin in exact reduplication with a dissimilatory process having taken place at a later stage or 92 This regularity is not presented as a rule, because it is not a productive morphophonemic rule of Bikol.


Lexical Reduplication

whether the echo-words were originally formed as such. The latter would be a case of preventive dissimilation (cf. Hurch 2005). The striking overlap in meanings of echo-words and full reduplication could be seen as an argument for the development out of exact reduplication via a dissimilatory change. The observation that there are various consonants in the second constituent which all alternate with one specific consonant in the first constituent leads to the conclusion that the direction of a former dissimilation was regressive (cf. Hurch 2005: 720). As dissimilation is by nature a sporadic phenomenon, it is not surprising that the echo-words "co-exist" with exact reduplications containing the relevant consonants. 3.5.2 Meaning Semantically, the echo-words are very clearly associated either with movement, for example gurong~gusong 'rubbing, out of annoyance', kiríg~kidíg 'make short, quick steps' or with sounds for example rawít~dawít 'chatter' (also the term for a form of poetry (cf. III.1)), purág~pudág 'galloping sound'. To a small degree they denote surface properties, as for example garáp~gasáp 'rough' or giríng~gitíng 'a scalloped edge or border'. In some instances "movement" must be understood in an emotional sense, like for example ra'án~da'án 'hold a grudge against s.o.' or purók~pusók 'be irritated, annoyed'. Not only movements and sounds, but also surface properties can be associated with plurality93. I.e. in the case of jagged or rough surfaces a lot of spikes or a lot of indentations for example. With terms of movement there is a large degree of overlap in meaning between the exact full reduplication and echo-word reduplication. A meaning difference between exact full reduplication and echo-word reduplication cannot be found easily. The only distinction that suggests itself is that the movements denoted by exact reduplication seem to be "back and forth" in most cases (see above), while the echo-words usually denote sporadic movements without a concrete direction or even movements in many directions, which can lead to a distributive or even a "chaotic" interpretation (for example Wurong~wusong ako. 'I pull my hair out.'). However, it has to be emphasized that the co-existence of lexicalized exact reduplication and echo-words with similar meanings is not exceptional. See for example Keane (2001: 47-48) for Tamil.

93 For some echo-words my consultants accept a simplex form as understandable, but not correct, „because the movement takes place repeatedly“.


Lexical Reduplication

Even if the origin of the echo-words (though exact reduplication or by spontaneous creation) is not clear, it can at least be stated that synchronically echo-word formation is not a productive process. The consultants I worked with refused to create "new" echo words from existing simple forms as well as from existing exact reduplication. And vice versa, the echowords cannot be replaced by exact full reduplications (i.e. without sound alternation). So, echo-words in Bikol differ completely from productive echo-word formation, which exists in many other languages, for example in Hindi (cf. Singh 2005) or Tamil (cf. Keane 2005) or also in English (cf. Nevins and Vaux 2003). Cross-linguistically typical meanings of echowords include the so called "etceteras" (for example in Tamil kaappi kiippi 'coffee and other beverages', paampu kiimpu 'snakes and other reptiles/pests', Keane 2001: 57) and/or "pejoratives" (for example English linguistics-shminguistics 'linguistics and stuff', or Hindi bhagavaan-vagavaan 'crazy things like God' (Singh 2005: 266). Levinson (2000) subsumes the function of these echo-words as shifting "the interpretation away from the prototype extension of the word" (Levinson 2000: 153). Bikol echo-words however do not have this function and they are not productive. They denote specific movements and sounds. In this respect they have are more similar to echo-word forms as German and English Zickzack 'zigzag' and Ticktack 'tictac', or Tibetan lab lob 'murmuring', yang nge ying nge 'shaky, unsteady' etc. (cf. Mattes and Vollmann 2006). Like in Bikol, these forms have no corresponding simplex form94 and they denote sounds, shapes and movements. 3.6 Summary The typical meanings which are associated with reduplicated structures in Bikol are plurality, movement or sound, and, to a much smaller extent, animals, plants and food. Tables 11 and 12 give an overview of the numeric distribution of forms and meanings of lexical reduplication in Bikol. Interestingly, increasing semantic arbitrariness corresponds to decreasing formal transparency, as can be seen in table 12: echo-words > full reduplication > bisyllabic reduplicated roots > partial reduplication. Even if this might not hold true for productive reduplication, the distribution of the meanings of lexical reduplication support the hypothesis of Bybee et al. (1994), that formal reduction goes hand in hand with semantic bleaching (cf. Excursus IV). 94 One exception is German Zickzack which can be associated with Zacke 'jag, indentation' (cf. Dressler et al. 2005: 457).


Lexical Reduplication

Formal type



percentage 82








bisyllabic root






Table 11: Formal types of lexical reduplication in Bikol

meaning “iconic” (plural, formal type

“typical” (animal, total

movement, sound) plant, food)













bisyllabic root








Table 12: Distribution of form and meaning of lexical reduplication in Bikol

This distribution is not at all surprising or extraordinary. These form-meaning correspondences are highly iconic and very frequent cross-linguistically. Pott (1862) groups his examples under the headings such as “body parts”95, “animals”, “musical instruments”, etc. Kocher (1921) lists among others “body parts”, “diseases”, “food”, “movements”, and “sounds”, which are denoted by reduplicative word forms in Romance dialects and colloquial speech. Uhlenbeck (1978: 32-33) notes that names of plants and animals etc. have “peculiar” structures in Javanese. Hess (1966) gives several examples of terms for animals, plants, tools etc. with fossilized reduplicated forms in the Salish language Snohomish (sno).96 In Kröll (1991) a similar structure can be found for Portuguese reduplication. Procházka (1995) provides an overview of the most common meanings associated with reduplications crosslinguistically and in Arabic. Interestingly, Arabic lexical reduplications have many meanings parallel to the ones in Bikol. Procházka analyzed the data of three Arabic dictionaries. Thus his data are a good reference to compare with my own analysis of the Bikol dictionary, 95 “Weibliche Brüste” (pp. 31-35) and “Andere Körpertheile” (pp. 35-40). 96 “A number of words have shapes which suggest that they were once derivatives with chameleon affixes” (Hess 1966: 355).


Lexical Reduplication

especially because Bikol and Arabic do not share any genetic or areal affiliation. 97 Furthermore, it is in the nature of actions or events, that the iconic or typical meanings often overlap or coincide. I.e. the movements are causing sounds and usually are accomplished more than once (i.e. plurality). The words for animals, plants etc. in Austronesian have the tendency to be of longer than average length (cf. Blust 2001). Reduplication is one of the easiest means of lengthening a word. As already mentioned above, the unusual frequent appearance reduplicative structures in these lexical domains does not occur exclusively in Bikol or Austronesian, but can also be observed in other languages like Arabic, Salish, Tibetan or Portuguese (cf. also Gonda 1950: 174, Pott 1862: 35-39 (body parts), 60-63 (animals)). Even if about sixty percent of the lexical reduplications can be clearly classified as “iconic”98, this is sufficient to say that lexical reduplication is also highly structured. The universally available form-meaning pattern of reduplication can be used, and is obviously used frequently, but of course it need not be activated.

EXCURSUS I: ICONICITY OF REDUPLICATION Defining Iconicity „Iconicity“ refers to a specific type of relation between form and content. The term goes back to Peirce's Theory of Signs. In an iconic relationship the sign shares certain properties with its referent, i.e. the sign itself has a property that it refers to. (cf. Peirce 1904). Peirce distinguishes „imagic“ and „diagrammatic“ iconicity. The imagic type of iconicity refers to a sign which directly resembles its referent, as it is the case with onomatopoetic words for example.99 The diagrammatic type of iconicity refers to the systematic arrangement of signs which mirrors the arrangement of their referents. For example in Caesar's „Veni, vidi, vici“ the linear order of the propositional elements corresponds to the temporal sequence of the events they refer to (cf. Greenberg 1963: 103, Haiman 1980: 528, Haiman 2000: 282). One of the underlying assumptions is that grammatical structure is an iconic reflection of conceptual 97 Except for the Arabic influence in the Southern Philippines (Mindanao) which causes also sporadic Arabic loanwords in the Austronesian languages. But this influence is too weak in Bikol for assuming any contact phenomenon for an explanation for the similarities of the reduplication semantics. 98 Any cases of doubt were counted as arbitrary. In a less strict procedure of classification the percentage would be even higher. 99 Imagic iconicity also plays a role in many lexical reduplications denoting sound, cf. IV.3.


Lexical Reduplication

structure, or even more basic, that the language somehow mirrors the human conceptualization of the world (cf. Croft 2003: 203, Newmeyer 1992: 758-759, cf. also Langacker 1987). In the tradition of structuralist and later generative linguistics, the „autonomy of grammatical structure“ (or the arbitrariness of the sign) was a crucial assumption, and therefore the very old topic of „motivation of the linguistic sign“ was not really a matter of debate for decades. Jakobson (e.g. 1971) must be mentioned here as an exception: he was the first to take interest in Peirce's semiotic concepts and profoundly reflect on the relation between linguistic signs and their functions.100 With the emergent functionalism in linguistics, motivation and iconicity (i.e. the non-arbitrariness of the sign) came back into the broader focus of interest (cf. Givón 1991: 83-86). The usage of the term "iconicity" in linguistics (and other disciplines) varies considerably, however an iconic relation always refers to some sort of similarity between signs and their referents. Haiman (1980) further distinguishes two kinds of diagrammatic iconicity: „isomorphism“ and „motivation“. Isomorphism denotes the one-to-one-correspondence (biuniqueness) between sign and meaning. If the principle of isomorphism is considered as a universal tendency in grammar (cf. Haiman 1980: 518), different forms must be assumed to always have different meanings, and vice versa, one form is expected to have only one meaning. The huge amount of synonyms, homonyms and polysemes in most languages is a challenge for this hypothesis. Isomorphism in its strict sense obviously does not exist. Haiman (1980) for instance tries to solve this problem by tracing back homonyms to one underlying common meaning. For more details see the discussion of the polysemous full reduplication in Bikol ( The second type of iconicity, i.e. motivation, which was already highlighted as an essential principle of language by Jakobson (1963, 1971), refers to the direct reflection of meaning in a grammatical structure. In morphology, this means in the first instance that increased quantity in form is an icon of increased quantity in meaning, and increased complexity in grammar is an icon of increased semantic complexity (referred to as "constructional iconicity" or "diagrammaticity" in Natural Morphology).101 Iconicity and reduplication The most important aspect of iconicity with respect to reduplication is definitely that of 100 Cf. Haspelmath 2006: 2. 101 Haspelmath (2006: 1) thirdly mentions the iconicity of cohesion, which plays a minor role with reduplication, however.


Lexical Reduplication

quantity, i.e. greater quantities in meanings are expressed by greater quantities of form. Compared to other additive morphological procedures, in reduplication it can be restated as „more of the same form“ corresponds to „more of the same content“ (cf. also Kouwenberg and LaCharité 2001: 59). What linguists usually mean by iconicity of reduplication is, to quote Botha (1988): „... reduplication is a means of word formation that manifests a measure of iconicity: form and meaning resemble each other in a quantitative respect“ (Botha 1988: 3). The relationship between the base and the reduplicated word form directly reflects the relationship between their referents. But despite its highly iconic character, reduplication also always exhibits a high degree of abstraction, even in the case of „plural reduplication“. Firstly, the material of the base is usually copied only once 102, but it is rarely (if at all) used to express duality. It usually refers to much more than two incidences of entities or events. Secondly, in the case of partial reduplication, formally only a part of the base is copied, while semantically it is the whole entity or event that is pluralized. The iconicity of complexity (i.e. more complex meanings are expressed by more complex forms) means for reduplication simply that the reduplicated word is semantically more complex than the unreduplicated word. In this sense there is no difference in iconicity between reduplication and any other kind of affixation which adds meaning to the base. 103 And in this sense there is no doubt that diminutive reduplication, as for example Bikol full/Curu-reduplication, which is labeled as „non-iconic“ by many authors, is also clearly motivated because the word has at least one additional semantic feature (e.g. [+small]) in contrast to its corresponding unreduplicated base. So, reduplication is iconic with two respects: with respect to quantity and with respect to complexity. As a consequence of my observations of Bikol reduplication and partly of comparisons with other studies, I propose the following definition of “iconic” reduplication: Every reduplicated word form which expresses any kind of quantity change with respect to the meaning of the base (i.e. intensity, plurality, diminution, etc.) is an example of “iconic” 102 With view exceptions of „triplication“. For example in Fiji for marking extreme intensive (cf. Schütz 1985), or Mokilese for a distinction of progressive and continuative (cf. Harris 1973). The Niger-Congo language Mwera (mwe) has simple reduplication with polysyllabic roots and triplication with monosyllabic roots (cf. Bybee 1994: 160). I am not aware of an example of regular „quadruplication“. 103 „... within a grammatical correlation the zero affix cannot be steadily assigned to the marked category and a 'nonzero' (real) affix to the unmarked category“ (Jakobson 1963: 270).


Lexical Reduplication

reduplication, because the change of quantity in meaning corresponds to a change of quantity in form. The iconic principle is however not restricted to productive reduplication. Lexemes with reduplicative structure must also be considered as “iconic” as long as they refer to any kind of quantity (as it is the case for example with words denoting sounds or movements, cf. also IV.3). The formal and the semantic aspect of the iconicity of reduplication can be considered as two sides of the same coin. Thus, the degree of iconicity of reduplication can vary because of either formal aspects or semantic aspects or both. The formal aspect of iconicity refers to the degree of identity between the base and the reduplicant. On the highest iconic level, the base is fully copied and no further changes take place. When only a part of the base is copied, or when the reduplicant or the base undergo further changes that cause a formal difference between the two units, or when additional material is added to the reduplicant, formal transparency decreases. And consequently the iconicity of the means decreases, too. The case of „overapplication“ or „failure of rules“ (Wilbur 1974) is a measure to remedy such deviations from iconicity (for more details see Excursus II). With respect to the semantic aspect, reduplication is usually considered to be highly iconic when it refers to the plurality or the intensity of entities, events, states or properties. Other categories, which do not resemble the word form in this respect, reduce the degree of iconicity. In extreme cases reduplication loses all aspects of iconicity, i.e. formal and semantic, and is fully opaque. (As for example reduplicated lexemes in Sayula Popoluca (saw), as described by Rhodes 2004). Between these extremes of the continuum all degrees of iconicity can be found. Thus, iconicity must be regarded as a continuous concept, with variable transitions between different poles (cf. Kouwenberg and LaCharité (2001)104.

transparently iconic reduplications

opaque, non-iconic reduplications

Figure 6. The iconicity continuum for reduplication

104 What I labeled here as formal and semantic aspects of iconicity, is referred to formal and semantic iconicity by Kouwenberg and LaCharité (2001). But since "iconicity" is defined as a kind of relation between form and meaning, it does not make sense to separate formal iconicity on the one hand and semantic iconicity on the other hand.


Lexical Reduplication

Kouwenberg and LaCharité (2001) observe in Caribbean Creole languages, that formal and semantic aspects of iconicity of reduplication go hand in hand. This definitely does not hold true in Bikol: Full reduplication cannot categorically be classified as more iconic than partial reduplication, and exact reduplication cannot be classified as more iconic than reduplication with additional phonological material. „Non-iconic“ reduplication All instances of reduplication which do not denote plurality or intensity in its broadest sense are usually classified as „non-iconic“ in literature (for example reduplications for the derivation of word classes, for detransitive marking, etc.). For all kinds of diminutive meanings, sometimes even the term „counter-iconic“ can be found, because superficially, diminution is the antonym of intensity or plurality and as such it is the opposite of what should be expected from reduplication. But what does „non-iconic reduplication“ really mean? As already mentioned above, in the sense of iconicity of complexity, diminutive or any other „arbitrary“ meaning of reduplication is fully iconic, as long as it renders the semantics more complex as compared to the base. So, this cannot be the type of iconicity that is referred to when the term "non-iconic" is used. What is meant by that expression is the iconicity as direct reflection of the plurality of the „same“ in form and content. However, in it will be shown that diminution can also be analyzed as iconic reduplication in this sense, because it is just one aspect of the concept „quantity“, that is typically expressed by reduplication. The analysis of diminution as a semantic extension of the concept of plurality and/or the classification of the general function of reduplication to mark a „change of quantity“, just raises the iconicity of reduplication to a more abstract level than the „simple“ correspondence between more of the same form and more of the same function, but it does not violate the iconicity principle itself. From my point of view, the expression “non-iconic reduplication” is only appropriate for instances of „arbitrary meanings“ and/or formally highly opaque reduplications. Both seem to be relatively rare, in any case. I suppose that for some authors the differences in the notions of iconicity were not consciously taken into consideration. But especially for the justification of statements such as „xy is an example of counter-iconic reduplication“ it must be clearly defined what is meant by iconic or counter-iconic, respectively. Mayerthaler (1977: 28) for example states that full


Lexical Reduplication

reduplication is perfectly iconic for intensive and plural, whereas partial reduplication is iconic for diminution, because only a part of the base is reduplicated. Furthermore some statements on the iconic nature of reduplication seem to be built on misunderstandings or on an incomplete knowledge of the phenomenon. For example Lakoff and Johnson (1980) interpret the iconicity of reduplication in the sense that more form always represents intensified features of what is denoted by the base. I.e. plural or intense actions, plural or big objects, but diminutive of small objects (i.e. increased feature [+small]; cf. Lakoff and Johnson 1980: 127-128). This interpretation of diminutive reduplication which is actually an intensification of the smallness inherent to the base might be appropriate in some languages. But it is definitely not the only possible explanation of the diminutive function of reduplication, as can be seen from the Bikol example and many others. „Iconicity“ of lexical reduplication Lexical reduplications do not have a corresponding simplex form, as opposed to productive reduplications. Kouwenberg and LaCharité (2001: 64) correctly state that due to the nonexistence of an unreduplicated form, no morphological procedure is involved in the making of these word forms and therefore they cannot be mixed up with productive reduplication: „Only through the study of and reduplicant pairs can we determine the semantic and formal effects of morphological reduplications“ (Kouwenberg and LaCharité 2001: 64). Therefore, lexical reduplications lack the complexity aspect of iconicity (see above). But apart from that, many lexical reduplications have semantic properties which are very similar to those word forms which are generated by productive reduplication, as was demonstrated for Bikol earlier in this work, and as is stated for various other languages. It has already been argued above that lexical reduplication should also be described in terms of iconicity. However, this is not an uncontroversial issue. Kouwenberg and LaCharité (2001) for example exclude lexical reduplication105 from their study on iconicity, due to the high percentage of non-iconic instances. In the languages that they have examined, they found more “arbitrary” meanings of lexical reduplications than “iconic” meanings. They calculate that only 25 and 40 percent of all lexical reduplications in different Caribbean Creole languages are iconic. But for Bikol, I classified more than 60 percent of lexical reduplications as iconic (cf. IV.3.6), not to mention the many cases that are somewhere on the continuum between fully iconic and fully non105 “Pseudo-reduplication” (Kouwenberg and LaCharité 2001: 61).


Lexical Reduplication

iconic. And in Tibetan, which has no productive reduplication rule, as much as 93 percent of the lexical reduplications can be considered as iconic (cf. Mattes and Vollmann 2006). Such high figures, pointing towards interesting regularities, must not be ignored. And the fact that lexical reduplication is at the same time very often arbitrary, does not weaken the argument, because productive reduplication can have and does have non-iconic meanings, to a considerable extent in some languages, as well. In my view, it is obvious that there are semantic regularities within the group of lexical reduplications and that they can be classified as iconic and non-iconic in the same way as productive reduplications, as long as their different morphological status is kept in mind. This aspect of lexical reduplication shows that the grammatical and the lexical levels cannot and need not always kept strictly apart, what Dressler (1968) called "grammatisch-lexikalischer Mischcharakter innerhalb der Hierarchie des Sprachsystems" (Dressler 1968: 48). As was already discussed in 5.3.6, in many completely independent languages a certain percentage of lexical reduplications refers to semantic domains such as animals, plants and diseases. Not all of these word forms can be unambiguously classified as semantically iconic, but their cross-linguistic appearance leads to the assumption that there is some kind of motivation for this form-meaning relation. But as long as the nature of motivation is not clear, I propose to use three different categories for a semantic classification of lexical reduplication: iconic, non-iconic, and “typical” (i.e. the class of animals, plants, food etc., cf. IV.3.6). Because of its particular formal properties, i.e. copying phonological material of the base or at least a part of it, reduplication has definitely the potential to be used in a highly iconic way. And indeed, it very often is used to express „plurality“ or „intensity“, or more generally „quantity“. But abundant examples from many languages also show that reduplication can be conventionalized in almost the same manner as any other morphological means (cf. Haiman 2000: 287).


Productive reduplication

4 Productive reduplication 4.1 Partial reduplication 4.1.1 Introduction After the description of the lexical reduplication types of Bikol, the analysis of productive reduplication is provided in the next chapters. While productive partial reduplication is very transparent and regular, productive full reduplication is much more complex from the semantic point of view. Formally, there are only two types of morphological partial reduplication: CV- and -Vr-. But CV-reduplication has at least three completely different functions. The most frequent one is the marking of imperfective aspect, the second one is plural marking for ma-derivations (as a suppletive variant of -Vr-), and the third one is limitative marking for numerals. My suggestion is to assume three “homonymous” CV-reduplications, not one polysemous type, because their functions and domains differ considerably. No ambiguity arises between them because their respective meanings are clearly defined by the environment in which they appear. Limitative CV- appears only with numerals, plural CV- is restricted to ma- statives and stands in complementary distribution with -Vr- (therefore it is described in, whereas imperfective aspect marking can appear in all environments. Other authors assume even a forth type, namely CV- nominal derivation. But as will be argued below, this is just a special use of imperfective reduplication, at least diachronically. -Vr- reduplication for plural actor marking is especially interesting, because it is infixed, and as such is a rare type crosslinguistically. 4.1.2 Imperfective CV- reduplication The most regular reduplication type which exists in Bikol is the imperfective aspect CVreduplication. It indicates that an action is not completed. In combination with morphemes for begun aspect (nag-, pig-, -in-, cf. (53)) it marks a present action which is not completed, and therefore ongoing. Without these affixes for begun aspect the imperfective reduplication is used to mark future or intended actions, cf. (54).


Productive reduplication

53) Nag-ngi~ngirit! BEG.AV-IMPFV~laugh '(You are) laughing!' (uttered as a comment on an ongoing process)


54) i-di~diretsyo niya an pag-istorya kan tungkol sa Ibalong NMLZ-story ARG.SPEC concerning LOC Ibalong CV-IMPFV~continue 3SG PB 'he has the intention to / he will continue the story about Ibalong'


Like all TAM- and voice-affixing, imperfective reduplication is fully productive. It can be applied to every content word, independent of its semantic category (cf. III.2.6.). And it occurs not only with roots denoting actions or events but also in nominalized word forms, denoting concrete objects or abstract entities (see below). It even works with the replacement word ano 'what', as in 55)A: Dali na, Nay, ta nag-a~ano ang ... B: ... nag-a~andar ang ...106 hurry already Nay because BEG.AV-IMPFV~what PB ... ... BEG.AV-IMPFV~go PB ... A: 'Quickly, come on, Nay, because it is what ...' B: '... it is running' [calongay]

Furthermore there are no restrictions for loan words or foreign words (e.g. nag-ba~bayle107 'dancing', nag-be-besbol108 'playing baseball'). Form The form of imperfective reduplication is achieved by copying root initial C 1V1. If the base begins with a consonant cluster, this is reduced in the reduplicant to CV, e.g. trabaho 'work' --> nag-ta-trabaho 'is working' (*nag-ra~rabaho; *nag-tra~trabaho) or drive --> pig-dadrive 'is riding' (*pig-ri~drive; *pig-dri~drive), as in 56) nag-a-bantay siya kan nag-a-pa-harong kan nag-ta~trabaho BEG.AV-PRES-watch 3SG.AF ARG.SPEC BEG.AV-PRES-CAUS-house ARG.SPEC BEG.AV-IMPFV~work [magana] 'he watches the work(ing) of the house constructors' 57) Ano an sa-saki-an na pig-da~drive=mo? what PB DERIV-ride-DERIV LK BEG.UG-IMPFV~drive=2SG 'What car do you drive?' (lit.: 'What is the vehicle ridden by you?)

106 The speakers are talking about a tape recorder. 107 From Spanish bailar 'to dance'. 108 From English baseball.


Productive reduplication

If the vowel of the base is a diphthong, only the first element, i.e. a simple vowel is reduplicated, but not the whole diphthong. Example (58) is pronounced as [pig-da~draif] (*[pig-dai~draif]). Other examples are nag-ba-bayle 'is dancing' (*[nag-bai-baile]) or nagka~kayod 'hard work' (*[nag-kai-kaiod]). It is not possible to analyze the bases as containing a biphonematic vowel sequence, because the language does not permit onsetless syllables (cf. III.2.2.1). All vowels in the language are separated by consonants. In the orthography many words with vowel sequences are found, but they are all pronounced with a glottal stop (for example [kaon] 'eat', [kua] 'take', [raot] 'bad, broken'). This means that the reduplicant is restricted to contain only a simple vowel and the diphthong is reduced to its full element. This reduction of diphthongs is reported also from other – nonAustronesian - languages, e.g. Brazilian Portuguese hypocoristic kin terms, e.g. pa~pai 'father', ma~mãe 'mother' (Comrie 1989) or Kwaza plural reduplication, e.g. bu~bui-da-ki 'they (pl.) went outside', da~dai-wa-ki 'they (pl.) took' (cf. van der Voort 2003: 74-78). This data supports the arguments against the well-established monophonematic interpretation of diphthongs. Berg (1986) draws a conclusion based on his analysis of speech errors that diphthongs are “two discrete atoms on the segmental level which are molecularized in the suprasegmental structure.” (Berg 1986: 198)109 Long before that, Pike (1947) had already suggested such a hierarchical view of diphthongs. The fact that in reduplication just one element of a diphthong can be reduplicated and not necessarily the diphthong as a whole, is only explainable if the diphthong is assumed to consist of two segments which despite their cohesiveness can be separated from each other by the speaker. Figure 7 shows the formalization of the rule for imperfective reduplication in Bikol, following the word-based rule model110:

109 Berg (1986) refers only to German and English diphthongs because his study only includes data from these two languages. 110 The square brackets each contain a formal representation and other relevant semantic and/or grammatical information of a certain word form. The word forms stand in a paradigmatical relationship, symbolized by ↔. Word-based rules have been proposed among others by Becker (1990) or Ford and Singh (1991). Cf. also Haspelmath (2002).


Productive reduplication

/C1(C2)V1(V2)X/ 'x'

/C1V1C1(C2)V1(V2)X/ 'imperfective x'

Figure 7. Rule for imperfective aspect reduplication

Most other related Philippine languages also have imperfective reduplications, but their forms vary slightly from language to language. In Tagalog (and some other Central Philippine languages), the imperfective CV-reduplicant always has a long vowel, irrespective of the vowel length of the base. In Ilokano (and some other Northern Luzon languages), imperfective aspect is marked by CVC-reduplication, which seems to be preserved as such from Proto-Extra Formosan (PEF), where CVC-imperfective was formally distinct from plural CV-reduplication (cf. Reid 1992: 78). –

PEF *i-beleng 'throw out' - *i-bel~beleng 'is throwing out' (Reid 1992: 79)

Ilokano i-belleng 'throw out' – i-bel~belleng 'is throwing out' (Reid 1992: 69)

Tagalog i-tapun 'throw out' – i-ta:~tapun 'will throw out' / i-t-in-a:~tapun 'throws out' (Reid 1992: 85)

Bikol i-tapok 'throw out' – i-ta~tapok 'will throw out' / i-t-in-a~tapok 'throws out'111

However, according to Blust (1998), CV- imperfective forms do not most likely originate in an older CVC- reduplication, as it exists in Northern Philippine languages. CV- imperfective is rather a development of PAN Ca- reduplication. Ca- reduplication is still found expressing future aspect, and durative or iterative meaning in several Austronesian languages (for example Thao (ssf), Puyuma (pyu) or Tetun (tet)). Blust associates this Ca- reduplication with CV-reduplication for future or contemplated aspect in some languages (for example Rukai (dru) and Tagalog) and assumes a change from Ca- to CV- as “a type of unmarking that is attested for various languages in numerals, verbs and instrumental nouns” (Blust 1998: 35). If the CV- reduplication in Bikol (and other Austronesian languages) indeed has its origin in a reduplicant with a fixed segment, this clearly contradicts Niepokuj's (1997) hypothesis that reduplicants with fixed segments unidirectionally develop from pure reduplication (cf. Excursus IV). A change from Ca- to CV- is easily explainable for Austronesian languages. /a/ 111 In Bikol, as in Tagalog, the reduplicant is CV even if the base initial syllable is closed. The coda is never reduplicated. For example nag-tu~tugtog 'playing and instrument' (*nag-tug~tugtog).


Productive reduplication

is probably the most frequent vowel. For all roots of the shape CaCVC, Ca- reduplication can easily be reanalyzed as CV- reduplication. Of course, the converse change could be explained in the same way. Comparative data suggest however, that CV- reduplication has evolved from Ca- reduplication and not vice versa (cf. Blust 1998: 35). If the imperfective reduplication appears together with a prefix which causes assimilation with the following consonant (as for example mang- or pang-, cf. III.2.1.2.), this assimilated consonant appears also in the base which actually does not meet the conditions for the phonological change. I.e. these examples represent a typical case of “overapplication of a phonological rule”. 58) na-mu~murak sa altar nin sakuya-ng pagka-moot ... BEG.AV.GENR-IMPFV~flower LOC altar ARG 1SG.LOC-LK DERIV-love [angelina] 'it is blooming on the altar of my love ...' In (58) burak 'flower' is marked for “begun general action” by the prefix nang- and for imperfective aspect by reduplication: nang- + CV- + burak --> namumurak 'decorating with flowers' (*namuburak). Other examples are tubod 'faith': nang- + CV- + tubod --> nangungubod 'believing' (*nangutubod) or nang- + CV- + sublí' --> nanunublí' (*nanusublí') (Mintz 2004: 36).

EXCURSUS II: THE “IDENTITY CONSTRAINT” Example (58) represents a common phenomenon in reduplicative structures, namely overapplication of a phonological rule. The phonological rule in question is nasal assimilation, which is very common across languages. In many languages there are such cases of reduplication where a certain phonological rule applies in an environment which usually does not trigger the respective process. Or vice versa, a phonological rule does not apply, although the environment would require it (“Failure of rule”, cf. Wilbur 1973: 18-26). If “overapplication“ was not taken into consideration, the output of the rules would be *na-mu~burak (nang- + mu~burak) in example (58) or *na-mayad~bayad (nang + bayad~bayad) in example (93). The initial consonant in the reduplicant regularly fuses with the final nasal // (cf. nasal assimilation processes in Bikol III.2.1.2). But there is no phonological condition for changing the base initial consonant to a nasal. The fact that it is substituted by the nasal anyway can only be explained by the preference for maintaining the 84

Productive reduplication

identity of base and reduplicant, also in cases where they would otherwise differ from each other due to regular phonological changes. A theory which bases the concept of reduplication predominantly is based on the assumption of a strong tendency to the identity between base and reduplicant is Optimality Theory (OT). But it was Wilbur 1973 who first highlighted the problem in her dissertation and described the character of “overapplication” and “failure of rule” and introduced the term “Identity Constraint”. She demonstrated, that in many languages the tendency “... to generate forms where Rr is an identical copy of Ro” (Wilbur 1973: 15) is strong enough to influence the phonological rules in different ways. For example in Bikol nasal assimilation, which usually takes place automatically at morpheme boundaries when the left morpheme ends in //, takes place even without a preceding nasal in order to harmonize the base with the reduplicant. In the Bikol examples it is the initial consonant of the reduplicant that is changed regularly and it is the initial consonant of the base that is adapted consequently even without the appropriate environment. However, the Bikol examples of “overapplication” of nasal assimilation could still be easily described without any “Identity Constraint” by rule ordering. I.e. if the prefixation and the nasal assimilation are assumed to precede reduplication, the existing forms are the logical result: 1. nang- + sublí' --> na-nublí'; 2. na-nublí' + RED --> na~nu~nublí', or 1. nang- + bayad --> na-mayad; 2. na-mayad~mayad. In this case the prefix + root derivation is assumed to operate as an input for the reduplicative process. And indeed, this is argued to be the case by several authors. For example French (1988: 52) in her Autosegmental analysis defines the following ordering of morphological procedures for Tagalog: “... derivational affixation (i.e., prefixation and suffixation only) occurs first, followed by reduplication followed by infixation.” (Cf. also Bloomfield 1933: 222). But if, in contrast, the root is reduplicated before it is prefixed (i.e. nang- + su~subli --> nanunubli or nang- + bayad~bayad --> namayad~mayad, the nasalization of the root can only be explained by backcopying. The ordering of the different derivational procedures is not completely clear for Bikol. However, some instances with polysyllabic prefixes where a syllable of the prefix is reduplicated instead of the root (cf. IV. lead to the assumption that prefixation precedes reduplication in Bikol.112 Furthermore, as Wilbur (1973: 61) argued before for 112 The contradictory data from Bikol with this respect gives reason to reconsider the assumption that


Productive reduplication

Tagalog (where the same phenomenon exists), the explanation of “overapplication” by rule ordering does not work for many other languages. If the observations on reduplications in many languages are taken into account, the tendency to preserve the identity of base and reduplicant turns out to be quite a general principle and the phonological appearance of reduplications in many cases cannot be explained by any other model. 113 So it is sensible to assume this constraint also for Bikol (and other Philippine languages) instead of explaining overapplication by a specific rule ordering. Furthermore, the ordering of the reduplication rule after the assimilation rule is quite problematic because almost all theories assume that all morphological rules apply before any phonological rule. The idea that rules should not be ordered strictly serially was introduced tentatively by Wilbur (1973). She formulates a less linear view of rules and their arrangements by claiming that a phonological rule must be assumed to have more information about the word form than just the output of the preceding rule, i.e. a rule must be allowed to “look back” (Wilbur 1973: 72). Later in OT, the idea of serial ordering of rules is completely abandoned. The constraints are ranked with respect to their hierarchical power, but temporal/serial ordering is irrelevant. They are assumed to be evaluated simultaneously. McCarthy and Prince's notion of “reduplicative identity” is a property of the output, which can never be lost in the process of word formation (cf. McCarthy and Prince 1995: 288). As in all formal treatments of reduplication, in OT the characterization of a reduplicant essentially refers to a corresponding “base” (cf. Shaw 2005: 161). Against the backdrop of the pioneering work of Wilbur (1973) on principles of the phonology of reduplication, the “BaseReduplicant Correspondence Theory” (BRCT) was developed within the OT approach, mainly by McCarthy and Prince (1995). The BRCT assumes the reduplicant and the base to stand in a reciprocal phonological relationship. This means that not only the base can influence the form of the reduplicant but also vice versa, the reduplicant can influence the base. The rationale behind this principle is to preserve as much as possible the similarity of base and reduplicant. The central notion of BRCT is that of “correspondence”. In the BRCT two different correspondence relations are distinguished: The correspondence of reduplicant morphological rules obey a fixed ordering. Cf. IV. 113 In Sundanese (sun), a Malayo-Polynesian language spoken in Indonesia, nasalization of reduplicated word forms usually does not affect the base initial consonant. But interestingly, there are free variations with both consonants being nasalized (cf. Robins 1959: 338). This free alternation is a strong argument for an identity preference of the speakers, and cannot be explained by rule ordering, which would not allow free variation.


Productive reduplication

and base and the correspondence of phonological input and output. Based on these principles, McCarthy and Prince (1995) establish several universal constraints with respect to “identity” and “faithfulness”, which are separately valid for every correspondence relation and therefore can also be ranked separately. The ranking of the universal constraints is a language specific matter. The constraints reflect the general tendency to preserve the identity of base and reduplicant, and of input and output (in the generative sense). These structure preserving constraints, which are very central in OT, guarantee cognitive simplicity, i.e. higher intelligibility (cf. Hurch 1998: 127). The superordinate constraints with this respect are: –

The MAX Constraint Family (i.e. “Every segment of S1 has a correspondent in S2.”)

The DEP Constraint Family (i.e. “Every segment of S2 has a correspondent in S1.”)

The IDENT (F) Constraint Family (i.e. “Correspondent segments are identical in feature F.”)

(McCarthy and Prince 1995: 271-272) The basic model of McCarthy and Prince's Correspondence Theory may be illustrated as follows (cf. McCarthy and Prince 1995: 273): Input:

/AfRED +

Stem / I-O-Faithfulness




B-R-Identity “The faithfulness/identity constraints demand completeness of correspondence and identity of correspondent elements.” (McCarthy and Prince 1995: 274) Following these constraints, the “ideal” reduplication is exact full reduplication. Every deviation from this ideal form has to be explained by one ore more other constraints which are ordered higher in the hierarchy than IDENT, MAX or DEP and thus rule out the faithfulness or identity requirements (for example a markedness constraint, cf. Excursus III). What leads to overapplication forms, like the ones in (58) or (93), is the fulfillment of the B-R-identity requirements. The reduplicant takes all the features of the base, or vice versa, the base takes the attributes of the reduplicant. Within Naturalness Theory (NT henceforth), one could explain the phenomenon by two tendencies: the one towards syntagmatic identity and the other one towards paradigmatic identity. In cases of overapplication or underapplication, these two organizational principles 87

Productive reduplication

are in conflict with each other. Reduplication types (or more generally, languages) differ in their preferences for one principle or the other. Among the formal theories, OT is no doubt a very adequate theory to describe the phonology of reduplication which often behaves exceptionally with respect to the overall phonology of the language. Such an exceptional behavior is not really surprising, as the phonology of reduplication belongs to the morphophonemic level. OT abandoned the mechanistic view of language processing of the “classic” Generative Grammar and brought into discussion a new non-linear representation (cf. II.4.2.4). From psycholinguistic studies it is known that language processing involves “top-down processes” as well as “bottom-up processes” at the same time. A model which assumes a parallel evaluation of different constraints therefore respects much more the results of cognitive research than the purely serial rule ordering of Generative Grammar. One of the problematic points in OT, as in most other formal theories, is that reduplication is described from a purely phonological perspective. Of course, with reduplication exceptional phonological phenomena can often be observed and therefore the phenomenon is indeed very relevant for phonology, also from a general perspective. But still, reduplication is a morphological procedure, and this fact should not be completely ignored in theoretical discussions, even if they focus on formal aspects. The strong identity constraint in reduplicative structures is motivated by the iconicity of the procedure: The multiple and identical occurrence of one event or object or state, semantically speaking, is formally expressed by multiple (usually two) identical phonological strings (cf. Excursus I). The more the base and the reduplicant differ from each other phonologically, the more iconic power of the morphological means is lost. The BRCT for example is never discussed with respect to its iconic motivation, which it undoubtedly has, but it is exclusively treated on a phonological level. In my judgment this mapping of the meaning onto the form should not be ignored also in „phonological theories“.114

114 The “Morphological Doubling Theory” (MDT) (Inkelas 2005, Inkelas and Zoll 2005) in contrast to other formal theories, starts from a purely semantic correspondence in reduplication, from which phonological correspondence can emerge.


Productive reduplication

EXCURSUS III: THE PREFERENCE FOR THE UNMARKED SYLLABLE IN REDUPLICATION The Bikol reduction of base initial consonant clusters to a simple CV-syllable in the reduplicant by deletion of the second consonant, i.e. the maintenance of the initial consonant, like in trabaho 'work' --> nag-ta~trabaho 'is working' (cf. example (56)) is not surprising at all. Despite the fact that this reduction contradicts the identity principle (cf. above). The universal preference for the core syllable (CV) has broad evidence cross-linguistically and in observation on language acquisition. CV is the first syllable which is acquired by children. All languages of the world possess CV syllables. Many languages do not allow any other syllable structure; of course many other languages do, but to very different degrees of deviation of the core syllable (cf. Vennemann 1988: 18). In general, in cases of cluster reduction (in reduplication, in language acquisition, speech errors, etc.) usually the initial consonant is maintained. This is true not only in Bikol or Tagalog (cf. Rubino 2005b: 18), but also in other languages (for example Vedic prā 'fill' --> pa~práu 'has filled', cf. Kulikov 2005: 432; Sanskrit svar 'sound' --> sa~svar-a 'has sounded', cf. Steriade 1988: 120, or Klamath (kla) pnia:k' 'wild onions' --> pi~pnia:k' 'little wild onions, cf. Barker 1963: 302).115 This cross-linguistic preference for the maintenance of the first consonant in cluster reduction has two cognitively plausible explanations: First, the initial consonant of a consonant cluster is the most salient one, and as such has probably more “lexical load” (i.e. for word recognition) than the other consonants of the cluster. If a choice needs to be made, it is reasonable to maintain the element that is more important for speech perception rather than the less important one.116 Second, the universality of syllable structures is traditionally explained by the “sonority hierarchy” or “strength hierarchy”. The most preferred syllable, following for example Vennemann's “syllable laws”, is one which has the sharpest sonority contrast between onset and nucleus. “A syllable head is the most preferred: (a) the closer the number of speech sounds in the head is to one, (b) the greater the Consonantal Strength value of its onset, and (c) the more sharply the Consonantal Strength drops from the onset toward the 115 Languages differ with respect to st- and sp- clusters. In Vedic for example it is the second, i.e. the less sonorant consonant, e.g. tí~sthati 'stands' (Kulikov 2005: 432), that is preferred. In Bikol this problem does not arise, because neither st- nor sp- is allowed, not even in loan words. Therefore a vowel is inserted in front of the cluster, which is then reduplicated (e.g. nag-i-spanish --> nag-i~i-spanish 'speaking Spanish'). 116 Psycholinguistic tests showed that in speech perception the initial consonant of a word activates possible “candidates” for the recognition of the uttered word (cf. for example Lahiri and Marslen-Wilson 1991). It would be an interesting to study how reduced consonant clusters influence the recognition of the lexical item.


Productive reduplication

Consonantal Strength of the following syllable nucleus.” (Vennemann 1988: 13-14) If a syllable with a consonant cluster is reduced to CV, the resulting syllable is of course closer to the “ideal” syllable if the first consonant with the greater consonantal strength is maintained than with the second or third consonant. Thus, the reduction of the consonant cluster in the reduplicant results in a more unmarked and universally preferred syllable structure. In addition, also the reduction of diphthongs to monophthongs in the reduplicant produces a less marked syllable, because monophthongs are universally preferred over diphthongs as syllable nucleus (cf. Vennemann 1988: 27). On the other hand, however, by improving the syllable structure through consonant cluster reduction and monophtongization, the similarity between reduplicant and base, and thereby also the iconicity of the procedure, is reduced. The preference for CV syllables is a very important and often cited argument in Natural Phonology. Basically, NT explains phonological processes from an articulatory or perceptual perspective.117 In general all phonological processes that can be observed in a language are caused by one of two antagonistic principles: “fortition” and “lenition”. The first principle leads to an improvement of perception, the latter to a simplification of articulation. Evidence is usually taken from phonetics, from first and second language acquisition, from speech errors etc. The CV-syllable can even be used as an example of phonological structures which are preferred for both, articulatory and perceptual reasons (cf. Hurch 1988: 13). The problem of dealing with consonant clusters in Bikol reduplication is a rather “young” one. The form of the reduplicant for imperfective reduplication is CV-. In native Bikol words, which exclusively have CV(C)- syllable structure, the conflict between producing the preferred syllable structure on the one hand and keeping the identity between base and reduplicant on the other hand does not exist. When loan words from Spanish, and much later from English, were adopted in the language and with them the syllable internal consonant clusters, the form of the imperfective reduplicant obviously did not change but was instead reinterpreted as C1V1-. The strong universal and language specific preferences for a CVsyllable structure leads to the reduction of the consonant cluster in imperfective reduplication,

117 Of course a lot of other parameters have additionally been taken into consideration by different authors. For detailed information about Natural Phonology see e.g. Donegan and Stampe 1979, Dressler 1984, Hurch and Rhodes 1996.


Productive reduplication

even if this violates the iconicity principle of morphological procedure118 (cf. Excursus II). One of the basic assumptions of NT is that the dominance of one preference inevitably leads to the suppression of another conflicting preference. “Die inhärente Widersprüchlichkeit von Prozessen und deren Vermittlung ist ein essentielles Bewegungsprinzip in der natürlichen Phonologie.” (Hurch 1988: 19) The fact that the formation of the preferred syllable structure in the reduplicant means a loss of identity between base and reduplicant at the same time causes problems for theoretical explanations which would not exist without the Spanish and English borrowings (cf. also French 1988: 1). As was already explained above (cf. Excursus II), the basic concept of reduplication within Optimality Theory is that of identity. Any deviation of perfect identity between the two constituents has to be explained by specific constraint ordering. In the case of the strict preference for a CV-syllable as reduplicant, a markedness constraint must be assumed to dominate the identity constraint. Based on the observation that reduplicants often have a less marked phonological structure than their base, OT introduced a specific term for cases in which markedness dominates faithfulness: “The Emergence of the Unmarked”. “If some markedness constraint M is crucially dominated by all relevant IO faithfulness constraints FIO, then satisfaction of M cannot produce unfaithfulness in the IO mapping, and some M-violating surface forms will be observed. But if the same M crucially dominates a faithfulness constraint, FBR, that governs the BR correspondence relation, then M will be obeyed in the reduplicant, even at the expense of inexactly copying the base. This situation is called the emergence of the unmarked (TETU).” (Alderete et al. 1999: 330) Languages allow the violation of constraints to different degrees and in different hierarchies. With the apparatus of language specifically ranked constraints the theory can choose the optimal form out of a number of several possible forms. The case of Bikol consonant cluster reduction is a clear case of TETU. The MAX Constraint is violated, while DEP and IDENT are still fulfilled. Not all of the segments of the base (S1) have a correspondent in the reduplicant (S2). For reduplicants which always have a mono-consonantal onset, Urbanczyk (2001: 99) introduced the markedness constraint *COMPLEX ONSET which dominates MAX. Of course there are several other theoretical solutions offered to the mentioned problem. 118 On naturalness in morphology, where the iconicity principle is a very important, see for example Dressler et al. (1987), Mayerthaler (1981), Wurzel (1984).


Productive reduplication

French (1988) for example provides an analysis of the phonology of Tagalog reduplication (which is identical to Bikol with respect to consonant clusters) within the framework of Autosegmental Theory. She argues for a template driven association of the segments of the base to the reduplicant in order to be able to describe the phenomenon of consonant cluster reduction (cf. French 1988: 32-36). I will not go into further detail with this theory. In my view it has no explanatory value but only solves the problem of how to formally illustrate the phonological processes of reduplication. Even if NT and OT have very different methods to describe phenomena like the consonant cluster reduction (NT avoids formalizations whereas they are the central exercise in OT), they have some basic ideas in common; for example the interaction of antagonistic forces (constraints/preferences) and the assumption that these are universal concepts which have different strength in the various languages.119 Base of reduplication When the imperfective reduplication appears with actor voice prefixes (mag-/nag- etc.), the root initial syllable that is reduplicated is in most cases the second syllable of the whole word. But also if the root initial syllable in prefixed words is not the second syllable (because of polysyllabic prefixes like abilitative or sociative maka-/naka-), it is nevertheless the second syllable of the word which serves as a base for the reduplication. This pattern is not regularly observable. Probably, there is an ongoing diachronic change in this respect. In my corpus both cases can be found, i.e. reduplication of the root initial syllable (59), (60) as well as reduplication of the second syllable of the prefix (61), (62), with a noticeable frequentative dominance of the latter case. 59) maka-wi~wili asin and NBEG.ABIL.AV-IMPFV~engross 'fascinating and romantic to look at'

romantiko-ng romantic-LK

pag-hiling-on, DERIV-see-UG [r-d: Bulanon]

60) dai

pa naka-ta~tapos kan k{in}u~kua niya-ng NEG still BEG.ABIL.AV-IMPFV~finish ARG.SPEC {BEG.UG}IMPFV~obtain 3SG-LK '(she) was not yet able to finalize her career.' (lit.: (she) was not yet able to obtain the end of her career.')

119 For a critical examination of the similarities and differences of OT and NT see Hurch (1998).


karera. career [paurog]

Productive reduplication

61)... kita-ng duwa an naka~ka-aram kan ka-totoo-han ... 1PL.EXCL.AF two PB BEG.ABIL.AV~IMPFV-know ARG.SPEC NMLZ-true-NMLZ [bunga] '... we are the two who know the truth.' 62)permi-ng man siya naka~ka-bohe, permi-ng siya naka~ka-dulag. always-LK also 3SG.AF BEG.AV.ABIL~IMPFV-escape always-LK 3SG.AF BEG.AV.ABIL~IMPFV-free [ibalong] 'he also always managed to escape, he always managed to free himself.' The same phenomenon is observed in Tagalog. Schachter and Otanes (1972: 227) list both variants without any comment on productivity or diachrony. French (1988: 64-68), within her autosegmental approach to the phonology of Tagalog reduplication, also does not go into detail with this respect, but at least she classifies the root initial reduplication as archaic and formal, and the reduplication of the second syllable of the prefix as the modern and conversational form. Indeed, in my corpus of spontaneous spoken speech, no example of root initial CV-reduplication can be found. In the written data as well the prefix-reduplication clearly prevails. A similar ongoing change can be observed with other prefixes; in my corpus very often with the causative prefix pa-, which most frequently is the syllable to be reduplicated instead of the root initial syllable. 63) Nag-pa~pa-isog nin daghán, nag-pa~pa-kusog nin boot ARG chest feeling BEG.AV-IMPFV~CAUS-brave BEG.AV-IMPFV~CAUS-strong ARG [r-d: basabas] 'it makes you brave the chest, it strengthens the feelings' But this pattern is variable, even within one sentence, as e.g. in 64) dai nang~ang~ahulugan120 na pa~pa-baya-an=ta an “Filipino” NEG BEG.AV.GENR~IMPFV-mean LK IMPFV~CAUS-abandon-UG=1PL.INCL PB “Filipino” 'this does not mean that we neglect the “Filipino”' and 65) mayo man nin pa-du~duman-an. NEG also ARG CAUS-IMPFV~DEM.DIST.LOC-UG 'no use' (lit.: 'this does not lead to any success')

[cf. Carpio 2000: 146-147]

As is the case with the naka-/paka- prefix, root initial reduplication in words with the causative prefix pa- is rarely found in the written corpus, and no example of this can be found in the spoken data. Based on this observation I claim that the original rule for forming the imperfective aspect by root initial CV-reduplication has almost completely changed and is 120 nangangahulugan < nang- + IMPFV + kahulogan 'mean/signify', cf. III.2.1.2.


Productive reduplication

substituted by a rule with the domain of the whole word form: Words with a prefix for begun action and actor voice (nag-, naka-, naki-, pig- etc.) reduplicate the second syllable for marking imperfective aspect. Other word forms reduplicate the first syllable. Imperfective reduplication and infixation The imperfective reduplication can appear in combination with the infixes -in- (for begun aspect) and -Vr- for plural actors. In this case -in- is infixed into the reduplicant, not into the root121. 66) Sato-ng b{in}a~balik-an 1PL.INCL-LK {BEG.UG}IMPFV~return-UG 'Our returning to the mother tongue.' but: *bi~b{in}alik-an

an PB

ina-ng mother-LK

tataramon. language [cf. Carpio 2000: 146]

The plural infix -Vr-, on the other hand, is always infixed into the root, i.e. in this case the imperfective aspect reduplicant is not affected by the infixation122: 67) an mga sira nag-ba~b{ar~}ayle ta sinda PB PL fish BEG.AV-IMPFV~{PL~}dance because 3PL.AF 'The fish are dancing because they are very happy' but: *nag-b{ar~}a~bayle

ma-ogma-hon ST-joy-INT [r-d: Bulanon]

Consequently, if all the three morphemes, i.e. CV- imperfective aspect, -in- begun aspect and












{BEG.UG}IMPFV~{PL}root-UG, as for example bakal 'buy' --> ba~bakal 'buying' --> ba~bar~akal 'buying, pl.' --> b-in-a~b-ar~akal 'is buying, pl.'.

68) B{in}a~b{ar~}akal=ko ang mga {BEG.UG}IMPFV~{PL~}buy=1SG PB PL 'I am buying (many) eggs.'

sugok. egg Output Constraints Some examples show that there is no blocking for reduplications which would create three subsequent identical syllables. Such a constraint might be expected, because full reduplication is blocked in cases where it would create four subsequent identical syllables 121 In unreduplicated bases, -in- is infixed after the root initial consonant, e.g. balik --> b-in-alik 'bought'. 122 For more detailed explanations and argumentations for the different behavior of the -in- infix and the -Vrinfix see IV.


Productive reduplication

(e.g. *singsing~singsing, but suru~singsing ' similar to a ring', cf. chapter IV. This means that in Bikol three subsequent identical syllables are accepted, whereas four identical syllables in a row are blocked. 69) dai-ng labot kun an sipon nag-no~nónó. NEG-LK care if PB mucus BEG.AV-IMPFV~run.nose 'not minding if the nose was running.'

[r-d: romdom]

70) nag-do'~do'do' pa ang aki child BEG.AV-IMPFV~breast still PB 'the child is still being breast-fed' Function Aspect marking for actions and events In the first instance, imperfective CV-reduplication is used to mark present and/or progressive actions (as in example (53)) as well as future actions (as in example (54)). In the first case, imperfective reduplication appears together with an affix for begun aspect (nag-, -in-, pig-, etc.), in the latter case no additional aspect marking appears (cf. III.2.5.1). As already mentioned in, aspect and voice marking may appear in any syntactic position and in combination with any root, independent of the meaning. It does not have any restrictions with respect to potential bases, a behavior that can only be expected for languages with no or only a weak distinction of word classes (cf. III.2.6). Unlike the morphemes for begun and not-begun aspects, imperfective reduplication is independent of voice marking and its form does not change in this respect. There is only one irregularity in the TAM-paradigm, namely the prefix ma:- for future tense/aspect in the actor voice instead of reduplication, e.g. Ma:-balik ako. 'I will return.' (cf. III.2.5.1). This morpheme has developed from CV- reduplication plus another affix. But there is a controversy about whether it goes back to the combination of the not-begun aspect prefix mag- plus the imperfective reduplication of root initial CV-, which is used in other languages, for example in Tagalog (e.g. mag-ta-trabaho ako 'I will work'), i.e. mag-CV~ > ma:-, or whether its origin is rather the combination of the old not-begun aspect infix -um-123 plus the imperfective reduplication of root initial CV-, i.e. *C-um-V:-. McFarland's (1974: 194) 123 For the description of the historical -um- paradigm in Bikol and its loss see Lobel 2004.


Productive reduplication

argumentation is that ma:- is a result of magCV-, because of an alternation of CV- with just a lengthened vowel in colloquial speech. Fincke (2002: 180-191) however provides a new explanation based among other things on a quantitative analysis of his data, arguing that ma:must be a result of *C-um-V:-. Word forms with the imperfective aspect reduplication are frequently used in attributive function (cf. III.2.5.1). As all modifiers, they are linked to their heads by na or -ng, for example124 71) naka-hiling kami ki babae-ng BEG.AV.ABIL-see 1PL.EXCL ARG.UG woman-LK 'we could see a walking woman'

nag-la~lakaw BEG.AV-IMPFV~walk [merr_asuwang]

Imperfective reduplication is also used to express habitual action, as in the following example, where the speaker refers to her family's habit of preparing fried rice: 72) Dai

man kami nag-ba~bawang sa s{in}anglag. also 1PL.EXCL BEG.AV-IMPFV~garlic LOC {BEG.UG}fry 'We (usually) don't put garlic into the fried rice, either.' NEG


Or in (73), where the TAM-marked word form refers to the habit of the witch to rape/ bewitch beautiful people: 73) ... an asuwang. Pig-ku~ku'a an arog saimo-ng ma-gayón ... PB witch BEG.UG-IMPFV~take PB similar 2SG.LOC-LK ST-beauty [p_asuwang] '... the witch. A beautiful one like you is taken (habitually) ...' The future/intended modality can also be expressed by imperfective reduplication within past contexts, as is illustrated by the following examples: 74) nag-gibo siya nin mga batas na su~sunod-on kan mga tawo. BEG.AV-make 3SG.AF ARG PL law LK IMPFV~follow-UG ARG.SPEC PL human.being [ibalong] 'he established laws that would be followed by the people.' If the repetitive or continuative aspect of an action or an event is emphasized, the prefix para-125 is used instead of the imperfective CV-reduplication. For example 75) An dagat nag-para-bato saiya. PB sea 3SG.LOC BEG.UG-REPET-stone 'The sea threw stones towards her constantly.'

[enot na tawo]

124 See also examples (3) and (4). 125 para- is also used as a prefix for deriving terms for jobs from actions, e.g. luto 'cook' --> para-luto 'a cook', or terms for persons who habitually accomplish an action, e.g. basa 'read' --> para-basa 'reader, bookworm'.


Productive reduplication

76) Pig-para-bicol-an=ko. BEG.UG-REPET-Bikol-REFL=1SG 'I keep on speaking Bikol.'


Mintz (2004: 37-38) and Lobel and Tria (2000: 101-105) list the combination of para- and the CV-imperfective morphemes in their paradigms (for present repetitive: nag-pa~paraetc.). In my corpus, however, no example of such a combination could be found. My suggestion is that the reason it is not very frequent in spoken language is because the more general meaning of CV- is already contained in the more specific meaning of para-, and whether the speaker is referring to present or past events is usually because it is clear from the context. But this topic is open for further research. CV-imperfective reduplication can also appear with the stative prefix, when marked for begun aspect, i.e. Na-li~lipot ako. 'I am feeling cold' or Na-i~intindi-han=mo? 'Do you understand?' As there is no undergoer-voice for statives, future aspect in these cases can only be expressed by the prefix ma:-, e.g. ma:-pagal=ka! 'You will be exhausted!'. I.e. reduplication is not involved in this respect. Continuative aspect in nominalized words Earlier in this work it was already mentioned that it is sensible to name every kind of morphological operation “derivation” in certain languages of the Philippine type, among them Bikol (cf. III.2.5). It was also already mentioned that the affixes are not restricted to certain lexical classes, but that any lexeme can undergo almost any kind of derivation. There are several possibilities for “nominal derivations”126 (the term nominal is understood purely semantically, cf. also III.2.6) and many of them are achieved by voice-affixes. For example laba 'wash' --> lab-han 'laundry', turog 'sleep' --> turog-an 'bed'; kaon 'eat' --> kaon-on 'food', inom 'drink' --> inom-on 'beverage'. As with derivations for actions, with nominalized forms the -on suffix can still be interpreted as expressing undergoer-features, and -an as expressing locative features (cf. grammar chapter III.2.5.5). I.e. even if the derivations are lexicalized to different degrees and not built spontaneously, they still offer high transparency. The same holds true for CV-reduplication in “nominalized” forms. These usually denote “an object to which an action is done or will be done or a place where an action is done or will be done.” 126 For example ka- -an, pagka-, para-, pag-, etc. (for details cf. Lobel and Tria 2000: 33-42, Mintz 2004: 4952).


Productive reduplication

(Lobel and Tria 2000: 33). For example lakaw 'walk' --> la~lakaw-an 'way, path', la~lakawon 'distance'; taram 'speak' --> ta~taram-on 'word/language'; lubong 'bury' --> lu~lubng-an 'grave', birik 'inside out, turned around' --> bi~birik-an 'spindle for cotton', la'ag 'put' --> la~la'ag-an 'container', ta'o 'give' --> ta~ta'o-an 'receiver'. In a language like Bikol with a very weak distinction of word classes, like Bikol (cf. III.2.6), it is not very surprising that word forms like ta~taram-on can be translated on the one hand as 'word/language' and be used in referential function, and at the same time can be used predicatively and be translated as 'will speak'. However, what is noteworthy is that one and the same form can be a lexicalized word with idiosyncratic meaning, and a transparent, regularly derived word. But also in combination with other derivational affixes, CV-reduplication can be used to express continuative or habitual meaning. The prefixes pag- and ka- derive words with different semantics, depending on the semantics of the base. Either abstract objects, like kaon 'eat' --> pag-kaon 'meal', abstracts, like tubod 'believe' --> pag-tubod 'faith' or “nomina actionis” like abot 'arrive' --> pag-abot 'arrival' or layog 'fly' --> ka~layog '(the performance of) flying'. Sometimes, these prefixes are combined with CV-reduplication to express continuity or habituality. For example puli' 'return home' --> pag-puli' 'the return home'/pagpu~puli' 'the returning home' (cf. Mintz 2004: 49). In (77) the word pag-ga~gadan refers to the period in which the hero of the story is already dead and exists only as a corpse, which is certainly continuous. In (78) the word ka-la~layog refers to the continuous flying of the bird, the central character of the story. 77) pag-poon kaidto sa pag-ga~gadan NMLZ-start DEM.DIST.UG LOC NMLZ-IMPFV~corpse 'from the time on Rabot was dead'


Rabot, Rabot

78) Dahil ngani sa ka-la~layog=niya, na pahá-on man siya ... because EMPH LOC NMLZ-IMPFV~fly=3SG already thirst-INT also 3SG.AF 'Because of her constant flying, she was already very thirsty ...'


[enot na tawo]

In most cases the conjoint appearance of pag- and CV-reduplication is lexicalized, i.e. pagderivations appear either with or without CV-reduplication, while a deviation from the lexicalized form is then judged by the speakers as ungrammatical, or otherwise CV- causes a difference in meaning, as for example in pag-dusa 'suffering, misery' versus pag-du~dusa 98

Productive reduplication

'sorrow, crisis' (from dusa 'grieve, lament, suffer') (cf. Mintz 2004). In some derivations the imperfective meaning of CV- is not very transparent, but the forms must be considered as fully lexicalized. For example in pag-do~doble 'reduplication' (from doble 'double, twice') or the cited pag-du~dusa 'sorrow, crisis'. With ka-, imperfective reduplication seems to be less lexicalized, and therefore more regular and transparent. However, in my corpus the imperfective reduplication rarely appears with nominalizations, and during elicitations speakers do not produce such word forms. Asked for a grammaticality judgment of a nominalization with reduplication, they mostly accept it as “interpretable”, though “not commonly used”. Therefore I suggest is, that imperfective reduplication is of limited productivity in combination with nominalizing prefixes like pag-127 and ka-, while it is very productive with the voice-markers -on and -an. But because of their obviously related semantics, in my view there are no reasons to separate the “nominal derivations” by CV- + -an/-on (named so for example by Lobel and Tria 2000: 33-34) from other TAM- and voicemarked word forms. Even if they are lexicalized in some instances, their derivation remains fully transparent for the speakers, for example ka~kaon-an 'restaurant' (“a place for eating”) or la~lakaw-an 'way, path' (“a place for walking”). Naylor (1986: 182) states for Tagalog CV:- reduplication that its semantic core is aspectual imperfectivity. And she ascribes this property to all word classes. "... imperfective (or perfective) aspect is a property of verbs as well as nouns and adjectives, ..." (Naylor 1986: 179). Diachronic development of aspect systems in Central Philippine languages Lobel and Tria (2000: 357-359) mention that CV-reduplication for incomplete actions might have existed already in the Proto-Central Philippine stage (in free alternation with -a-128). Not much research on the diachrony of morphology of Austronesian languages has been done so far, and the few results that do exist differ from each other. Blust (2003) summarizes the results of several studies, from Brandstetter (1916) to Zeitoun and Huang (2000). None of these studies mentions CV-reduplication for imperfective or progressive or unfinished aspect for Proto-Austronesian. But Starosta (1995), Blust (1998), and Ross (1998) propose PAN Careduplication for imperfective verbs, or future tense and durative aspect respectively (cf. Blust 127 The nominalizing prefix pag- itself is extremely frequent and productive. 128 Hiligaynon for example only uses -a- for imperfective aspect (cf. Wolfenden 1971: 127-129).


Productive reduplication

2003: 440-441). Zorc (1977: 232, cited Reid 1992: 74-75) reconstructs CV-imperfective reduplication for Proto-Malayo-Polynesian (Proto-Extra-Formosan, in Reid's terms). Reid (1992: 83) assumes the origin of CV-reduplication for the Central-Philippine languages in CVC-reduplication, as it is preserved in the Northern Philippine language Ilokano for example (cf. IV. Reid (1992) provides a scenario of the development of the aspect system from the formal and morphosemantic point of view in some Philippine languages, focusing on Tagalog and Ilokano. As the synchronic aspect system of Bikol is comparable to the one of Tagalog, I assume the diachronic developments to be identical or at least very similar. Reid (1992) reconstructs an aspect system for the Proto-Extra Formosan (PEF) stage, which has had more similarities with the contemporary Ilokano aspect system than with the Tagalog system. I.e. the older aspectual distinction was the one between completed and non-completed actions, while the distinction between begun and not begun actions was introduced later (and is expressed by the infix -in- in the Central Philippine languages). I.e. for PEF, Reid assumes the reduplication only to appear only in word forms which denote present continuative actions ([-completed/+continuative]). In Reid's scenario, the distinction between [-continuative] and [+continuative] was later replaced by a distinction between [-begun] and [+begun], and reduplication became a marker for [-completed] actions (imperfective aspect) in Central Philippine languages (cf. Reid 1992: 81-86). Whatever the exact formal development of the imperfective CV-reduplication is, it is an interesting fact that there is no evidence in Bikol, and many other related languages, that this partial reduplication had its origin in full reduplication – a path that was argued by Niepokuj (1997) to be obligatory for all partial reduplications (for counterarguments to her hypothesis see Hurch and Mattes 2005). A development from PEF CVC- to CV-, as proposed by Reid (1992), would at least follow Niepokuj's path of reduction. However, having a PAN Ca- affix as the origin of the CV- reduplication, as proposed by Starosta (1995), Blust (1998), and Ross (1998), is a big challenge for diachronic theories of reduplication.

EXCURSUS IV: GRAMMATICALIZATION OF REDUPLICATION The two most frequently cited and influential works on the origin and the development of 100

Productive reduplication

reduplication are Niepokuj (1997) and Bybee et al. (1994). They focus on different aspects of the development, though they share the basic idea, namely that partial reduplication is always a result of grammaticalization of full reduplication. Niepokuj (1997)129 Niepokuj considers reduplication to be a special case of affixation and therefore assumes reduplication to develop in the same way as affixes do, i.e. via grammaticalization of former independent lexical items. She considers full reduplication as equivalent to the composition of two lexical items and proposes the following scheme for the grammaticalization of reduplication:

i. Stage 1:




ii. Stage 2:






iii. Stage 3:






iv. Stage 4:


Figure 8. Stages of grammaticalization (Niepokuj 1997: 63)

Stage 1 represents exact full reduplication. In stage 2 one of the two elements has already undergone some processes of formal reduction and so it is defined independently of the base (i.e. partial reduplication). In stage 3 the reduplicant has fully obtained the properties of an affix. At this stage the reduplicant can develop fixed segments. Stage 4 means a reduction to gemination. This development of stage 4 out of stage 3 is well documented for reduplication in Indo-European languages, but has received little attention for other languages.130 Niepokuj establishes her model with respect to the formal development of reduplication and explicitly excludes semantic developments due to a lack of data. This is covered however by the following authors: Bybee, Perkins and Pagliuca (1994) Bybee, Perkins and Pagliuca also regard the development of reduplication as parallel to that of affixes in general. They mention the problem that the development of grammatical morphemes from a fuller lexical source, which they generally assume, is difficult to apply to reduplication, since it is not possible to trace a reduplicative gram back to a single word or a 129 Niepokuj (1997) is discussed here before Bybee at al. (1994), because her dissertation was written already in 1991, but published later. 130 One exception is the very auspicious approach for Arabic by El Zarka (2005).


Productive reduplication

specific phrase. Therefore they suggest, as Niepokuj does, „the fullest, most explicit form of reduplication, total reduplication, to be the originating point for all reduplications, with the various types of partial reduplication as reductions and thus later developments from this fullest form.“ (Bybee et al. 1994: 166) The authors additionally provide a model on the semantic development which is related to the formal reduction of reduplicants. Based on a comparative and typologically balanced sample of sixteen languages they conclude that full reduplication in most cases expresses very specific meanings, which are bleached through the process of grammaticalization. Partial reduplication has more general meanings and a greater variety of uses. The meanings which are cross-linguistically typically expressed by partial reduplication are meanings which generally occur later in the evolutionary path (i.e. also with other affixes), cf. Bybee et al. 1994: 167). Based on their analysis they propose the following semantic path of grammaticalization of reduplication131:













Figure 9. Development path of reduplication (Bybee et al. 1994: 172)

Undoubtedly the development of reduplicants as it is described by Niepokuj (1997) and by Bybee et al. (1994) can be observed in many languages, both on a formal and functional level. It is also true that a formal reduction often goes also hand in hand with an increase in semantic abstraction (cf. also Kouwenberg and LaCharité 2001: 79). In this sense the notion of grammaticalization of reduplication is absolutely justified, and describes the origin and the development of many reduplication types very well. Nevertheless the grammaticalization model of reduplication is problematic in several respects and there are important reasons to doubt that this model explains exhaustively the origin of all types of reduplication. Partial reduplication can obviously arise independently of full reduplication, and can take different 131 Bybee et al. (1994) exclusively cite examples of “verbal” reduplication, but without being explicit about this choice. So, unfortunately, it is unclear how “nominal” types of reduplication fit into their model.


Productive reduplication

directions of development. Many instances of partial reduplication in very different languages cannot be explained as originating in full reduplication, among them Bikol CV- reduplication, which probably originates in Ca- reduplication (see above). Other examples are cited in Hurch (2003), Haugen (2005), Hurch and Mattes (2005), among others. From the cognitive point of view, these reductions are clearly different from the process of erosion of lexical units to affixes, which is assumed in the general grammaticalization model. The erosion process is explained to be caused by the greater memory strength that a certain phonological form obtains by its frequent use and consequently an increased predictability by the speaker and the hearer. Since the reduplicant has a different phonological shape each time it is used, this explanation is not at all appropriate for the reduction of the reduplication pattern. Another important difference between affixes and reduplicants with respect to grammaticalization lies in the iconic character of reduplication (cf. Excursus I). In the process of grammaticalization the original lexical item gradually looses its semantic transparency. This is the main reason for the unidirectionality of grammaticalization (cf. Haspelmath 1999, 2004). Following Keller's invisible-hand theory (1990), speakers have no conscious access to grammatical morphemes, as opposed to lexical ones, i.e. they can intentionally “manipulate” the lexicon but not the morphemes. But this is not fully correct for reduplication. Its iconic principle is never fully lost during the process of formal reduction and semantic bleaching. At any stage of development reduplication keeps its formal feature of being a copy of the base, however small the copied portion might be. This repetitive formal property always offers an iconic interpretation and so it easily enables reduplicants to develop, formally and semantically, in the direction opposite to the one postulated grammaticalization path (called "antigrammaticalization" by Haspelmath 2004: 27). But if we accept Haspelmath's claim that such „changes against the general direction of grammaticalization are extremely rare“ (Haspelmath 1999: 1043) (and in my opinion the arguments for the principle of the irreversibility of grammaticalization is very convincing) then there must be something wrong with Niepokuj's directionality. On the basis of the important differences between reduplication and affixation I claim that reduplication should not be seen as undergoing grammaticalization in exactly the same way as affixes do in general. As Hopper and Traugott (1993: 1) point out, origin and development of grammatical units, although related to each other, have to be treated separately. After the preceding 103

Productive reduplication

considerations this means that reduplication can arise in a language as full or as partial (origin), and than it can develop along the path described by Bybee et al. and Niepokuj, but also in the opposite direction (development). As Hurch (2003) points out, reduplicants which are synchronically equal must not automatically be assumed to have developed equally. Furthermore, as reduplication is such an iconic means, it must be assumed to arise in different languages independently from each other, and even its appearance in several languages of one family does not automatically point towards a common inheritance of the reduplicants, but can also be a product of convergence (cf. Blust 1998: 30).

4.1.3 CV-reduplication with numerals In Bikol, as in many languages, numerals behave somewhat differently from other lexemes with respect to morphological operations. They can be derived regularly by TAM-affixes, for example tulo 'three' --> pig-tulo 'divide by three' or lima 'five' nag-lima 'quintuple' (as in Pigtu~tulo=niya ang butong.132 'He is dividing the bamboo into three parts.'). Full reduplication with numerals has a distributive function (e.g. apat 'four' --> apat~apat 'four each', as in Apat~apat ang lapis sa kamot nindo. 'They hold four pencils each in their hands.'). (For more details cf. IV. But in addition to these generally applicable reduplications, there is another reduplication type which appears only with numerals: CV-reduplication which has limitative function, exclusively in combination with numerals, for example tulo 'three' --> tu~tulo 'exactly three'. It is used to express that there is exactly a certain number of objects, not more and not less. 79) Sa~saro-ng chupa133 ang bagas! chupa PB rice LIMIT~one-LK 'There is exactly one chupa of rice (left)!' 80) T{in}a'o-an kan ina su kada aki=niya tig-tu~tulo-ng dulce. {BEG.UG}give-UG ARG mother PB every child=3SG DISTR-LIMIT~three-LK candy 'The mother gives to each of her children exactly three candies.' /C1V1C1(C2)V1(V2)X/

/C1(C2)V1(V2)X/ numeral


132 BEG.UG-IMPFV~three=3SG PB bamboo 133 An old measuring unit for rice.


Productive reduplication


'exact x'

Figure 10. Rule for numeral limitation

Like full reduplication for distributive, CV-reduplication for limitation is productive with all numerals, regardless of whether they are native Bikol lexemes or borrowed from Spanish or English, e.g. ki~kinse 'only fifteen' or fa-five 'only five'. However, some speakers prefer the Bikol or the Spanish numerals with any derivation and do not accept English numerals with Bikol prefixes or reduplications. Spanish numerals, as most other Spanish loanwords, are not considered to be “foreign” by many speakers, whereas English loans are consciously perceived as such. However, numeral derivation seems to be the only category which has a slight constraint with respect to stratal conditions. Usually, for counting, the native Bikol numerals are used for numbers "one" (saro) to "ten" (sampulo), the Spanish system is used from "eleven" (onse) on. English is getting more and more common for counting and calculating in general. As is the case for imperfective CV-reduplication, numeral CVreduplication, reduces consonant clusters and diphthongs, i.e. te~twelf 'only twelf' or fa-five 'only five'. But usually, if English numerals are used, the whole phrase is expressed in English. In order to express “only X” (not less and not more), CV-reduplication is combined with the suffix -i (), for example du~duwa-he134 'only exactly two', e.g. 81) Du~duwa-he sana su na-ku'a=niya-ng dulce. PB BEG.ST-receive=3SG-LK candy LIMIT~two-LIMIT only 'She only received (exactly) two candies.'135 The origin of CV-distributive for numerals is not fully clear, however it is possible that it has developed from the PMP Ca-reduplication for “human” numerals. Besides PAN Careduplication for verbal meanings (cf. imperfective reduplication, IV.4.1.2) and for instrumental nouns, Blust (1998, 2003) also reconstructs a PAN Ca- prefix for counting humans. A formal change of this Ca- prefix to CV- is documented for various Austronesian languages (cf. Blust 1998:33), and in many Northern Luzon languages there exists CVreduplication for human pluralization (cf. Reid, to appear). Bikol, and other Central Philippine languages, do not have, at least synchronically, any specific grammatical category for counting humans. Thus, it is quite possible that a former numeral reduplication for 134 /h/-epenthesis: cf. III.2.1.3. 135 Lit.: It was only (exactly) two, her receiving of candy.


Productive reduplication

humans here took over another function, namely limitation. As Blust (1998: 33) points out, by evolving from Ca- into CV- reduplication, the pattern looses its specific markedness and can therefore more easily undergo functional changes. 4.1.4 Infixal -Vr-reduplication for plural actors Form A completely different type of partial reduplication, which is also very productive, is the infix -Vr-, denoting plural actors. The reduplicant consists of the first vowel of the root and the consonant /r/ and is inserted between the first consonant and the first vowel, e.g. bayle 'dance' --> b-ar-ayle 'dance (of several people)'. For example 82) ... kan nag-t{ur~}ubo na ini-ng ... when BEG.AV-{PL~}grow already DEM.PROX-LK '... when these taro plants grew ...'

mga PL

linsa136 ... Taro

83) Dakul-on-on na mga tawo an nag-{ar~}atender. many-INT-INT LK PL people PB BEG.AV-attend 'Very, very many people participated.'



When the infix is inserted into a word with initial consonant cluster, the infix follows the whole consonant cluster, as for example: trabaho 'work' --> nag-tr~ar-abaho kami 'we worked' or swimming 'swim' --> nag-sw~ir-imming kami 'we went swimming'. /C1(C2)V1(V2)X/


/C1(C2)V1rV1(V2)X/ 'x by plural actors'

Figure 11. Rule for plural actor

There is no consistency regarding the description of the structure of the infix among the various authors dealing with Bikol. Fincke (2002) analyzes the reduplicant as -Vr-, Mintz (1971, 2004) and Lobel and Tria (2000), however, describe an -rV-reduplicant, inserted after the first syllable. Indeed, there is no difference at the surface between an underlying -rV- or -Vr- (ba~ra-kal or b-ar~akal). But there are strong arguments for assuming an underlying -Vr-structure: First, the language has other infixes (-in-, -um-, -umin-), which all have a -VC136 Taro (linsa) is a vegetable plant with big leaves, typical for the Bikol region.


Productive reduplication

structure and are all inserted between the first consonant and the first vowel. Thus, for reasons of descriptive economy and cognitive plausibility, the -Vr-structure is clearly preferred. Second, if we assume the form -Vr-, the reduplicated vowel is reduced to the full vocalic element of a diphthong, for example bayle [baile] 'dance' --> b-ar~ayle [b-ar-aile] 'dance, pl.'. This is also the case with CV-reduplication (cf. IV. If we assumed a -rVinfix, however, we would have to suppose that this splits up the diphthong of the base, i.e. ba~ra-yle [ba-ra-ile]. Since the splitting of diphthongs is very improbable, because it is hard to explain it from either a phonological or from a cognitive point of view (the diphthong forms one syllabic constituent, i.e. the nucleus), these examples clearly support the description of the infix as -Vr-. And third, it can be gathered from comparative and diachronic data that -Vr- is the more appropriate description for the reduplicant: Dahl (1976) for example, reconstructs -VC- infixes for Proto-Austronesian in general. “... all infixes have the form -vc-, and as they are inserted after the initial C of the wordbase, they also add a supplementary syllable CV at the head of the structure” (Dahl 1976: 10). Blust (2003: 472) even lists a reconstructed *-ar- infix for Proto-Malayo-Polynesian with the function “plural actor”. This is exactly the function of the synchronically productive -Vr- infix in Bikol. The PMP *-ar- infix still exists in its original form in some other Malayo-Polynesian languages like Sundanese, where it generally marks plural. E.g. budak 'child' --> b-ar-udak 'children'; sare 'sleep' --> s-ar-are 'sleep. pl.', niis 'cool oneself' --> n-ar-iis 'cool oneself, pl.' (cf. Moravcsik 2000: 545, Robins 1959). Thus, it is very probable that *-ar- is the ancestor of the -Vr- reduplication. Much less clear than the formal description of the reduplicative infix is a plausible hypothesis of its origin and development.

EXCURSUS V: INFIXES137 Infixes (lat. īnfīgere 'to insert') are defined as affixes which are inserted into the stem. The infix is a universally rare affix type, though their occurrence is a typical feature of the Austronesian language family and an areal feature of Southwest Pacific and Southeast Asia (cf. Sapir 1921: 72, Ultan 1975: 172). The existence of infixes in a language always implies

137 Sapir (1921: 72): “ the very curious type of affixation known as “infixing””.


Productive reduplication

the existence of external affixes, as is expressed in Greenberg's Universal 26138. Consequently, they are assumed to develop mostly from other affixes. For Austronesian, infixes are even reconstructed even for the proto-language (cf. Ultan 1975: 163, also Blust 2003: 471-472). The reason why infixes are clearly disfavored cross-linguistically lies in their nature to produce discontinuous morphemes. Discontinuity is a highly complex pattern and much more difficult to process than continuous patterns. Furthermore the place of the infixes, i.e. inside another morpheme, is the least salient part of the word. “The question thus arises why discontinuous constructions exist at all (...) and, in particular, why infixation exists?” (Moravcsik 2000: 549). Interestingly, infixes from different languages have similar phonological properties: They are consonantal in most cases, and if they contain a vowel, it is often a result of a secondary development. In several examples of Ultan's cross-linguistic corpus the infixed vowels “harmonize with the root vowel” (Ultan 1975: 163). Such an interpretation would not consider Bikol -Vr- as reduplication with a fixed element, but as a fully specified infix whose vowel harmonizes with the base vowel. However, as there are no vowel harmony phenomena known in Bikol, and since reduplication is a very common means in the language, I adhere to the description of the infix as a reduplicant with fixed segmentism (cf. Excursus VII). Another cross-linguistic observation is that infixed consonants are predominantly nasals or liquids. One reason for this homogeneous appearance lies in their diachronic development: Very often metathesis is involved in the genesis of infixes (see below), and sonorants are typically more susceptible to metathesis than other consonants (cf. Ultan 1975: 162-164).139 Synchronically, infixes show revealing relations to external affixes, for example they often have alternative external positions or external alternates respectively, and they frequently cooccur with reduplication (for more detail cf. Moravcsik 2000: 547-549, Ultan 1975: 173-175). Together with other observations (i.e. the typical consonantal quality etc., see above), this suggests to the assumption that infixes usually develop from external affixes or reduplication, either by metathesis or by entrapment. Metathesis is a purely phonetic or phonological process and is governed by the Sonority Hierarchy Principle. Ultan (1975: 179) mentions the 138 “If a language has discontinuous affixes [infixes, circumfixes, etc.; author's note], it always has either prefixing or suffixing or both” (Greenberg 1963: 92). 139 The Bikol data fit this observation. The segmentally specified infixes are -in- and -um-, containing a nasal consonant. The reduplicative infix contains the fixed liquid consonant /r/.


Productive reduplication

example of the Indonesian former prefixes *um- and *in- which underwent metathesis with the base initial consonant due to the rigid constraints on consonant clusters. Entrapment on the other hand refers to the situation in which two or more affixes appear together within one word form and in which the outer affix becomes unproductive and is re-analyzed as part of the base. Consequently the inner affix is then interpreted as an infix (cf. Moravcsik 2000: 549).140 Of course the synchronic description of infixes also poses a challenge for all theoretical approaches. OT for example simply denies the existence of underlying infixes, but describes them as external affixes which shift into the base by highly ranked phonological constraints on prosodic structures like for example a NOCODA-constraint (cf. Blevins 1999: 384). In the case of Bikol -Vr- infixation OT would have to assume a prefix Vr-. This would violate the strong CV-constraint of Bikol (cf. III.2.2.1, Excursus III), causing a shift into the base. In cases of underlying prefixes that appear as infixes at the surface, the LEFTMOSTNESS-constraint („A prefix is located at the left edge of the word.“) is outranked by a prosodic constraint (cf. Blevins 1999: 397). For Bikol and related Philippine languages, a prefix underlying the existing infixes is a possible explanation, at least in a diachronic view (cf. IV. However, the assumption that infixes in all languages are surface forms of underlying external affixes is not maintainable. Blevins (1999) argues for the Austronesian language Leti (lti) that there are no phonological reasons to justify describing infixes as results of mutation of underlying prefixes. Also from the diachronic point of view there is evidence, for example from Indo-European languages, that infixes can come into existence in internal position, and need not be traced back to external affixes (cf. Moravcsik 2000: 549).

140 Haspelmath (1993) argues for the tendency of „externalization“ of inflectional affixes. He gives several examples of a diachronic change in affix order, all of which have the same direction: Internal inflection shifts towards the periphery, but never towards the center (Haspelmath 1993: 289). Even if this was only stated on the order of affixes, not taking into consideration the assumption that infixes arise from external affixes conflicts with the general idea of a tendency of affixes „outwards“.


Productive reduplication Diachronic development of Bikol -VrThe question with respect to -Vr- pluractional is whether this infix has its origin in a former prefix or whether it has developed from a primal *-ar- infix, as it is reconstructed for PAN. Some neighbor languages of Bikol have the same infixal reduplication for plural, for example Hiligaynon. There the infix -Vl- can have plural as well as intensifying meaning. Other Philippine languages, even as closely related as for example Tagalog, do not have infixed reduplication at all. Moravcsik mentions that a “special affixing pattern that has a tendency to give rise to infixes is reduplication” (Moravcsik 2000: 549). In the case of -Vr-reduplication the diachronic data suggest a development in the opposite direction, i.e. an infix gaving rise to reduplication. If we accept that -Vr- has indeed developed from *-ar- (reconstructed infix for plural actor *-ar, cf. Blust 2003), this genesis would be an instance of a development of reduplication on the basis of a fully specified affix.141 This scenario is usually not taken into consideration as the origin of reduplication. But, of course, reduplication with fixed segments is per definition a “mixed category” insofar as reduplication is combined with some specified material. Still, the possibility of an external origin of the infix should not be completely abandoned. Reid (to appear) describes a very interesting development of an infix for human noun plural in Balangaw (a Northern Luzon language). The infix has the form -an- or -in- and is probably the result of a “reanalysis of earlier C1V1C2V2-reduplicated forms with weak vowel deletion and resulting metathesis of glottal stop initial consonant clusters ...” (Reid to appear: 13). Other languages of Northern Luzon still have reduplicants for the plural of (human) nouns. For instance in Ilokano objects are pluralized by C1V1C2-reduplication, for example sabung 'flower' --> sab~sabung 'flowers'; balay 'house' --> bal~balay 'houses' etc. (Reid to appear: 4142). This most interesting comparable data with this respect is found in Agta. The Northern Luzon language has a plural actor prefix with the form Ca- (e.g. datang 'arrive' --> nagda~datang 'they all arrive', cf. Healey 1960: 9). If we assume the plural actor morpheme in Bikol and Agta to have the same origin, the two forms suggest that the Bikol infix might either have arisen from a former prefix and become an infix by metathesis, or that the Agta 141 Furthermore, another reduplication type of Bikol and other Austronesian languages, namely CV-, has probably developed from a reduplicant with fixed segmentism (Ca-). Cf. IV. 142 Reid cites examples from an Ilokano Grammar by Morice Vanoverbergh (1955).


Productive reduplication

prefix is an example of an infix that had been re-analyzed as a prefix (which is universally preferred over infixes). In any case the development of -Vr- defies the widespread assumption that every kind of partial reduplication must have its origin in full reduplication. The fact that there are good reasons to assume -Vr- to have developed from a fully specified affix -ar-, in addition and in an analogous fashion to CV- reduplication as originating in Ca- reduplication (cf. IV. and IV., leads to the conclusion that reduplication and affixation are not necessarily completely independent operations, but that they can also interact to a certain degree. Nevertheless, this observation is not an argument against a conceptual differentiation of affixation and reduplication (cf. II.4.3). Lobel and Tria (2000: 358) state that Bikol -rV-143 is in some sense a cognate of Bikol and Waray-Waray (war) para- and Tagalog pala- for repetitive actions. Unfortunately, they do not give any reference or further explanation for this statement. Formally, para- can easily be segmented into the two morphemes pa- (causative) and -Vr- (plural actor). For Hiligaynon it is indeed indicated by Wolfenden (1971) that the combination of causative pa- and plural -Vlresults in repetitive/habitual meaning. 84) Hiligaynon:

Nag-p{al~}a-luto' si BEG.AV-{PL~}CAUS-cook PB.PERS.AF 'Mother was always cooking.'

Nanay. mother (Wolfenden 1971: 154)

However, Wolfenden does not comment on this combination at all and therefore the example can only cautiously be taken in support of the hypothesis of the origin of repetitive para/pala-. Although I have no evidence of the causative pa- being indeed a (historical) component of the repetitive prefix (synchronically, such a composition is not at all transparent for the speakers), there is at least some evidence of a diachronic relationship between causative and iterative morphology in Austronesian as well as in non-Austronesian languages. „... in a number of languages (e.g., Leti, Taba144, Tetun, Kambera145) it [the causative prefix] also functions to alter the lexical semantics of the base verb and its argument(s) by adding notions such as 'increased intensity or duration' (referring to 143 -Vr- in my formulation (see above). 144 East Makian (mky) 145 (xbr)


Productive reduplication

the activity denoted by the verb), or 'incerased activity/agentivity' (referring to the first argument of the verb).“ (Klamer 2002: 944)146 An additional cue for a later and compositional origin of para- out of older affixes is that there is no reconstructed iterative affix for older language stages, whereas both causative *paand plural actor *-ar- are reconstructed already for PAN (cf. Blust 2003: 471). But these are only preliminary ideas. Whether pluractional -Vr- reduplication is indeed related to the prefix para- and if so, how the morphemes are exactly related to each other, is still unresolved and open for further research. r-l-metathesis As already described in the Grammar sketch (cf. III.2.1.4), -Vr- reduplication “triggers” /r/-/l/ metathesis, when infixed into an l-initial word. For example layog 'fly', if marked for plural actors, becomes r-al~ayog, not *l-ar~ayog. When the infix is inserted after a consonant cluster including /l/, metathesis also takes place, as can be seen from the example plano 'plan' --> nag-pr~al-ano kami 'we planned' (*nag-pl~ar-ano). Metathesis is often described as a sporadic and irregular phenomenon, belonging primarily to the field of language acquisition, speech errors and sound change (The „Metathesis Myth“, Hume 2004: 203). But Hume (2004, and in several other publications) points out that metathesis by all means occurs by all means productively and regularly in many languages. Bikol data support for this statement, as the l-r-metathesis is fully regular, i.e. an l-r-sequence is impossible and must be avoided with the aide of metathesis. It is not a reduplication specific phonological rule, but -Vr-infixation is the only morphological operation which produces the phonological conditions for the metathesis. The language seems to have a phonotactic constraint for /r/ to always precede /l/ (cf. III.2.1.4).147 When a word with initial /l/ is derived for plural actor by the reduplicative infix -Vr- the output of the rule would violate this constraint. Thus, metathesis re-establishes the required consonant sequence. A restriction pertaining to the sequence of these two phonemes at all distances within a word is exceptional, however, and very interesting for Austronesian and typological phonology. Phonological changes around the two liquids /r/ and /l/ in general and metathesis in particular are cross-linguistically very common. However, metathesis concerning two word 146 For non-Austronesian languages see Schrammel (in prep.). 147 This constraint does not concern borrowed words with l-r-sequence, for example laringhitis < Sp. laringitis / Engl. laryngitis.


Productive reduplication

initial non-adjacent consonants, as is the case in Bikol, is very rare. Hume summarizes that metathesis preferably takes place at the end of a word or root, and „overwhelmingly involves adjacent sounds“ (Hume 2004: 229). The reason for such a phonological process often lies in an ambiguity of the two affected sounds. And indeed this is the case in Bikol. /r/ and /l/ are phonetically very similar, as they are both liquids and share the place of articulation; they only differ in the manner of articulation. Such high similarity can cause difficulties for the hearer in discriminating between the sounds. In such cases the speaker and/or the hearer have to rely on the knowledge of the temporal ordering of sounds in the language (cf. Hume 2004: 216, Hume and Johnson 2001: 6). In the case of Bikol a phonotactic restriction for the ordering of the two liquids /r/ and /l/ might be such a strategy to improve language processing. Infixal plural reduplication and other infixes If one adopts a rule-based approach to morphological operations, then plural infixation precedes aspect infixation, i.e. -in- is infixed between the initial root consonant and the vowel of the plural reduplicant. For example 85) ... na na-raot na an saiya-ng t{in}{ar~}anom na ... LK BEG.ST-destroy already PB 3SG.GEN-LK {BEG.UG}{PL~}plant LK '... that his plantation of Taro was destroyed.' not: *t{ir~}{in}anom

linsa. Taro. [ibalong]

86) h{in}{ar~}atod=mi pa su mga escuela duman sa BU-extension {BEG.UG}{PL~}accompany=1PL.EXCL still PB.SPEC PL student DEM.DIST.LOC LOC BU-extension [valentine's day] 'we still accompanied the students there to the BU extension ...' 148 Interestingly, in terms of rule ordering, the -Vr- infixation seems to precede the -ininfixation, but succeed imperfective CV-reduplication which itself precedes -in- infixation (cf. IV. This leads to the following ordering of derivations: 1. CV- reduplication (imperfective aspect), 2. -Vr- reduplication (pluractional), 3. -in- infixation (begun aspect, undergoer voice). An example with all three affixes is already given in Such an assumption is problematic with respect to a rule based approach of word formation. Such ordering is neither cognitively plausible, nor can it be described elegantly.149 It might be more 148 BU is the common acronym for “Bicol University”. 149 Haspelmath (2002: 6-9) gives the goal of elegant as well as cognitively realistic description as a crucial one


Productive reduplication

realistic to assume that Bikol requires certain patterns for positions of morphemes in complex words which are “filled” by the available segmental material. Consider for example the pattern C1{in}V1~C1{V1r~}V1(C2)C3V2C4 for a begun and not completed action, performed by plural actors (e.g. bakal 'buy' --> b-in-a~b-ar-akal, cf. (68)). Within such a word based approach, the fact that the -in- infix precedes -Vr-, and that furthermore -in- is infixed into the imperfective reduplicant whereas -Vr- always is infixed in the base, fits perfectly with Haspelmath's hypothesis (1993) of the „externalization of inflectional affixes“. In word forms which contain both aspect and plural, aspect is always closer to the periphery than plural. Function In Bikol it is quite obvious that plural is derivational. There are arguments not to assuming inflection for languages such as Bikol at all (cf. III.4.2.5) and, related to that, there is no conjugational concord system anyway. So, as any other morphological operations, the plural formation by the reduplicated infix -Vr- is derivational, adding the meaning of plural actors, undergoers or experiencers to the basic meaning of the lexeme. Newman (1990: 53) chooses the term “pluractional” for this kind of optional derivational morphological marking. In Bikol, we have three different plural markers: Firstly mga, the general marker, for actions, things, states etc., which can be attached to any kind of word form or phrase (cf. III.2.5.3), secondly -Vr- for plural actors, undergoers or experiencers, which for logical reasons cannot pluralize “nominal word forms” (cf. IV.4.1.4), and thirdly full reduplication for plurality of the action or the property itself (cf. IV. As other pluralization processes (by the general plural marker mga or by full reduplication for plural action), -Vr- pluractional is always optional, because the expression of “plural” is not obligatory in Bikol. However, if the infix -Vr- is used in a word denoting an action, an event or a state of the proposition, it is obligatory to pluralize the correlating actors, undergoers or experiencers (by mga) if these are overtly expressed in the sentence150. This kind of -Vr-/mga “agreement” is optional the other way round: mga pluralized actors need not to be marked in the corresponding action word (cf. III.2.5.3). Usually, in Bikol, all plural markers are only used if the speaker wants to put a certain emphasis on the pluractionality. Usually -Vr- refers to a larger number, at least three. Newman (1990) points out that: for morphological research. 150 To simplify matters, it is referred to “plurality of actors” in the following, including experiencers and undergoers.


Productive reduplication

“The term 'plural' when applied to verbs tends to be used indiscriminately for two distinct categories, namely inflected plural verb forms required by a conjugational concord system (...) and derived plural verb stems denoting semantic plurality ... .” Newman (1990: 53) Because of the semantic difference of -Vr- pluractional and full pluractional, these two plural reduplication types can be combined with each other, and indeed, in my corpus there are many examples of co-occurrence of full and partial plural reduplication within one word. 87) Nag-ka~k{ar~}amas~kamas pag-sulay kan lada NMLZ-fix ARG.SPEC chili BEG.AV-IMPFV~{PL~}PL~hurry 'they are in a hurry with the fixing of the first chili,' 88) May nag-r{al~}ayog-layog diyan sa EXIST BEG.AV-{PL~}PL~fly DEM.MED.LOC LOC 'There was flying151 around there ... above me.'

... ...

na LK

enot, first [r-d: lada]

i-ta'as=ko. CV-above=1SG [valentine's day]

Whereas mga can be used as a plural marker with any root and any derived word in Bikol, independent of its basic meaning, -Vr- pluractional cannot occur with “nominal” roots and derivations (cf. Mattes 2005 and 2006b). But this is not a categorical difference between -Vrand mga, but only a logical consequence of the specific meaning of -Vr-. It clearly refers to plural actors, undergoers or experiencers of the expression marked in this way and never to the plurality of the referent itself of an expression denoting things or events. This becomes even clearer if we look at words where roots marked for plural actors by -Vr- are “nominalized”, for example pag-d-ir~ipan “gathering (of many people)” as in 89) ta ma:-atender kan saro-ng pag-d{ir~}ipan na pag-tomar-posesión152 because FUT.AV-attend ARG.SPEC one-LK NMLZ-{PL~}gather LK [merr_ asuwang] 'because we would attend one gathering of the assumption of office.' In this example it is very clear that not the gathering itself is pluralized (i.e. that it does not mean 'the gatherings'), because the word is further modified by the numeral saro 'one'. The -Vr-plural can only refer to the subjects of the gathering (i.e. 'one gathering of many people'). In order to pluralize the “nominalized” word itself, the general plural marker must be used (i.e. mga pag-dipan 'the gatherings'). 151 Intended: flying of many spirits. 152 tomar posesion is borrowed from Spanish: tomar 'take' and posesion 'possession'. It is lexicalized as an idiomatic expression with the meaning 'assume office' (cf. Mintz 2004).


Productive reduplication

McFarland (1974: 186) gives the following example for “plurality of the events”, though: 90) Inda kun na-m{ur~}ula-hon not.know if BEG.AV.GENR-{PL~}red-INT 'I don't know, if I blushed or got pale.'

ako o nan-r{ul~}ungsi.153 1SG.AF or BEG.AV.GENR-{PL~}pale

But the data both from my corpus and from elicitations do not support the view that -Vrreally expresses plurality of the events (which is rather expressed by full reduplication) 154. I suggest a different interpretation of the above example from McFarland, that the plural refers to the plurality of corporal spots. In any case, the translation clearly suggests a singular event. Another explanation for this example might be the use of -Vr- reduplication for expressing additional emphasis or intensity. Even if this is not the primary function, it appears sometimes for this purpose. As already mentioned above, Hiligaynon also has -Vl- reduplication not only for pluractional but also for intensity155. The marginal relevance of intensive function compared to pluractional meaning of the infix leads to the assumption that the intensive function is a new one, introduced by the speakers by means of transposition. This is not very surprising due to the close semantic relationship between intensity and plurality, both being subcategories of the very basic category “quantity” (further discussion cf. IV. Plural reduplication for ma-derived word forms -Vr-reduplication has a complementary variant for ma-derived word forms. In order to express plural possessors of a property or a state which is derived by ma-, the root initial CVis reduplicated. For example, Ma-hamot ang burak. 'The flower smells sweet.' --> Maha~hamot ang mga burak. 'The flowers smell sweet.' (*ma-h-~ar-amot/*m~ar-a-hamot). When the plural CV- is added to the ma- derived property, the accent on the root shifts one syllable to the left, if the basic accent is on the second syllable of the root. If the basic accent is on the first syllable of the root, it remains there, i.e. the reduplicant itself is never stressed, for example ma-gayón 'beautiful' --> ma-ga~gáyon; but ma-ímot 'stingy' --> ma-i~ímot.

153 In this example two instances of nasal assimilation can be observed, as described in nang- + pula -> namula, nang- + lungsi --> nanlungsi. Furthermore in nanrulungsi, the r-l-metathesis has taken place, as described in 154 For the semantic distinction of plurality of events and in events see Cusic (1981: 61). 155 In Hiligaynon -Vl- has the additional function of specifying the „normal use“ of an object (cf. Wolfenden 1971: 146). This function of -Vr- is absent in Bikol.


Productive reduplication

/maC1(C2)V1(V2)X/ state/property


/maC1V1C1(C2)V1(V2)X/ ↔


'plural x'

Figure 12. Rule for the plural of ma-derived word forms

Properties and states which are not derived by prefixation, like dakula 'big' are regularly pluralized by -Vr- (d-ar~akula), and also properties derived by the spatial dimension prefix ha- are infixed by -Vr-. Interestingly, the infixation does not take place after the first consonant of the root (as is the case with TAM-derived words), but after the first consonant of the prefix, i.e. h-ar~a-li'pot 'short, little (pl.)', but not *ha-l-ir~i'pot. The different pluralizations of properties with ma- and with ha- derivations lead to the assumption that the ha-prefix is more likely perceived as part of the root than the ma-prefix. This goes together with the observation that ma- is much more productive than ha-, which is limited to a small set of lexical items. 4.1.5 Summary Productive partial reduplication is a very important morphological procedure in Bikol grammar. Compared to productive full reduplication, the different partial reduplication types have highly specified functions. Except for imperfective aspect, which can be used to derive “nominal” word forms, the partial reduplications are fully morphosemantically transparent and do not cause any ambiguity. Because of the regularity of the partial reduplications, they might be easily classified as “inflectional reduplication”. However, reduplication, as any other morphological procedure, is not obligatory in a syntactic frame and, conversely, it does not underlie any syntactic restrictions. Like all other morphological operations it must be categorized as derivational. But if we allow for a continuous concept of inflection and derivation, partial reduplication in Bikol definitely has a more inflectional character than full reduplication (cf. for example Bossong 2001). This observation is a supporting argument for Bybee et al.'s hypothesis (1994) that partial reduplication typically has inflectional functions, whereas full reduplication more often carries derivative functions. Although there are several reconstructions of morphemes which plausibly could be assumed as ancestors of the current reduplicants, their diachronic origin and development is 117

Productive reduplication

not entirely resolved. But the least that could be said is, that historical and comparative data are sufficient to cast doubts on the widely unquestioned assumption that partial reduplication is always traced back to full reduplication.

EXCURSUS VI: VERBAL PLURALITY All meanings of productive reduplication and most meanings of lexical reduplication are nuances of the semantic category “plural”. As such, the reduplication system in Bikol as a whole can be considered highly iconic (cf. Excursus I). The analysis of Bikol reduplication shows that the large amount of terms such as for example “iterative”, “continuative”, “distributive”, “frequentative” must be reduced considerably in order to economically and adequately describe the phenomenon. A clear, reduced terminology is even more important for larger cross-linguistic studies, where highly differentiated semantic nuances cannot be correctly compared to each other. Furthermore, as will be demonstrated with Bikol full reduplication (cf. IV.4.2), it is also more adequate to reduce the terminology, i.e. it is sufficient to assign the meaning “plural” to reduplication, or even more general “change of quantity”, because the nuances are not a primary component of the reduplicant, but rather a result of the combination of various factors (cf. IV.4.2.5). In the following, the concept of “verbal plurality” will be summarized briefly. Plurality can be expressed lexically, but in most languages there are additional morphological strategies to mark plurality. In these cases, „plural“ usually belongs to the grammatical category of number. Usually, in descriptions and reference grammars plural is associated with nouns (or entities, semantically speaking). Nominal plural is very well studied and described, but only little is known so far on the nature of verbal plurality (semantically related to events), though it is stated to be very widespread in the world's languages, i.e. in many languages of North America, Africa, the Caucasus, South-East Asia and the Pacific (cf. Corbett 2000: 245). The question arises, whether verbal plural is a numeral category like nominal plural, or whether it has to be classified as a type of verbal aspect. Corbett in his comprehensive book on Number leaves this question open, but he argues that verbal plural should be seen as a numeral category, not least because he considers it worthwhile to draw


Productive reduplication

parallels between nominal and verbal plural (Corbett 2000: 247). This is also Dressler's main interest, arguing that the fundamental properties of noun and verb plurals are essentially the same (1968: 52-53). However, within the category of verbal plural, Corbett establishes a main distinction between participant plural and event plural and associates the former more with „number“ and the latter more with „aspect“ (Corbett 2000: 247). The general problem with discussing the issue of verbal plurality, as opposed to nominal plural, is that there is no standardized terminology. Furthermore, compared to nominal plural, verbal plural seems to be a much more complicate phenomenon to describe, because its functions can bear much more nuances, syntagmatically conditioned (cf. Dressler 1968: 54). One reason for this is that verbal plurality can refer not only to the “subjects” and “objects” of the verb, but also to the action or the event itself. Furthermore it can express plurality in the temporal and/or in the spatial sense. In addition, these different references of verbal plurality can often not be clearly separated from each other. For example, multiple actors can participate in one and the same action, but they can also individually perform multiple actions. And they can do so simultaneously, or serially, and/or at different places (cf. also Jensen 1952: 19). On the contrary, nominal plural usually clearly refers to the number of objects referred to by the noun (cf. Dolinina 1997: 487). Besides these semantic reasons, verbal plurality is much less described and understood, because it rarely exists in the well documented Indo-European languages. As mentioned already in the Grammar Sketch (cf. III.2.5.3), Bikol has no obligatory number marking. Personal pronouns have suppletive forms for singular and plural for each person - 1st, 2nd and 3rd (cf. table 8). Thus, whenever plurality is expressed156 it is optional and not grammatically determined. The absence of plural marking may not be automatically interpreted as singular. Words without any plural marking are in fact neutral with respect to number. Plural marking is a derivational means which can be applied in order to emphasize or disambiguate the plurality of entities, events or states. Such derivational semantic plurality is very widespread in the languages of the world (cf. for example the compilations of Jensen 1952, Dressler 1968), more than one might expect from the West-European languages, which predominantly have inflectional plural marking.157 156 By the general plural marker mga, by full reduplication, by CV-reduplication with numerals or ma- word forms or by -Vr- reduplication for plural actor. 157 Therefore number is considered to be a prototypical inflectional category (cf. Dressler 1989: 6).


Productive reduplication

Languages which make use of derivational morphology for verbal plurality on the contrary seem to have a tendency to lack obligatory plural marking for nominal categories (cf. Cusic 1981: 236). Verbal plurality is the focus of interest in illuminating studies on plurality like Dressler (1968), Cusic (1981) and Newman (1990). Of course the term “verbal plurality” is somewhat problematic with respect to languages which do not have a noun-verb distinction in the classical sense. However, since noun and verb plurals are essentially the same (see above), Dressler (1968) states correctly that the conclusions on the nature of plurality should be equally valid in languages with a clear noun-verb distinction and languages without such.158 As shown in previous studies (Mattes 2005 and 2006b, cf. also III.2.6), there is no clear distinction between nouns and verbs in Bikol. But on the lexical and on the derivational level, two semantically defined classes can be established, via semantic, syntactic and morphological criteria: „nominal words“ (referring to entities) and „verbal words“ (referring to events, states or properties). Therefore with respect to Bikol the term verbal plurality is used when referring to the plural of verbal words, but it must not be understood in the traditional sense of word classes. The issue of the semantics of verbal plurality inevitably touches on the notions of aspect and actionality (aktionsart). Whereas aspect usually belongs to the grammatical level of language and expresses the temporal view of the speaker on an event (mainly perfective vs. imperfective), actionality, referring to inherent temporal and spatial properties of the event, is rooted in the lexicon (cf. Bertinetto and Delfitto 2000: 190). Both dimensions play an important role with respect to plural reduplication in Bikol. The CV-reduplication type regularly marks imperfective aspect. It belongs to the tense-aspect paradigm and among all reduplication types this is clearly the one with the most inflectional character. This reduplication type, like all other reduplication types in Bikol, belongs to the category plural.

Imperfective aspect is generally semantically closely related to verbal

plurality: Firstly, imperfective aspect and repetitive (plural) actionality are often combined (as opposed to perfective aspect and repetitive, which usually do not occur together), and 158 “... selbst wenn sich Sprachen finden sollten, die eine ungefähre Scheidung von Nomen und Verbum nicht kennen, so würde dies die Ergebnisse unserer Untersuchung nicht stark modifizieren, da es gerade ihr Ziel ist, das Wesen der Pluralität bei Nomen und Verben als letztlich identisch zu erweisen” (Dressler 1968: 52).


Productive reduplication

secondly, imperfective aspect can be used to exclusively express repetitive or continuative meanings in many languages (cf. Dressler 1968: 60, 92). The other types of productive reduplication with their highly derivational character must be classified as actionality. In order to draw a clear distinction between inflectional and derivational plural verbs, Newman (1990:53) introduced the term „pluractional“ for the latter. The author mentions, that although these verbs are sometimes related to the plurality of nominal arguments connected with the verb, most often plurality refers to the action expressed by the word itself. Cusic (1981) uses the term „event plurality“, referring to the same concept as Newman. With respect to Bikol, I decided to use the term „pluractional“ for several types of plurality, which are optionally expressed with actions, events or states. The reader must be aware that in a language like Bikol, which has a highly flexible system with respect to the usage of lexemes on the morphological and on the syntactic levels, a term like „pluractional“ has to be understood in a much wider sense than in languages with a more rigid system (as for example in Chadic, the language family on which Newman's analysis is based). In most languages that mark pluractional (or event plurality) this category has various nuances. This can also be seen with the plural reduplication types and their various meanings in Bikol. A great variety of terms denoting verbal plurality can be found especially in the research on reduplication. This makes comparison of data quite difficult as almost every author uses different expressions for the underlying concepts. Dressler (1968) and Cusic (1981) carefully describe exactly what they mean by iterativity, repetition, continuation, etc. A central parameter for the notion of the verbal plural nuances is the concept of „phases“. Cusic highlights the main distinction between repeated and repetitive actions. Repeated actions are multiple, identical, and completed actions which are executed several times („event-external plurality, ..., distributable over multiple occasions“, Cusic 1981: 79). The discontinuative character of the event is in focus (cf. Cusic 1981: 96). A repetitive action on the other hand refers to one event consisting of multiple phases, but without focusing on the breaks between the phases („event-internal plurality“, Cusic 1981: 78, cf. also Dressler 1968: 63). Repetition is further divided into iterative actions, with a greater focus on the phases of the event, and continuative/durative actions, with less prominent phasality (Cusic 1981: 96, compare with Dressler159 1968: 62-65, 74-77). These are the concepts of the nuances of plural which 159 Dressler (1968) uses a slightly different system for classifying repetitive and iterative etc.


Productive reduplication

underlie the expressions used in this study for Bikol. In the literature we find an abundance of related terms, like frequentative, alternative, diversative, habitual etc. Of course these terms could also be applied in a more detailed description of the examples of Bikol. But for the purpose of economy in description I decided to reduce the terminology as much as possible down to a few basic notions. Especially in Bikol, context and semantic features of the base and other parameters which were neglected here (as for example mimic and gesture), play such an important role in communication, that a highly detailed analysis of words out of context and situation is insufficient anyway (cf. IV. Cusic (1981) also interprets diminution and intensity as nuances of verbal plurality, as does Dressler (1968). Both meanings are considered to be derived from the repetitive meaning. Intensive action is perceived as repetitive, but non-distributive performance of an action. Diminutive action on the other hand consists of lots of little actions, in tentative action the action itself is executed „little/not enough“ (Cusic 1981: 105-109). Kouwenberg and LaCharité (2005) explain diminutive as an extension of the distributive nuance of plurality; this will be demonstrated in much more detail in The notion of distributive focuses on various temporal and/or spatial loci (cf. Cusic 1981: 102). Cusic (1981) summarizes the meanings of event plurality under three main categories: Plural marking on actions can express the plurality of the events (i.e. „external plurality or iterativity in the sense of a series of perfective or imperfective actions.“), or the plurality in events (i.e. „internal plurality or imperfectivity in the sense of internal structures of the events“), or a combination of the two (Cusic 1981: 61). More precisely, the definitions identify the following cases: „1. Plurality is internal to an event if a single event on a single occasion consists of internal phases; 2. plurality is external to an event but internal to an occasion if a single bounded event (internally plural or not) is repeated on a single occasion; 3. plurality is external to event and occasion if a single bounded event is repeated on separate occasions.“ (Cusic 1981: 67) In some languages, these different concepts of verbal plurality are expressed by a single form. For example in Dyirbal (dbl), one suffix (-day) is used for plurality of the events, of the actors or of the objects (Dixon 1972: 249). In Bikol however, these different dimensions of plurality are marked by different reduplication types: Plurality of actors is marked by -Vr-, the plurality in events (the imperfective aspect) is marked by CV-reduplication, and the plurality of events


Productive reduplication

is expressed by full reduplication (repetitive, reversative, diminutive, ...). The plurality of events by full reduplication covers a wide range of nuances. The exact meaning turns out to be a result of the interaction of the semantics of the base and the plural meaning of reduplication (cf. IV. Bikol is not exceptional in this respect. Conceptually, the dimension of plural has an elaborate set of subcategories. Cusic (1981: 74) lists sixteen different plural meanings, Dressler (1968: 62-84) describes about thirty different nuances of verbal plurality. One of the reasons for the wide variation is of course that not only terminologically, but also conceptually the boundaries of actions (e.g. whether they are single or multiple) are often unclear. Dolinina (1997: 490) points out that this „uncertainty“ of the interpretation of verbal plurality can easily lead to other meanings. The author cites examples from Russian, where the verbal plural can have resultative and potential meaning. A similar extension/abstraction of meanings takes place in Bikol as well, where the plural by reduplication can have diminutive, and beyond that imitative meaning and marke politeness, etc. (cf. IV.4.2.2). Cusic furthermore provides a good elaboration of the role of „boundedness“ for verbal plurality, and the factors contributing to the concept of boundedness, as for example telicity or transitivity (cf. Cusic 1981: 41ff.). These are indeed the essential factors for the semantic classification of the meanings of full reduplication in Bikol in section

4.2 Full reduplication 4.2.1 Introduction This chapter is about productive full reduplication. Usually, full reduplication is the most simple reduplication type, at least formally. But not in Bikol: Full reduplication is constrained to one prosodic word shape, and has an alternate for all others, namely Curu-reduplication. Also from the functional point of view, full reduplication in Bikol turns out to be the most complex one. At first glance it has a very wide array of meanings, some of them contradicting each other, i.e. augmentative and diminutives. However, one aim of this chapter is to explain this polysemy of full reduplication in terms of Cognitive Semantics and to show that the various meanings can indeed be reduced to one very general function, namely “change of


Productive reduplication

quantity”. After the description of full reduplication and its variant Curu-, I argue for a distinction between two types of full reduplication (for plural and diminution on the one hand and for intensification on the other hand), which can appear as homonyms in many cases. Then some semantic models will be discussed which try to explain the relatedness of reduplication to plural, diminutive and intensive through one basic concept. It will be demonstrated how the semantics of the reduplication base interacts with the semantics of the morphological process in this case. Finally, I will argue for the non-arbitrariness of the realization of the different meanings by one form and show that the case of Bikol full reduplication supports the hypothesis that polysemy is a cognitively ideal strategy of grammar. 4.2.2 Form and meaning of full reduplication Full reduplication in Bikol means that the whole root is copied exactly, as for example layog 'fly' – mag-layog~layog 'fly around, aimlessly'. At first glance, full reduplication in Bikol appears to have a very wide range of possible interpretations. It is used to express intensity, plurality (i.e. iterativity, continuity, reversativity, distributivity, etc.), diminution (i.e. attenuation, similarity, imitation, etc.) and furthermore it has several lexicalized derivations160. Intensity means that the connotation “very much”, “intensely” or “exactly” is added to the meaning of the base, as for example in mahal 'expensive' – mahal~mahal 'very expensive'. Continuity refers to one single event or action which takes place over a longer time span (“event-internal plurality”, cf. Excursus VI). Iterativity means that an event or an action takes place several times. Cusic (1981: 57) categorizes iterativity as “multiple action” (“event-external plurality”, cf. Excursus VI). Dressler (1968: 62) stresses the discontinuous character of iterative actions. For example batok 'bark' – batok~batok 'bark again and again'. Reversativity denotes multiple events or single events that include a movement back to the starting point, as for example balik 'return' – balik~balik 'come and go', or bitin 'hang up' bitin-bitin 'swing back and forth'. Distributivity refers either to an action which takes place at several places, as lakaw 'walk' – lakaw~lakaw 'go here and there, stroll around' or to a numeral or temporal indication meaning “every” or “each”, as in bulan 'month' - bulan~bulan 'every month, monthly' or lima 'five' – lima~lima 'five each'. Attenuation refers to actions 160 For an analysis of lexical full reduplications see IV.3.4.


Productive reduplication

performed less carefully or less intensely than in their proper sense, for example samod 'weep' – samod~samod 'weep a little bit' or to properties or states which apply only partly, for example pagal 'tired' – pagal~pagal 'somewhat tired'. Closely related to this is the imitative meaning. It refers to everything that is either an imitation or a copy of what the base designates („something that looks like X“), or to any property or state that is just feigned („pretend to do/be X“), or to something that is used as a substitute for the object designated by the base. For example buta 'blind' – buta~buta 'feigned to be blind', agom 'husband' – agom~agom 'common-in-law husband/wife'. An imitative function of reduplication is frequently found in other languages. It seems to be especially common in Austronesian languages. Rosen (1977: 5-6) assumes a relationship with metaphoric functions of reduplication, as they are found in Indonesian. Very often, the listed meanings overlap or coincide. So, for example, iterativity or continuity can also be interpreted as intensity (“Wiederholung bedeutet Verstärkung”, Dressler 1968: 77), as in the following sentences: 91) Sa

San Carlos, Albay, ini an ha-hanap-hanap-on San Carlos Albay DEM.PROX PB IMPFV~RED~search-UG 'Here in San Carlos, Albay, (everyone is) continuously searching ...' LOC

92) Kaya' saro-ng ma-uran-uran na thus one-LK ST-RED-rain LK 'Thus, one rainy night, ...'


banggi, ... night, [ibalong]

Actions where continuity coincides with distributivity are also frequent. McFarland for example describes the function of “doubled roots” as “expressing a durative or iterative event with no particular goal in mind: lakaw~lakaw 'to walk around'” (McFarland 1974: 174). Dressler (1968: 72) uses the term “dispersive” (as a subcategory of distributive) for these actions or events which take place “everywhere/here and there”. It has been argued already that iterativity, continuity, reversativity and distributivity, as well as the diminutive nuances of full reduplication, can be subsumed under the term pluractional. Other authors use the term “verbal plurality”, but especially for languages like Bikol, which do not have clear verb and noun distinctions, “pluractional” is probably more appropriate. Full reduplication always refers to the plurality of the events themselves (cf. Cusic 1990: 61) and as such stands in opposition to the CV- imperfective, which refers to the 125

Productive reduplication

plurality in the events, and to -Vr- pluractional, which refers to the plurality of actors (more details cf. Excursus VI). As already mentioned in III.2.1.2, assimilation takes place with prefixes ending in -ng. As in the case of partial reduplication (cf. IV., with full reduplication this assimilation rule “overapplies” to the base initial consonant. For example nang- + bayad 'pay' --> namayad 'pay (to several persons)'; namayad + RED --> namayad-mayad 'pay (open bills) to many people'. 93) Na-mayad~mayad siya sa mga utang BEG.AV.GENR-INT~pay 3SG.AF LOC PL debt 'He repays his (many) debts.'

niya. 3SG

This overapplication takes place in order to preserve the identity of reduplicant and base. For more detailed discussion of the “Identity Constraint”, cf. Excursus II. Fully reduplicated forms are subject to all normal morphological and syntactic procedures. They can take any regular affixation and they can appear in any syntactic position. The following examples give an impression of the great variety of usage of full reduplication: 94) Nag-du~duman siya BEG.AV-IMPFV~DEM.DIST 3SG.AF ‘S/he goes there every month.’

bulan~bulan. PL~month

95) pigla-ng nag-s{ar~}abay~sabay na nag-tuga an mga bulkan suddenly-LK BEG.AV-{PL~}PL~together LK BEG.AV-errupt PB PL vulcano 'Suddenly all vulcanos erupted simultaneously ...' 96) Lima~lima an lapis=ko sa PL~five PB pencil=1SG LOC ‘I have five pencils in each hand.’

kamot. hand

97) Nag-batok~batok su ayam sa bilog na BEG.AV-PL~bark PB.SPEC dog LOC entire LK ‘The dog was barking the whole night long.' 98) ini

ha~hanap~hanap-on IMPFV~PL~look.for-UG '... here (they are) continuously searching' DEM.PROX


banggi. night

an PB

[r-d: kaon]


Productive reduplication

99) Siram kan pating o paging k{in}onot na nag-lana~lana delicious ARG.SPEC whale or stingray {BEG.UG}cook.spec LK BEG.AV-INT~oil 'The good taste of the very creamy whale or stingray, cooked in coconut milk ...' [r-d:kaon] 100) ... pag-aloy~aloy - bados na ... NMLZ-DIM~time - pregnant already 'only a short time – pregnant again' 101) Nag-tu~turog~turog=ka man BEG.AV-IMPFV~DIM~sleep=2SG.AF PART ‘You are just pretending to sleep!’


sana. just

Full reduplication is also used for marking politeness. This can be considered as an extension of the diminutive function, for example 102)Bagay~bagay su bado saimo. DIM~fit PB.SPEC dress 2SG.LOC 'This dress suits you well!' Interestingly, in Bikol it seems to have achieved the status of politeness marking, which as such is perceived to be independent from the diminutive meaning, as can be demonstrated with the following example from my corpus: The regular comparative is formed by the prefix mas-, i.e. mahal 'expensive' --> mas-mahal 'more expensive'. But uttered in a joke in association with 'Christ', the normal comparative would be absolutely inappropriate. Reduplication is used instead: 103) Mahal~mahal man su Pilipinas kesa DIM~expensive also PB.SPEC Philippines than 'So the Philippines were more expensive than Christ!'


Kristo Christ [pilipinas]

The phenomenon is known to exist in other languages, among them German and Italian, where requests are often mitigated via diminution (e.g. Fammi un piacer-ino? 'Could you do me a little favor?'), as described by Dressler and Merlini Barbaresi (1994: 150-152). In Snohomish the diminutive is often used by speakers to refer to his or her possessions in order to indicate humility (cf. Hess 1966: 351). Dressler and Merlini Barbaresi (1994) analyze the „morphopragmatics“ of diminution in several languages. They define a morphological rule to be relevant for morphopragmatics, „if it contains a pragmatic variable which cannot be suppressed in the description of the 127

Productive reduplication

meaning.“ (Dressler and Merlini Barbaresi 1994: 55) This is actually the case with Bikol diminutive reduplication used for politeness reasons. Morphopragmatics investigates the connotative meanings which are transported by morphological categories. As the authors demonstrate, in several languages the diminutive marking connotatively expresses a lack of precision or seriousness. Dressler and Merlini Barbaresi claim that with the diminutive marker, a [- serious] feature is added which is one of several means of lowering the responsibility of the speaker towards the speech act, and the commitment to its illocutionary force (cf. Dressler and Merlini Barbaresi 1994: 144). The connotation of [- serious] and [precise] leads to different pragmatic uses of the diminutive in different languages. In German for example it can mitigate requests (e.g. Kann ich ein Stück-chen Kuchen haben?) or it expresses a pejorative attitude (e.g. Austrian German Erzähl mir keine Gschicht-ln! 'Don't tell me lies (= stories-DIM)!'). In Bikol imprecision and reduced seriousness is a prerequisite for politeness. In the Philippine culture a directly advanced opinion is usually perceived as impolite. The diminutive marking allows the speaker to remain imprecise and consequently to leave the right of interpretation to the speaker. Spitzer expresses the politeness function of the „impressionistic diminutives“ (cf. Spitzer 1921: 202) in an almost poetic way: „Die Suffixe wirken wie Vorzeichen in der Musik, sie bestimmen die „Tonart“ der menschlichen Rede.” (Spitzer 1921: 201) The grammaticalized politeness system in Bikol is not as sophisticated as in many other Asian languages (i.e. Indonesian languages, Korean, Japanese). Nevertheless, there are many different strategies for politeness and the rules of their usage would be a very interesting field for further investigation. Besides the morphopragmatic use of diminutive reduplication I just mentioned, there are the specific politeness particles (e.g. po, tabi), social constraints on suppletive forms of „yes“ (opo, o'o, oho, iyo) and rules for the selection of respectful forms of address (madam, ati, tita, etc.). The origin of full reduplication is easily explained by its high iconicity. Blust (2003: 474) lists several functions of full reduplication for the PMP stage, namely 'do [verb] in numbers, do in quantity', 'random, indeterminate (with interrogatives)', 'plural of nouns', 'distributive, X at a time (X=numeral or temporal term)', 'simulative'. This list already contains almost all meanings that can be found in Bikol. 128

Productive reduplication Different accent patterns for different meanings? Mintz (1971: 149-150) mentions the variations in the stress pattern of exact full reduplication as corresponding to different meanings. He notes that intensive “repetition” 161 has initial stress in the second constituent, e.g. panó'~páno' 'very full', or gútom~gútom 'very hungry' (xx-Xx), whereas diminutive “repetition” has a parallel accent pattern, e.g. búta~búta 'feigning blindness', or bangóg~bangóg 'feigning deafness' (xX-xX or Xx-Xx). Lobel and Tria (2000: 90) only mention the imitative function of full reduplication anyway, without any reference to accent. Thus little information about any coherent prosodic differentiation of meanings is available from the literature. Therefore I tried to verify Mintz' statement with my data, and came to the conclusion that it has to be discarded. There are two accent patterns associated with full reduplication, namely rising-falling (xX-Xx) and rising-rising (xX-xX). Intuitively, it seems to be reasonable to expect that such a prosodic distinction would also distinguish meanings. And this is indeed the case in some other languages. For example in Sranan (srn), a Jamaican Creole, verbs are fully reduplicated in order to express diminutivepejorative-imperfective (DPI) as well as augmentative and iterative meaning. But “The three verbal types are accentually distinguished from each other, …” (Adamson and Smith 2003: 87). DPI-reduplication has parallel stress (férfi~férfi 'painted a bit'), augmentative reduplication has the main accent on the second constituent (ferfi~férfi 'painted too much'), and iterative reduplication has the main accent on the first constituent (férfi~ferfi 'painted several times'). For Bikol, however, no systematic prosodic distinction of meanings can be observed. Even in Mintz' own dictionary, numerous counterexamples to his classification (see above) are listed, e.g. bungóg 'deaf' --> bungóg-búngog 'to feign deafness' or labí 'overdone' -> labí-labí 'very much overdone' (Mintz and Del Rosario Britanico 1985). My corpus of spontaneous speech also contains contradictory data in this respect, for example saráy-saráy ‘keep carefully’ and pag-alóy-alóy162 ‘only a little later’. When confronted with two segmentally equal but differently stressed reduplications, the consultants usually do not associate different meanings. And vice versa, when confronted with one form, they often provide both intensive and diminutive or plural meanings. In elicitation tasks the speakers 161 The terms „repetition“ vs. „reduplication“ are rather burdened by the long discussion of the classification of repetition and reduplication in the current theories on reduplication (cf. Hurch (ed.) 2005). Here I am citing Mintz’ terminology which does not correspond to my own classification (cf. IV. 162 Prefixes in Bikol do not have any influence on the stress pattern of the base. Suffixes however cause a leftwards shift of stress (cf. III.2.2.3).


Productive reduplication

sometimes produce different prosodic forms to emphasize the difference between diminutive and intensive, but only in a direct contrast. In spontaneous speech, embedded in a whole phrase or sentence, the accent pattern seems to be neutralized. The initial result of my analysis of a long list of full reduplications (by elicitation, by spontaneous speech, by accounting for Mintz’ dictionary entries) is that in Bikol both stress patterns are documented with all meanings. Bikol does not, at least today, prosodically differentiate between different and even contradictory interpretations.163 “Full reduplication” or rather “repetition”? The distinction between repetition and reduplication is difficult to define and there is not much agreement between different scholars how to establish the border between the two phenomena. Therefore, Kouwenberg (ed., 2003) points out that a clearer distinction should be an aspiration of future research. However, she states that “the analysis will depend crucially on whether a distinction can be made between two identical words on one hand (repetition), a single word consisting of two identical parts on the other (reduplication).” (Kouwenberg (ed.) 2003: 2) Huttar and Huttar (1997) emphasize the affinity of repetition and reduplication. For them, the crucial difference lies in the phonological coherence of the expression. “By reduplication we mean the repeating of all or part of a word (more than a single segment), the result still being a phonological word, with its pitch and stress pattern. This is opposed to iteration, the repeating of a word, each word having its independent phonological (and semantic) qualities”. (Huttar and Huttar 1997: 395/396) The accent patterns in Bikol, as argued above, actually do not allow for a distinction between reduplication and repetition, as Mintz (1971) has proposed. The non-parallel (xx-Xx) accent patterns, which are possible for every lexeme, point towards a general classification as reduplication rather than repetition, at least in Huttar and Huttar's (1997) sense. Gil (2005) establishes some other useful criteria in his article on repetition and reduplication in Riau Indonesian:

163 I am very much aware of the fact that for a definite conclusion a thorough prosodic examination would be necessary. But a larger number of non-uniform patterns in Mintz and Del Rosario Britanico (1985) and in my own corpus at least allow for a strong hypothesis about this.


Productive reduplication





unit of output

greater than word

equal or smaller than word


communicative reinforcement

present or absent




iconic or absent

arbitrary or iconic


intonational domain of output

within one or more

within one intonation group

intonation groups 5

contiguity of copies

contiguous or disjoint



number of copies

two or more

usually two

Table 13: Criteria for the distinction between repetition and reduplication (Gil 2005: 33)

With respect to the “word” definition, the Bikol examples fall into the category of “reduplication”. It is always just the root that is copied, and further affixation applies to the double root. A derived word form is never repeated as a whole (cf. examples (95)-(101)). Also with respect to Gil's 6th criterion (see also Hurch 2002) they must be categorized as “reduplication”: To my knowledge, no case of more than two full copies is attested. Gil's criteria 3 through 5 are not applicable in Bikol. The interpretations are all iconic, albeit to different degrees (as will be argued below), and intonation can vary for each example (see above). Gil's second criterion, the “communicative reinforcement”, would classify intensive formation as “repetition”. Intensive word forms always express emphasis, often with the connotation of not only ‘more’ or ‘bigger’ than the base meaning but also ‘more than appropriate’. For example 104) Gabos-gabos niya k{in}ua su mga prutas. INT-all 3SG {BEG.UG} PB.SPEC PL fruit ‘S/he has taken the whole fruit!’ (and was not supposed to) 105) May tulo~tulo agom niya. EXIST INT-three husband 3PL ‘She has three husbands.’ However, in my opinion there is no reason why communicative reinforcement should not also be possible by reduplication, at least in Bikol.


Productive reduplication

If we try to weigh up the criteria for the distinction between reduplication and repetition for Bikol, established by various authors, it turns out that this distinction is often not clear. It is obvious that the concepts of reduplication and repetition have to be regarded as a continuum. Even if cross-linguistic criteria are taken into account it turns out to be necessary to define the borderline between repetition and reduplication individually for every language (cf. also Kouwenberg 2003 (ed.): 2). 4.2.3 The Curu- reduplicant: an alternative to full reduplication Prosodic conditions for the selection of the Curu-reduplicant Exact full reduplication is only allowed for bisyllabic C1V1.C2V2(C3)-bases where the two syllables are not identical.164 With bases consisting of three or more syllables, bisyllabic bases with initial165 or internal consonant sequences (i.e. C1C2VX, cf. examples (109), (111) or C1V1C2.C3V2(C4), cf. example (112) (113)), and bases consisting of two identical syllables (i.e. C1V1.C1V1, cf. examples (113), (114)) only the initial consonant is reduplicated and followed by the segment string /uru/ (labeled Curu-reduplication in the following)166. Bases which consist of only one syllable (cf. examples (105), (106))167 appear fully reduplicated as well as with Curu- reduplication. Both forms can be elicited and are accepted as grammatically correct. The consonant cluster pl- does not require necessarily the Curureduplicant (plano~plano/ puro~plano ‘desperate plan’). Borrowed fricatives are substituted by “native” consonants in the reduplicant: f- and v- by p- and b-, and ʃ- by s-, i.e. puru~freezer ' like a freezer', nag-sa-shopping 'doing shopping', nag-ba-valibol 'playing volleyball'.168 Thus, the most economic formulation, following the principle of the “Elsewhere Condition” (cf. Kiparsky 1973) of the rule is: 164 C1V1.C2V2(C3) has to be understood as having C1 and C2 (and C3) or V1 and V2, may be identical, e.g. babo 'play with a small child' or baga 'spleen', as long as the two syllables differ in at least one segment. 165 This is only possible with loanwords and foreign words, cf. III.2.2.1. 166 Alternatively the vowel is also reduplicated, only -r- remains as fixed segment in the prefix. At the present stage I am not able to identify whether this is a free alternation or an ongoing change. e.g. siri~singsíng or siro~singsíng (compare to (115)). 167 This is also the case only with loanwords or foreign words, cf. III.2.5. 168 Besides the strict constraint for CV-syllables in the (imperfective) reduplicant, this is a further example for „The Emergency of the Unmarked“ (cf. Excursus III) in Reduplication. The English phoneme /f/ is accepted by all speakers and pronounced by a part of the speakers, but only in the root. In the reduplicant this highly marked borrowed consonant is substituted by a native Bikol one.


Productive reduplication

1. exact full reduplication with C1V1.C2V2(C3)-bases 2. C1uru-reduplication elsewhere Examples for bases which select the Curu-reduplicant are: a) bases with three or more syllables: 106) balyo ‘change, transfer’ – buru~balyo ‘keep on changing’, 107) dakula' ‘big’ – duru~dakula' ‘somewhat big’, 108) karabasa ‘pumpkin’ – kuru~karabasa ‘small pumpkin’ b) bases with one syllable: 109) tren ‘train’ – turu~tren ‘toy train’, 110) ref ‘refrigerator’ – ref-ref / ruru~ref ‘something like a refrigerator’ c) bases with initial or internal consonant cluster: 111) prutas ‘fruit’ – puru~prutas ‘something like fruit’, 112) banggi ‘night’ – buru~banggi ‘every night’ d) bases with two identical syllables: 113) sopsop ‘suck’ – suru~sopsop ‘sucking continuously’, 114) rara 'poisonous' – ruru~rara 'somewhat poisonous' A few lexicalized exceptions to this rule can be found, however. For example lambi' 'extra fat found around the stomach and waist area' – lambi'~lambi' 'wattle, flesh hanging down from the throat or head of fowl', parte 'part, section' – parte~parte 'proportionally', or samba~samba 'praying mantis' (cf. Mintz 2004). The conditions for the selection of the Curu-reduplicant pertain especially to borrowed words (see above). This might lead to the impression that borrowed words or loanwords are treated differently from native Bikol words. But this is not the case. Bisyllabic loanwords without consonant cluster are regularly reduplicated (for example hapi~hapi 'enjoy' (without simplex, from English 'happy'), yelo~yelo 'yellowish' (from English 'yellow'), or relo~relo 'toy watch' from Spanish reloj 'watch'). Conversely, native Bikol words which meet the conditions for the Curu-reduplicant are equally excluded from exact full reduplication (for example hapros 'touch' - huru~hapros 'to touch gently and repeatedly'). 133

Productive reduplication

The selection of the Curu-alternate of full reduplication is a means of avoiding a sequence of certain equal structures which are perceived as “unpronounceable”. This is exactly what Hurch (2005: 719) defines as “preventive dissimilation”. “... a morphophonemically governed selection device of competing allomorphs or morphemes in word formation ... (...) ... the simple banning of specific affixes due to the clash of their sound structure in combination with a basis.” While reduplication per se is an operation which is based on the iconicity of parallel structures and furthermore mostly has a clear preference for the maintenance of identical forms (cf. “Identity Constraint”, Excursus I and II), it is the dissimilatory force that leads to the selection of the Curu-alternate under the mentioned conditions. Languages seem to differ considerably in their preferences with respect to identity or dissimilation. Interestingly, Cebuano, which has the same reduplicant (Culu-), has only one condition for the selection: a root with three or more syllables. For example padala 'send' – pulu~padala 'send once in a while' (Edrial Luzares 1979). Unlike in Bikol, consonant sequences or two identical syllables in the base are no constraint for exact reduplication, e.g. pista~pista 'hold a little feast', dasmag~dasmag 'collide repeatedly', or yabyab~yabyab 'shake out playfully' (Trosdal 1990). The Agta diminutive reduplication Cala- has no constraint at all. It may appear with any bases and does not stand in a (complementary) relationship with full reduplication (cf. Healey 1960: 6)169. Functions of Curu-reduplication Interestingly, the Curu- reduplicant, which is in complementary distribution with full reduplication, does not have all the functions that are covered by exact full reduplication. Like full reduplication, Curu- reduplication can be used to express numeral augmentation, i.e. plurality of events and states or numeral and temporal terms (iterative, distributive, continuative), and attenuation (diminution and imitation). But it cannot be used to express intensity. If a word needs to be intensified but meets the prosodic conditions that do not allow exact copying, an alternative to reduplication has to be chosen. Usually the highly productive intensifier-suffix -on is selected, e.g. malisioso 'malicious' – malisioso-hon 'very malicious' (*malisioso~malisioso), sopsop 'suck' – sopsop-on 'suck intensively' (*sopsop~sopsop), or, less often, the modifier maray 'good' is used, e.g. maray na dakula 'very big'. Curu- can only 169 In Healy's grammar of Agta (1960) no productive full reduplication is described, but at least five different types of partial reduplication are. This is noteworthy because it contradicts the universal hypothesis of full reduplication being an implication of partial reduplication (cf. Moravcsik 1978: 328, Rubino 2005a).


Productive reduplication

express diminution or pluralization, e.g. muru~malisioso 'somewhat malicious' or suru~sopsop 'sucking continuously'. Parallel to exact full reduplication, the precise meaning of the reduplicant can be different in different contexts or in different usages of the word. For example ang suru~singsing refers to something similar to a ring (referential use), but in the following sentence suru~singsing is used predicatively and has the meaning “wearing a ring which belongs to someone else”: 115)Suru~singsing=ko su singsing ni Harry. DIM~ring=1SG PB.SPEC RING ARG.PERS Harry 'I am wearing Harry's wedding ring.' In some related languages, for example Hiligaynon and Cebuano, the same reduplicant (Culu-) alternates with full reduplication. The origin of Curu-reduplication The origin and development of the Curu-reduplicant is unclear. As already mentioned, variations of the Curu-reduplicant exist in some neighbor languages of Bikol. Others, for example Tagalog, do not have such a reduplicant with fixed segments (cf. Excursus VII). Instead, Tagalog, or Ilokano for example, have CV(C).CV reduplication for bases which are longer than three syllables. Thus, one possible scenario might be the development of this Curu- or Culu- reduplicant from a former CV(C).CV- reduplication. Another explanation is that Curu- developed from a combination of CV- reduplication and the plural infix -Vr-. For example balyo 'dance' --> (*b{-ar~}a~balyo >) buru-balyo 'dance a little bit'. Interestingly, Mintz describes the Curu-reduplicant as CV- plus -rV-reduplication in his newly edited dictionary (2004), but unfortunately without any argumentation. However, data from Cebuano might support this hypothesis. As in Bikol, in Cebuano bases with three or more syllables may not be exactly reduplicated. But unlike Bikol, Cebuano has two functionally distinct variants: Cu- is used for marking the diminutive and Culu- for marking the distributive. For example bakikaw 'awkward' --> bu~bakikaw 'somewhat awkward' and padala 'send' – pulu~padala 'send once in a while'. The two examples show clearly that in the case of Culu- a plural meaning is added. This supports the analysis of Culu-/Curu- as going back to CV- + -Vr-/-Vl-. A further interesting observation is that Agta, in addition to Ca- plural actor and other reduplication types (for plural, intensive and diminutive) also has a diminutive reduplication 135

Productive reduplication

with fixed segments, i.e. Cala- (e.g. pirak 'money' --> pala~pirak 'a little money', cf. Healey 1960: 6). Bikol has the plural actor infix -Vr- and the diminutive reduplication Curu-. Cebuano and Hiligaynon have the plural actor infix -Vl- and the diminutive reduplication Culu-. Related languages that do not have a plural actor affix, neither have a diminutive reduplicant with fixed segments. This comparative data is noteworthy as it suggests a relationship between the plural actor morphemes and the plural/diminutive affixes. Synchronically, a variation of Curu-, namely CVru- or CVrV- can be observed in Bikol, e.g. siri~singsíng or siro~singsíng instead of suru-singsing (cf example (115)). The variation is observable not only from one speaker to the next, but it can also be found in utterances from one and the same speaker. Speakers usually judge all variants as correct. The variation might be a symptom of an ongoing change from Curu- to CVrV-, which can indicate a speaker's reanalysis of the Curu- reduplicant as CV- + -Vr-. But at the present stage this is pure speculation.170

EXCURSUS VII: REDUPLICATION WITH FIXED SEGMENTISM Curu- reduplication (as well as -Vr- reduplication, cf. IV.4.1.4) is a special type of reduplication, namely reduplication with “fixed segmentism”. This means that the “reduplicative morpheme contains segments that are invariant rather than copied.” (Alderete et al. 1999: 327). While many languages, even closely related ones such as Tagalog, do not possess this type of reduplication at all, Bikol has as many as two different reduplication types with a prespecified segmental string: Plural/diminutive Curureduplication, where the base initial consonant is reduplicated and then followed by /-uru/, and plural actor -Vr-, where the first vowel of the base is reduplicated and then followed by /r/.171 If we further take into account the unproductive reduplications of Bikol, there is a third type of fixed segmentism, i.e. the echo-words, which substitute either the initial or the second consonant of the fully reduplicated base, following certain alternation rules (cf. IV.3.5.1), for example rukay-dukay 'search or rummage through', garap-gasap 'rough', etc. In the literature on reduplication several theoretical approaches to these types can be found, and unfortunately the definitions of what is really meant by “fixed segmentism” are often unclear and confused. 170 I cannot attest a variation in usage patterns based on speakers' age. 171 A possible historical relationship between the two types is discussed in


Productive reduplication

The seminal work on this topic has been provided by Alderete et al. (1999) and commonly cited since then. Even though the paper has some argumentative weaknesses in some details (especially because the authors constrain themselves to argue only within OT-suppositions), it clarifies the terminological confusion found elsewhere. The authors draw a fundamental distinction between two different types of fixed segmentism which must be analyzed separately: the phonological type and the morphological type. The phonological type of reduplication with fixed segmentism contains a phonologically defined default segment, i.e. if the reduplicant which resulted in the exact copying of the base does not fulfill certain phonological constraints, a consonant or a vowel is substituted by a “default” segment (or segment string). This default segment renders the reduplicant less marked. Thus, in terms of OT (see Excursus III) reduplication with a default phoneme is a case of “The Emergence of the Unmarked” (TETU). In Yoruba (yor) /i/ with high tone generally is the default vowel (for example there is i-epenthesis in loan words in order to split up impossible consonant clusters as in girama < Engl. grammar). Thus in this language a Císyllable is less marked than any other CV-syllable. Deverbal nominalizing reduplication has the form Cí-: gbona 'to be warm/hot' --> gbí~gbóná 'warmth/heat', dára 'be good' --> dí~dára 'goodness' (Pulleyblank 1988: 265). This means that in Yoruba the preference for unmarked syllables is stronger than the preference for identical base and reduplicant. The reduplicant can be either described as C- which triggers an /i/-epenthesis (i.e. dára --> *d~dara --> dí~dára) or as CV- which triggers a vowel change. In any case, both explanations show that this type of reduplication with fixed segmentism is a purely phonological matter. Alderete et al. point out that “... the choice of fixed segments is determined, often contextually, by phonological markedness constraints that are part of UG.” (Alderete et al. 1999: 355). In contrast, the prespecified segment (string) of the second type of reduplication with fixed segmentism is determined morphologically and is therefore comparable to a normal affix. It just has reduplicated material in addition. Unlike the phonological type, these affixes may for example have “marked” phonological structures and they can have allomorphic alternatives. There are no phonological conditions which determine the value of the fixed segment. Taking into account these criteria of distinction, the types of reduplication with fixed segmentism in Bikol can be clearly categorized as morphological. The -Vr- infix as well as the Curu-reduplication do not have a phonologically unmarked structure (the default vowel 137

Productive reduplication

for infixes would be a nasal, and the default vowel of the language is probably /a/), furthermore Curu- has a regular complementary alternate, i.e. exact full reduplication, in cases of a prosodic constraint. Of course, reduplications with fixed segmentism, whether phonological or morphological, are a challenge for the OT approach to reduplication which describes reduplication in terms of Correspondence Constraint. Important constraints such as “Dependence”, “Anchoring” and “Identity” are violated (cf. Alderete et al. 1999, McCarthy and Prince 1995, Urbanczyk 2005, etc.). In Steriade's full-copying model (1988) (cf. II.4.2.3), Curu-reduplication must be analyzed as a result of full reduplication, followed by the reduction to CVCV- (the elimination of disallowed units) and then by a substitution of -VCV- through -uru-. Such an explanation for the processing of reduplicants with fixed segments seems very artificial, and as such is not a very attractive.

4.2.4 Two different types of full reduplication in Bikol In the previous section I showed that the Curu-variant of full reduplication can only be used for plural and diminutive but not for intensive. This functional difference between full reduplication and Curu-reduplication is a crucial motivation for assuming two underlying types of full reduplication. Whereas exact full copying can mean plural, diminution, and intensity, Curureduplication expresses only plurality and diminution, but not intensity. As Curu- is a suppletive alternate to full reduplication, it must be assumed that there is a distinction between one type of full reduplication which has an alternate Curu- and which expresses plurality and diminution on the one hand, and a second type of full reduplication for expressing intensity on the other hand. These two types appear as homonyms in the case of bisyllabic C1V1.C2V2(C3) roots, however. E.g. mahal~mahal 'somewhat expensive' or 'very expensive'172. Intensity in Bikol is most frequently marked by the very productive suffix -on173 (e.g. mahal-on), but can optionally be expressed by full reduplication if the root has a 172 The two contradictory interpretations are disambiguated by the context, cf. IV. 173 -on can be applied recursively, cf. III.2.2.3.


Productive reduplication

C1V1.C2V2(C3) structure. Plurality (including continuity) is usually marked by full or Curureduplication, but can alternatively be marked by the prefix para-174. Diminution can be alternatively marked by the prefix medyo- (medyo-malisioso 'somewhat malicious'), but imitation is expressed exclusively by full/Curu- reduplication, e.g. turog~turog 'pretend to sleep', or turu~tren 'toy train'. Summarizing, the two types of full reduplication in Bikol are: - type I (full/Curu-) for plural and diminution and - type II (full) for intensity (blocked for all words besides C1V1.C2V2(C3)) Interestingly, the grouping of these meanings is slightly different in related languages. In Ilokano for example, CVC-reduplication expresses plural and intensive (pusa ‘cat’ --> pus~pusa ‘cats’), whereas CV-reduplication (plus prefix) marks diminution (sangit ‘to cry’ -> agin-sa-sangit ‘pretend to cry’) (Hayes and Abad1989). In Tagalog full reduplication (plus suffix) is used to express the imitative (bulaklak ‘flower’ --> bulaklak~bulaklak-an ‘artificial flower’), whereas bisyllabic reduplication marks intensive (baliktad ‘upside down’ --> bali~baliktad ‘all topsy-turvy’), plural (hiwalay ‘separated’ --> hiwa~hiwalay ‘thoroughly scattered’), and diminutive (ma-talino ‘intelligent’ --> ma-tali~talino ‘rather intelligent’) (Schachter and Otanes 1972). In Hiligaynon full reduplication (with a Culu- alternate) expresses plural (balay ‘house’ --> balay~balay ‘every house’) and diminutive/imitative (lakat ‘walk’ --> lakat~lakat ‘walk slowly’, tawo ‘human being’ --> tulu~tawo ‘puppet’) (Wolfenden 1971). Agta does not have a clear categorization. CVC-reduplication marks plural (mag-saddu 'leak' --> mag-sad~saddu 'leak in many places') and intensive (ma-bangog 'fragrant' --> ma-bang~bangog 'very fragrant'), CV-reduplication also marks intensive (dagkal 'big' --> da~dagkal 'very big'), and additionally diminutive (mag-arut 'flow swiftly' -> mag-a~arut 'flow slowly')(Healey 1960). In Cebuano, full reduplication (with an alternating Culu- reduplication) has the same functions as in Bikol: diminutive/imitative (abhung ‘bad smelling’ --> abhung~abhung ‘somewhat bad smelling’, balay ‘house’ --> balay~balay ‘doll house’) and event plurality (bakak ‘lie’ --> bakak~bakak ‘lie repeatedly’), whereas full reduplication without the Culu- alternate expresses intensive (awas ‘overflow’ --> awas~awas ‘overflow abundantly’)(Buyne and Yap 1971). 174 For example Igwa-ng bayong nag-para-layog (EXIST-LK bird constantly/always flying ...' [enot na tawo].



'There was a bird which was

Productive reduplication


type I

type II


plurality, intensity

intensity, diminution/imitation


plurality, diminution/imitation



plurality, diminution/imitation



plural, diminution/imitation

intensity (?)175


plurality, intensity



plurality, intensity, imitation


Table 14: Categorizations of meanings with reduplication types in several Philippine languages

4.2.5 Specifications of the intended meanings One form – several meanings? The result of my examination is that we have to assume two types of full reduplication in Bikol: type I for plural and diminutive, with a suppletive alternate Curu-, and type II for intensive, which is only possible with C1V1.C2V2(C3) bases (cf. IV.4.2.3). My analysis showed furthermore that these two types do not differ in stress or other prosodic features. Thus, there are word forms in Bikol which can have several meanings, and even opposite ones, due to the fact that first first type I and type II are superficially identical. For example lugad ‘wound(ed)’ --> lugad~lugad ‘small wound’/‘heavily wounded’, laog ‘inside’ --> laog-laog ‘just inside’/‘completely inside’. Additionally, type I by itself covers a wide range of meaning nuances. Botha (1984) has observed an analogous coincidence of divergent meanings in one form in Afrikaans, as did Kouwenberg and LaCharité (2005) in Caribbean Creoles. Thus the grouping of augmentation and diminution with plurality is not an uncommon phenomenon. What seems to represent high semantic complexity in one form reflects, under more detailed examination, only slightly different realizations of one underlying concept. I would even go as far as proposing to use the term “quantity” for the all-embracing concept176. The function of reduplication in Bikol can then be simply and fully correctly labeled as “change of 175 Wolfenden (1971: 103-104) mentions full reduplication for intensity. However, the given examples all express plurality (i.e. repetition or distribution). 176 „... the concept of plurality of situations is a special case of the concept of quantity, one of the basic philosophic categories“ (Xrakovskij 1997: 3).


Productive reduplication

quantity” (in contrast to the quantity of the simplex form). As Xrakovskij (1997: 4) differentiated further, „quantity“ refers to plurality if it is countable, and it refers to magnitude if it is measurable. Kiyomi identifies the two contrary meanings of reduplication in MalayoPolynesian languages as “two manifestations of the same semantic principle of “a …er degree of …”, which is projected in the opposite directions” (Kiyomi 1995: 1151), although she considers plural reduplication to be an iconic process and diminutive reduplication as a non-iconic process. Based on my own data I came to view full reduplication in Bikol as having the very general function of modifying quantity. The direction of the modification is secondary: It can be an increase or a decrease.177 This leads to the question of how the hearer can correctly interpret the meaning intended by the speaker. Disambiguation of homonymous full reduplication type I and type II through the context When C1V1.C2V2(C3)-roots are used as the base for full reduplication type I or II, one superficially identical form can have opposite meanings. For example dangóg ‘hear’, where dangóg-dángog can mean ‘hear very clearly’, but also ‘hear by gossip’. The reduplication of (h)aloy 'time span', haloy~haloy, can either refer to a 'very long' or to a 'very short time span'. The reduplicated form of lúgad ‘wound(ed)’, lugad-lugad can refer to a ‘small wound’ or to ‘heavily wounded’; or laóg-laóg from laóg ‘inside’ can mean ‘just inside’ as well as ‘completely inside’. The question is thus, how do these forms get disambiguated in communication? Miller (1978) in his article on the semantic relations of words addresses the question of how it is possible that people are obviously able to quickly and accurately recognize which meaning a word expresses on a particular occasion. He concludes that the point of disambiguation of polysemes must be the context or the situation in which the words are uttered: “Most words can be accurately disambiguated on the basis of information in sentences in which they occur.” (Miller 1978: 98) Comparing sentences in Bikol in which the same reduplicated or repeated word expresses diminution in one case and intensity in the other case makes it clear that gradation particles (na, pa, man, lang, sana …) seem to play an important role in disambiguating the meaning, as well as the situational context, of course. 177 I am grateful to the audience of the ICAL 2006 for inspiring discussion of the topic, which deepened my insight into the nature of the semantics of full reduplication.


Productive reduplication

116) ... pag-aloy~aloy - bados na ... NMLZ-DIM~time - pregnant already 'only a short time – pregnant again'


117)Na-lingaw-an=mi palibasa alóy~alóy na naka-duman sa beach ST-forget-REFL=1PL.EXCL because INT-time LK BEG.AV.ABIL-DEM.DIST.LOC LOC beach 'We had forgotten, since it was a long time ago that we have been to the beach ...' [valentine's]

118)Lugad~lugad man sana! DIM~wound PART only 'It's only a small wound!' 119)Lugad~lugad siya na nag-uli. INT~wound 3SG already BEG.AV-return 'He returned heavily wounded! (from a fight) 120) Dumog~dumog pa ang ni-lab-han. DIM-wet still PB DERIV-wash-BEG.UG 'The washing (on the line) is still somewhat wet.' 121) Dumog~dumog na siya dahilan sa uran. INT-wet already 3SG.AF because LOC rain 'S/he is already soaking wet because of (walking in) the rain.' 122) may ba'go-ng dakop na pasayan ma-init~init pa EXIST new-LK catch LK shrimp ST-DIM~hot still [r-d: kaon] 'there are newly caught shrimps that are still a little warm' man (~ ‘also’), lang (~ ‘only’) and sana (~ ‘just’) trigger a diminutive interpretation, whereas na ‘already/immediateness marker’ and pa ‘still’ can trigger either. In these cases it is necessary to take into consideration semantics of the whole proposition. Jacobs (1983) attributes to the gradation particles the function of referring to scales on which entities of propositions are ordered. This means that with respect to a sentence like (118), speakers agree that there is a scale of injury severities, and that the wound to which the proposition refers to places rather low on that scale. This knowledge is transported by the gradation particle sana 'only'. Jacobs points out, that the assessment criterion for such scales is usually “probability” (cf. Jacobs 1983: 129). However, at least for sentences with full reduplication which always express a change of quantity, the criterion “variation of time” also plays an important role. Look for example at sentence (122): The particle pa 'still' refers to the scale of temperature which is decreasing continuously in time from its maximum, at the moment when the shrimp


Productive reduplication

were caught and killed to the moment they achieved their final temperature as dead fish. Or sentence (120), where the scale refers to the continuous drying of the washing, with a maximum of wetness at the moment it is hung on the line to the moment it is completely dry and can be taken off the line. Disambiguation of the meanings of the two full reduplication types with bisyllabic bases is supposedly additionally possible through emphatic intonation and/or by extra-linguistic factors like facial expression and gesture. Blake mentions this also for Tagalog: “… these reduplicated forms have sometimes an emphatic, sometimes a diminutive meaning, sometimes either according as they are pronounced with more or less emphasis.” (Blake 1925: 53). The Bikol consultants I work with do consider the context as well as emphasis, gesture and facial expression as very important factors in the interpretation of full reduplication. Unfortunately, due to the relatively little amount of full reduplication in the spontaneous speech corpus and lacking video material, I was not able to examine these criteria systematically. Differentiation of the meaning nuances of type I through the interaction of the semantics of the base and reduplication Another question is how the speakers disambiguate between the many meaning nuances which can be observed with type I, i.e. full reduplication and its alternate Curu-. As described above, full reduplication can mark plurality (i.e. distributivity, iterativity, continuity, ...) and diminution (imitation, attenuation, politeness). After careful consideration of the data, I came to the conclusion that the exact meaning of the reduplicated word is predominantly a result of the interaction of the semantics of the base and the semantics of the reduplicative procedure. This means that with one single semantic interpretation of full reduplication, namely “change of quantity”, all meanings that are indicated above can be described if we take into consideration that the function of reduplication interacts with other conceptual devices. These devices are provided primarily by the semantic components of the base. And reduplication acts upon the (semantic) features of the lexeme (cf. Cusic 1981: 135). What are these relevant semantic features? The primary distinction is drawn between lexemes referring to entities and lexemes referring to events178. On a purely lexical level we could label this distinction noun-verb-distinction (cf. III.2.6). Further, within this second 178 The term “is used for (...) states, processes, and actions” (Himmelmann 1991: 5).


Productive reduplication

group of lexemes, the critical feature „boundedness“ is taken as a criterion for drawing another boundary between subgroups of the „verbal lexemes“. Cusic (1981: 52) distinguishes four semantic verb classes, with respect to the criteria „telicity“, „process“, and „phases“, namely activities like run, swim, push, etc., accomplishments like paint a picture, read a book, etc., achievements like reach the summit, win the race, etc., and states (including nonactions, i.e. cognition, mental state, etc.) like know, want, love, etc. Here, activities and states are subsumed under the category „unbounded“, while accomplishments and achievements are embraced by the category „bounded“. The third lexeme group consists of numerals and time indications. 1) With “entities”179, full reduplication expresses diminution or imitation, for example harong 'house' --> harong~harong 'small house, shelter', tugang 'sibling' --> tugang~tugang 'stepbrother/stepsister', kwarta 'money' --> kuru~kwarta 'play money'. The plural of entities can only be marked by the plural marker mga, i.e mga harong 'houses', mga tugang 'siblings', etc. The distributive is marked with the prefix kada-180 'each', e.g. kada-harong 'every house' /C1V1C2V2(C3)/

/C1V1C2V2(C3)C1V1C2V2(C3)/ ↔





'small/imitated x' / C1uruC1X/

/C1X/ entity



'small/imitated x'

Figure 13. Rule for the diminutive of entities

2a) With unbounded181 “events”, full reduplication usually expresses diminution, i.e. attenuation or imitation, for example turog 'sleep --> turog~turog 'pretend to sleep', dumog 'wet' --> dumog~dumog 'wettish', basa 'read' --> basa~basa 'skim over', luto 'cook' --> 179 The term “is used to embrace lexical semantic classes (...) such as persons, things, institutions, etc” (Himmelmann 1991: 5). 180 Borrowed from Spanish cada-. 181 “Unbounded” refers to time and space, i.e. non-punctual and/or atelic events. (“stative verbs are unbounded”, Cusic 1981: 237). The terms „bounded – non-bounded“ were introduced first by Allen (1966).


Productive reduplication

luto~luto 'play cooking', etc.182 The plural of the unbounded events in the sense of repeated action can be marked by para- (cf. examples (75), (76)), the plural of actors is marked by -Vr- (cf. IV.4.1.4). /C1V1C2V2(C3)/ unbounded event/ state

/C1V1C2V2(C3)C1V1C2V2(C3)/ ↔


unbounded event/state

'a little bit of/pretend to x' / C1uruC1X/

/C1X/ unbounded event/ state


unbounded event/state ↔


'a little bit of/pretend to x'

Figure 14. Rule for the diminutive of unbounded events and states

2b) With bounded183 “events”, full reduplication usually expresses the plural type of repeated actions184 or continuity, for example balyo 'change' --> buru~balyo 'keep on changing'. It can express event-internal as well as event-external plurality but the plural refers always to the event itself. Plurality of actors is marked by -Vr-. Diminution is often logically excluded with bounded verbs, because they usually denote events which either take place or do not, but which usually cannot be performed to different degrees. For example tungab 'appear' --> tungab~tungab 'repeatedly appear and disappear': Something or somebody can either appear or not appear, but nothing in between (cf. also Botha 1984: 126). /C1V1C2V2(C3)/ bounded event


/C1V1C2V2(C3)C1V1C2V2(C3)/ ↔


bounded event

'repeated x'

182 This categorization refers to the basic semantics of the lexemes. In the text, basically „unbounded“ lexemes can refer to bounded events (for example English read (unbounded) vs. read a book (bounded)). (Cf. Cusic 1981: 43-51). 183 “Bounded” refers to punctual and/or telic events. (“Motion verbs which are directed, goal-oriented, punctual, or limited to a particular movement were said to be bounded ...”, Cusic 1981: 237). 184 In some cases the boundary between repetition and continuity is fuzzy. For example batok~batok 'bark again and again' ~ 'bark all the time', cf. example (97). Cusic (1981: 99) mentions that the distinctiveness of the repeated time units can be obscured and the repeated action is perceived as durative.


Productive reduplication

/ C1uruC1X/

/C1X/ bounded event

bounded event


'repeated x'

Figure 15. Rule for the plural of bounded events

3) With numerals and time indications, full reduplication expresses distributivity, for example sampulo 'ten' --> suru~sampulo 'ten each' or ta'on 'year' --> ta'on~ta'on 'every year'185. To express non-distributive plural with words like banggi 'night' or ta'on 'year', the general plural marker is used: mga banggi 'nights', mga ta'on 'years', etc. /C1V1C2V2(C3)/ numeral/time

/C1V1C2V2(C3)C1V1C2V2(C3)/ ↔



'x each/every x'


/ C1uruC1X/





'x each'

Figure 16. Rule for the plural of numerals

As already mentioned, bounded and unbounded are categories which are not clearly distinct for every lexeme. Therefore the categorization of the semantic interaction of reduplicant and base cannot predict unambiguously every instance of full reduplication. Additionally, unbounded events can in some contexts also be interpreted as continuative. Bybee et al. (1994: 165) mention the reason for this overlapping: Iteration applies better to punctual or telic predicates, but continuation can apply to both telic and atelic. Furthermore, this semantic analysis can not directly account for lexicalized meanings, which frequently occur (e.g. hálo' ‘mix’ --> haló'~hálo' refers to a specific dessert), and for non-referential usages such as politeness and respect (cf. (102) and (103)), which must be understood as pragmatic extensions of the diminutive meaning (cf. Jurafsky 1993, Dressler and Merlini Barbaresi 1994). Furthermore it must be kept in mind that this classification is based mainly on examples out of a context. The concrete meanings often change in different syntagmatic 185 Time indications can alternatively be prefixed with kada- for the same purpose: duro-dominggo = kadadominggo 'every sunday'.


Productive reduplication

environments. Cusic (1981) points out that semantic properties such as plurality must be located at „higher levels of structure such as verb phrase and sentence in pragmatic contexts having to do with the kinds of events or temporal objects which are represented.“ (Cusic 1981: 364) This classification is very similar to the analysis of reduplication in Afrikaans (Botha 1984) or in Salish (for Snohomish, cf. Hess 1966). Cusic (1981) demonstrates an analogous classification of reduplicative meanings in the Hokan language Diegueño (dih)186. Reid (to appear: 4) observes a comparable differentiation of meanings via the interaction with the base for CVC-reduplication in Northern Luzon languages. There, the CVC-reduplication marks primarily the continuative or imperfective aspect with non-completed verbs, and repetitive or distributive meaning with completed verbs. With "adjectival forms" (i.e. properties and states in my terminology) it can also mark the comparative. Whether the discussed meanings of reduplication can be universally associated with the semantic types of bases is an open question for further research.187 In the preceding I demonstrated, that reduplication cannot be associated with just one certain meaning which is simply added to the meaning of the base. It is rather the interaction of the lexical meaning of the base and the general reduplicative device to provoke a change of quantity with respect to the quantity inherent in the simplex lexeme. All things considered, it is not only the reduplicant which unidirectionally modifies the base, but also vice versa - the base contributes to the exact meaning of the reduplicated word form. Without a detailed analysis, Rosen assumes such an interaction for Indonesian: To understand, what a reduplicated word actually means, „The native speaker is obviously guided by clues which probably depend on the semantic and morphological nature of the specific words.“ (Rosen 1977: 9). Levinson (2000) reexamines Botha’s results within the framework of his theory of “generalized conversational implicatures”. He uses the term “I-implicatures” to describe the information about the general (i.e. default or stereotypical) meaning of a linguistic expression that is at the recipient’s disposal and that enables her or him to correctly interpret the utterance of the speaker who seeks for the most simple expression (in unmarked situations). 186 For the data cf. Langdon 1970: 98-101. 187 Being able to answer this question is one aim of the typological research project on reduplication in Graz.


Productive reduplication

Levinson’s “M-principle” however denotes that marked expressions are used for referring to abnormal or non-stereotypical situations. One typical morphological procedure that is used to satisfy this M-principle is reduplication. It gives the recipient the signal that the meaning of the expression deviates from the unmarked (normal) meaning, which would be communicated by the simplex form. Through the interaction of I- and M-inferences, Levinson describes how the recipient correctly identifies the exact meaning of the reduplicated expression, although reduplication itself carries a very general meaning (quantity, deviating from the normal one, in the case of Bikol): “... whether the inference goes in the intensity/increase direction or the attenuation/limited dimension seems to depend on the direction of I-inference from the unreduplicated form, the reduplication then picking up the complement.” (Levinson 2000: 153) Dressler in his work on verbal plurality points out already 1968, that verbal plurality is always ambiguous, i.e. it contains the possibility of expressing different nuances (Dressler 1968: 58). He assumes: „Diminutive und intensive Nuancen rühren wohl vom Doppelgesicht der Pluralität her, der Unterteilung und Vermehrung.“ (Dressler: 1968: 83) In Bikol this view can be further supported by examples which clearly show that diminution and plurality cannot always be separated from each other anyway, as for example: bagáy ‘things, stuff’ --> bagáy-bágay ‘odds and ends, bits and pieces’, gapó’ ‘stone, rock’ --> gapó’-gapó’ ‘small stones (in the rice)’. raba'~raba' was described by a consultant as "destroyed a little bit, but because it’s divided in many little portions". Equivalently, huru~hapros was described (by another consultant) as "to touch gently and repeatedly" as opposed to the unreduplicated form hapros which would mean "touch once and hard". Semantic and Cognitive explanations for polysemy of plural and diminutive Givón reminds us that polysemy and homonymy emerge because „The message is constantly changed by creative elaboration.“ (Givón 1991: 106). The creative use of reduplication (at least full reduplication) in Bikol was already demonstrated in the preceding chapter. Obviously, a combination of semantic meaning and context provides enough information to guarantee the correct interpretation of the utterance. Nevertheless, it remains an interesting issue that the concepts of intensity, plurality and diminution are encoded by almost or completely identical forms. The fact that this can be found not only in Bikol, but also in other related and non-related languages is strong evidence that the mentioned meanings are based 148

Productive reduplication

on one cognitive concept. For example Lionnet (1968) in his article on the “Intensive” in Tarahumara (tar) describes that the older categories plural (by reduplication) and frequentative (marked by i-) have merged together and now have the function of the semantically similar “intensive” (cited Miller 1985: 503). Regier (1998) has developed a semantic network model to explain the cross-linguistically typical variation in meanings of reduplication, without referring to formal differences. He assumes intensity to be a semantic extension of plurality. Kouwenberg and LaCharité (2005) on the other hand assume another path of semantic extension, where diminution is a logical subcategory of the feature “distributivity” of plural. Regier Regier (1998) explains the cross-linguistic frequency of co-occurrence of certain meanings with reduplication by the interaction of iconicity and semantic extension. He takes smallness (“baby”), repetition, and plurality to be the three basic iconic meanings of reduplication. From these basic meanings, all other semantic variants are attained through semantic extension (to e.g. attenuation, contempt, continuation, intensity, lack of control etc.). SOUND













spread, scatter

lack of control


Figure 17. Regier’s model: The interaction of iconicity and semantic extension (Regier 1998: 888)

This model refers to the typical variety of meanings of reduplication, and it provides possible relationships between the respective meanings. However, the author does not provide arguments for why the relationships should be assumed to exist in this way and what 149

Productive reduplication

assumptions lead to the three basic semantic categories “baby”, “repetition” and “plural”. In my view there is no reason to separate repetition from plural; but it would be sensible to establish one basic category “plural”, which has “subsemantics” like repetition, continuation, etc. The problem with this view is that the concepts are mostly very flexible, i.e. they do not exclude each other but are continuously ordered on a scale as for example iteration, when it is performed quickly and with short breaks (for example enduring barking can be perceived and described as iterated punctual actions, but also as one continuous action, if the time span between the single actions is very short). Plurality and intensity can likewise have fuzzy boundaries, for example in sentence (119), where 'heavily wounded' can refer either to one single very heavy injury or to many serious or not so serious wounds, where the high degree of injury results from the high number of wounds. Regier's model is obviously based on the assumption that reduplication for augmentative has another origin than that for diminutive. Rainer (1998: 284) claims that augmentative reduplication results from the high iconicity of the procedure, whereas diminutive reduplication has its origin in child language. Kouwenberg and LaCharité (2005) however propose another model for relation of reduplication with its typical meanings. Kouwenberg and LaCharité Kouwenberg and LaCharité (2005) developed another model, based on the assumption that diminution is an extension of plurality via the feature dispersive/discontinuous/ distributive/scattered (ma-lamaw~lamaw ‘occasionally oversleep’ (“here and there”)). Their evidence comes from the data on Creole languages, where only the full reduplication type exists. Extending iconic dispersive interpretations: Discontinuous occurrence > attenuation, tentativity > approximation, similarity Figure 18: Kouwenberg and LaCharité’s model (Kouwenberg and LaCharité 2005: 540)

Although without such cognitive explanation, the categorization of diminution as a semantic nuance of plurality has already been stated by Dressler (1968) and Cusic (1981) in their studies on verbal plurality (cf. also Excursus VI). Dressler (1968: 61-62) focuses on the discontinuative property of iterative actions which leads to an association with infrequence 150

Productive reduplication

and consequently with un-seriousness and diminution in general. Corbett (2000: 239) mentions the frequent use in many languages of the plural marker for the expression of „approximative“. He explains the fact that plural („approximative“) marking is used for reasons of politeness in some languages with the quality of „vagueness“ of plural as compared to singular. Further explanations of the extension from diminution to imitation and to pragmatic attenuation and politeness, also independent of the process of reduplication, exist for example in Jurafsky (1993) and Dressler and Merlini Barbaresi (1994). The classification in (figure 19) of Bikol full reduplications reflects the path illustrated by Kouwenberg and LaCharité rather than by Regier, because in the former diminution is considered to be a logical extension of plurality. In Bikol diminution and plurality are expressed by one and the same formal type. I do not consider the two models as being competing explanations, but rather as illustrations of two possible semantic realizations of one general concept. The appearance of different meanings in one form does not automatically presume a common evolution. Both paths of development are conceivable: Either the separate developments of distinct reduplication types for plurality, diminution, intensity, etc. which have merged into one, because of their conceptual similarity, or the origin of one single form for one of the mentioned meanings, with subsequent extension to other, related meanings. Independent of the semantic development of reduplication which subsumes meanings like plurality, intensity, diminution etc. in one form, it allows an interesting insight into the cognitive organization of the concept [quantity]. Klamer (2004: 304) and Robert (2005: 120) point out that such underspecification is always the result of semantic bleaching. The polysemy can also be a clue that full/Curureduplication is an instance of a reduplication in change „on the go“, following Klamer's argumenation (2004: 300). This means that the various meanings might become differentiated only later. But because of the many strategies which are available for disambiguation, there is actually no urgent need for the speakers to change the system. On the contrary, polysemy can be interpreted as an economic strategy of language.


Productive reduplication Polysemy as a strategy in optimization of language Haiman (1980), Leiss (1997, 2005), and Levinson (2000) all attempt to explain why polysemy (not only in reduplication) is not a flaw of language but can rather be analyzed as an economic strategy on the part of the speaker, allowing not to load the mental lexicon unnecessarily. Levinson (2000) assumes that morphemes with general semantics are preferred over maximal overt marking, because language seeks optimization by storing and giving as little information as required. It is the usual strategy of the recipient to enrich “what is said” by reshaping the range of possible states of affairs associated with “what is said” to a narrower range of possible states of affairs associated with “what is communicated” (cf. Levinson 2000: 116). As we have seen in the case of Bikol full reduplication, inferences of semantics of reduplication and of the base, and the context are sufficient for reaching the appropriate interpretation of full reduplication. Haiman (1980: 516) considers the mapping of different meanings onto one forms as itself iconic: “Similar morphological shape or syntactic behavior of (apparently disparate) categories may be an icon of their underlying semantic homogeneity”. He demands linguistics to demonstrate that “homonymy is only apparent, and that the superficially disparate categories mapped onto the same form are in fact semantically related …”. (Haiman 1980: 527-28) Of course others also demand explanations of striking homonymies. For example there is a ongoing discussion of the widespread syncretism of case-morphemes in Indo-European languages. Leiss (1997 and 2005) also focuses on the search for explanations of (apparent) homonymies188, however she goes one step further. She accounts for the high functional elasticity of morphemes by “underspecification”, i.e. the “overt marking of just one part of the compositional meaning of a functional category” (Leiss 2005: 1), and argues that this is motivated by reasons of cognitive economy (see also Croft 2003: 105). She even states, like Levinson (2000) that it is not at all exceptional, but rather the normal case, that functional morphemes are polysemous. Miller (1978: 98) already mentions that the most polysemous words are those words which are most easy produced and understood, and most frequently occur. I.e. polysemes do not create problems for the speakers, because they are rendered 188 Somewhat provokingly she argues for example for the semantic uniformity of the –s-morpheme in English (genitive, plural and 3rd person singular indicative) by the features [whole] an [parts of the whole] (Leiss 1997).


Productive reduplication

monosemes by their context. In Bikol it is the interaction of the semantics of the base and the reduplication and/or the context which have the function of the “monosemation”, as was discussed in the preceding sections. Leiss considers polysemy and syncretism not just normal but an optimal strategy of language, even more adequate than a strict one-to-one correspondence of form and meaning. This claim is in opposition with the old claim of the one-to-one correspondence of form and function (i.e. the non-arbitrariness of the sign, as already discussed by Aristotle and Plato). However, it is clear that virtually every language has homonymies and polysemies, although to different degrees. Newmeyer diagnoses language with the untenability of isomorphism „in its strongest form“. (Newmeyer 1992: 761). Croft (2003: 106) points out that polysemy is both, economically and iconically motivated. Economically, because there is only one form to be stored, and iconically, because this form refers to related meanings. And the „set of related meanings can be thought of as a connected region in a conceptual space mapping out linguistic meanings.“ (Croft 2003: 106) But, if it is possible to analyze the various functions of one form to one underlying general function, as I showed for full reduplication in Bikol, we arrive once again at a one-to-one correspondence anyway. Therefore I consider as very realistic the view that polysemy and syncretism are not an “accident” of language which speakers of the language can handle somehow, but that it is rather an advantageous conceptual organization principle. But of course one cannot neglect the fact that there are still arguments in favor of the existence of both tendencies in language, i.e. the formal differentiation of single semantic features (e.g. the almost perfect one-to-one correspondence in some agglutinating languages). In my opinion both strategies can play a role in language. Deciphering one of them does not implicate the absence of the other in a language. It merely explains the strategy which is chosen with respect to a certain part of grammar. I.e. underspecification is an economic option that languages can use (and do use frequently). For example Bikol full reduplication is underspecified and there are enough strategies available for disambiguation. In case a certain meaning needs to be emphasized, there are alternative operations, which provide a one-to-one correspondence (i.e. para- for iterative, medyo- for diminutive, maray or -on for intensive etc.189).

189 Only the imitative has no morphological alternative.


Productive reduplication

The cited authors only focus on functional morphemes, i.e. segmentally specified morphemes. However, my analysis of full reduplication in Bikol shows that the assumption of underspecification and disambiguation can explain that polysemy can also be highly motivated and not at all arbitrary in morphological processes such as reduplication. The procedure of copying the base indicates a change of the quantity associated with the simplex form. The direction of the change in quantity, i.e. less or more, unidimensional or multidimensional, etc. is not specified for the reduplicative process itself, but is suggested by the semantics of the base or finally by the context. The sycretism of diminutive and augmentative meaning in one form is not as exotic as one might assume looking at the Bikol examples. Spitzer (1921) provides a detailed examination of the diminutive and augmentative suffixes in the Romance languages, where in some instances the two diametrically opposed meanings can be expressed by one form. He also mentions Czech (cez), where the two meanings can be expressed by one word form, e.g. telisko 'small body' or 'very big body' (Spitzer 1921: 191). Spitzer explains the fact that a language tolerates such ambiguities by the possibility of disambiguation through the context: „Ich kann mir das nur so vorstellen, daß eine vermittelnde Bedeutung noch dem Sprechenden bewußt ist, von der aus je nach dem Satzzusammenhang die vom Sprecher gewünschte Spezialisierung vom Hörer mitverstanden wird.“ (Spitzer 1921: 191) If full reduplication in Bikol is compared to the various types of partial reduplication, it can be concluded that the latter is much more regular and much more specified with respect to their functions. I.e. the form of full reduplication is the most unmarked form out of all reduplications and at the same time has the most polysemous nature. This supports Miller's claim (1978) that the most polysemous elements have the most unmarked forms (see above). 4.2.6 Summary The aim of this chapter was, besides the description of the forms and functions of full reduplication in Bikol, to argue that the various meanings of this highly iconic morphological procedure can be reduced to a single very general one. The appropriate interpretation is then guaranteed via several components, for example, the semantics of the base and/or the context. Furthermore I showed that the polysemous reduplication is not at all exceptional and that there are different models to explain this phenomenon, bringing into the discussion the view


Productive reduplication

of polysemy as a cognitively plausible strategy exercised by language users, as is argued among others by Haiman and Levinson, and, more extremely, by Leiss. Crucially the result of this analysis is that Bikol does not conceptually distinguish between “more” (intensive, plural) on the one hand and “less” (diminutive, imitative) on the other, which might be expected and occurs in many other languages. Rather, Bikol distinguishes between “plurality” (and distributivity with its semantic extension of diminution) on the one hand and “intensity” on the other hand. This is important evidence for the hypothesis that the above meanings are very closely related, and that “more” and “less” are not necessarily decisive diametrically opposed categories (cf. IV.4.2.4).The result of my investigation of Bikol full reduplication can be illustrated as follows: [change of quantity] I: full/Curu-Reduplication

II: full

plural (bounded events/actions; numerals)



intensive (events/states)


diminutive (entities, unbounded events/states)




Figure 19. Semantic categorization of the two full reduplication types of Bikol

The Austronesian languages are rich in reduplication, bearing a great diversity of forms and meanings. A thorough analysis of their form and meaning correspondences can provide some 155

Productive reduplication

insight into general theories of reduplication as well as the theories of cognitive Semantics. In this respect Bikol, together with many other languages which encode the many aspects of quantity by the processes of repetition or reduplication, is able to deliver strong arguments for the cognitive conceptualization of quantity with its various specifications and extensions. Moreover it is highly interesting for the discussion of polysemy, which is an important issue for cognitive semantics in general. Due to the virtually ideal iconicity of these phenomena, the various and at the first glance opposite meanings of one morphological procedure, must not be regarded as a defect in the organization of the grammar, but rather a crucial organizational principle of grammar that is cognitively motivated (cf. IV. 4.3 Combinations of various reduplication types As already mentioned, the high productivity of reduplication in Bikol allows a very free combination not only of reduplication and other morphemes, but also of different reduplications types. Apart from logic, there are no restrictions for combining reduplications, as is demonstrated by the following examples. In particular they can be found abundantly in poems, where the coincidence of several reduplications is obviously used as a stylistic device. For illustration I give to verses of two rawit-dawits of my corpus. Instances of multiply combined reduplications can also be found easily in spontaneous speech (cf. examples (88)), but I suggest that they are more characteristic of poetic speech (cf. also examples (67) and (87)). 123) First verse of “Ki agom” (“The husband”), by Nino Saavedra Manaog Nag-tu~t{ur~}uro an su'lot=mo-ng BEG.AV-IMPFV-{PL~}drop PB wear=2SG-LK 'The shirt worn by you is dripping; basa~basa190 an buhok=mo; INT~wet PB hair=2SG your hair is soaking wet; floating

palda shirt

nag-bu~b{ur~}ulos BEG.AV-IMPFV~{PL~}flow


basa sa angog=mo, wet LOC eyebrow=2SG the wet off your brow, PB

190 basa 'wet' is a Tagalog term, but commonly used in Bikol instead of tumog.


Productive reduplication

Saka sa pisngi=mo; and LOC cheek=2SG and off your cheek; Nag-ta~takig~takig BEG.AV-IMPFV~PL-shiver Your lips are shivering;

an ngabil=mo; PB lip=2SG

Mari digdi nag-para-sain=ka, Ne? approach DEM.PROX.LOC BEG.AV-REPET-where=2SG.AF dear Come here - where have you been, Ne?' 124) Second verse of “Tigbak na balulang” (“The killed rooster”), by Abdon M. Balde, Jr. sa harong na sana LOC house already just 'just at home the gamecock

ini-ng DEM.PROX

sabungero gamecock

nag-ma~malayumay sa pag-re~retiro; BEG.AV-IMPFV~recuperate LOC NMLZ-IMPFV~retreat is recuperating in the retreat; pig-bu~buru~bulnot an sanribo-ng BEG.UG-IMPFV~PL~pull PB thousand-LK pulling thousands of grey hairs


pig~su~suru~sikwat, mga kuko-ng BEG.UG-IMPFV~PL~prod PL fingernail-LK prodding, the nails of the corpse;

gadan; corpse;

pig-pu~puru~pildit, pig~hu~huru~hapros BEG.UG-IMPFV~PL~squeeze BEG.UG-IMPFV~PL~caress squeezing, caressing tulang na nag-lumoy, ugat na bone already BEG.AV-soft root already the bones are already soft, the root is already droop.'


na-luyos. BEG.ST-droop

Productive reduplication

5 Summarizing comments on reduplication in Bikol Reduplication in Bikol is grammaticalized to a high degree (e.g. imperfective reduplication), although it does not have obligatory grammatical functions as in some other languages. For example in Indonesian, it is ungrammatical to use an unreduplicated word in some contexts (cf. Rosen 1977: 1), and in the Uto-Aztecan language Comanche (com), reduplication is used for plural agreement. For example nanīsuyakeku 'pretty,


--> na~nanīsuyakeku 'pretty,


(cf. Key 1965: 96). But the functions of productive reduplication in Bikol are very common crosslinguistically (cf. Key 1965, Moravcsik 1978, Pott 1862, Rubino 2005b): pluractional, diminution, distributive and limitation with numerals, imperfective aspect. All these functions can be subsumed under the category of “change of quantity”. Comparable data outside the Austronesian family is found for example in Arabic languages, in Salish languages, or even in an Indo-European language, Afrikaans. The detailed analysis of the Bikol reduplication system led to the following hypotheses: –

There is a universal tendency for reduplication to express certain semantic categories, and this is true for productive as well as for lexical reduplication. These categories belong to the field “quantity”. Furthermore there are some “typical” classes of lexical reduplication, such as animals, plants and sounds.

It is not only the morphological level of language which is highly structured, but also the lexicon. It is more plausible to regard the relation of morphology and lexicon as a continuum rather than a dichotomy.

Reduplication by itself can be semantically unspecified to a large extent. The specific meaning of the reduplicated word form is then drawn through other factors: The interaction of the semantic features of the base and the reduplicant, and in many cases, simply the context and the situation. Probably also other, extralinguistic features, such as facial and manual gestures, play an important role. This is open for further research.


Summarizing comments on reduplication in Bikol

The widespread disagreement with respect to the terminology together with the richness of nuances of verbal plurality, causes great difficulties especially in a typological perspective. In the description of a single language, it is certainly not problematic to define a useful set of terms. But if one tries to indicate the functions of reduplication universally, these individual sets of terms cause severe problems. For this reason I tried to establish a matrix of some basic features which contribute to the meanings of reduplication. Of course I am only able to do this with respect to Bikol and not in a universal perspective, but it should be seen as a first attempt to reduce the terminological confusion with the aim of the comparability of data. I suggest that one basic concept like the concept, like “quantity”, can be sufficient to describe a high percentage of reduplications, when other contributing factors are analyzed separately, as I did in section The precise combination of relevant features is probably different for different languages and cannot be universally predefined. In addition, I suggest that context and situation have much more influence, also in other languages, than is commonly perceived. From the diachronic point of view, reduplication in the Philippine languages is highly interesting and challenging. All partial reduplication types in Bikol have not in all likelihood developed from full reduplication, but probably originated in fully specified affixes. With two reduplication types, ongoing language change can be observed: The change of the base of imperfective CV-reduplication (cf. IV., and the strengthening of the reduplicative structure of the Curu-reduplicant (cf. IV.



CHAPTER V: CONCLUSIONS 1 Reference to the central questions of the research on reduplication Concluding the study of the reduplication types in Bikol I summarize my results with respect to the central questions of research on reduplication in general, which were listed in 3.2. It turned out that it is very difficult to speak about “reduplication” as a single independent entity. Reduplication is such a heterogeneous phenomenon, that there are no uniform answers to the fundamental questions. An adequate answer to the question of whether reduplication can and should be described as affixation or as something else is especially tricky for a language like Bikol, which has reduplicants with fixed segments (as -Vr-, and Curu-, but also CV- which probably has evolved from Ca-). First, affixes are formally invariable and above all they are defined independently of a base. But reduplicants have a variable phonological form which is directly dependent on the base to which it is attached. Second, the constituent features for the notion of grammaticalization of affixes are the phonological reduction, accompanied by semantic bleaching (cf. Hopper and Traugott 1993), as well as the transition of an independent (syntactic) constituent to a dependent (morphological) one. In the case of reduplication however, there does not exist a single specified phonological form which can be reduced to a shorter form and consequently there is no concrete content related form which can be semantically generalized. Therefore I opt for a conceptual distinction between reduplication on the one hand and affixation on the other hand, although it is obvious that both morphological phenomena stand in a close relationship, as can be seen from the many transitions from specified affixes to reduplications in Bikol (and vice versa in other languages). Depending on the theoretical background, reduplication can only be defined as a rule, which produces a certain “pattern”. Or, in case of an affixal concept of reduplication, a certain template has to be assumed (as for example by Marantz 1982, cf. For the formalization of reduplication rules I propose a word-based model as the most adequate to describe reduplication, which I applied for every type in 5.4. Morpheme-based rules, which require a serial ordering of the morphological operations cannot be applied adequately in


Reference to the central questions of the research on reduplication

many cases, as was argued in and other sections. The question of whether partial reduplication always emerges from former full reduplication has been addressed several times in the previous chapters. For Bikol, this question can be clearly answered in the negative. Quite on the contrary: All Bikol partial reduplication types probably have their origin in a fully specified affix (cf. IV., IV. and IV. Nevertheless, there are undoubtedly many partial reduplication types in the languages of the world which have there origin in full reduplication. But the case of Bikol and other Austronesian languages is a strong argument to abandon the universality of the development, which is still widely accepted. For the same reason I also doubt that the universal implication of full reduplication in systems with partial reduplication can be seriously maintained. For example Healy (1960) describes five productive partial reduplication types in Agta, but no full reduplication. And this is easily explainable if Agta reduplicants have their origin in the same PAN specified affixes as those of Bikol or Tagalog. The fact that not only do most of the languages possess reduplication, but that also the functions of this morphological means are strikingly similar in (almost) all languages, is explained by the high degree of iconicity which reduplication offers. The repetition of the base or a part of the base is predestinated to express plurality. However, the very general concept of plurality has many nuances and therefore plural marking can result in many different concrete meanings, and also undergo far reaching semantic extensions. Consequently, the degree of iconicity is reduced in some reduplication types, and additionally there also exist fully arbitrary reduplications. Both, the arbitrariness and the motivation of the linguistic sign are existent principles in every language. With respect to reduplication, I tried to demonstrate the high motivational level of this means of word formation in Bikol, even in cases where it is not obvious at first sight. Furthermore I argued that lexical reduplications can also be interpreted in terms of iconicity for the most part. Due to its potentially high iconicity, reduplication is very often associated with „primitivity“, at least by non-linguists (but not exclusively). I am convinced, however, that the complex reduplication system of Bikol can be taken as proof that reduplication is far from being a „simple“ structures. I also mentioned some languages with reduplication systems that are even more elaborate and/or opaque, for example Kwaza (cf. van der Voort 2003) or Mixe languages (cf. Rhodes 2004). 161

Reference to the central questions of the research on reduplication

The question of whether there is one general meaning which is universally associated with reduplication can certainly not be answered on the basis of a single language or language family. However, for reduplication in Bikol I argued for assuming a very general function of reduplication, namely a “change of quantity”. Especially for the polysemous full reduplication I tried to demonstrate how other factors, i.e. the semantic features of the base and the context, contribute to the disambiguation of an underspecified reduplicant. Although this analysis was made exclusively for Bikol, I assume that it is similarly viable in many other languages, although reduplication can have (and has) also functions other than the marking of plurality or, more generally, the marking of a change of quantity. On the basis of my in depth analysis of the superficially apparent wide spectrum of meanings of reduplication and the insight that superficially “opposite” meanings are just two aspects of one common concept, I strongly hypothesize that the association of reduplication with plurality is even stronger than commonly assumed, because it includes the many meanings that are commonly considered as “arbitrary” or “non-iconic”. The question of the origin of reduplication was addressed separately for every type. Some types, as -Vr- reduplication, and probably CV-reduplication, clearly go back to specific affixes, which have been reconstructed for earlier language stages. For the Curureduplication, the origin is very much unclear, and the comparison with corresponding reduplicants in related languages allow only for speculations. Nevertheless, the Bikol data in comparison with those of related languages strongly suggest that we cannot maintain the view that all reduplication types resulted from one and the same diachronic development, but that they can undergo very different paths of development. Finally, the question of productivity could not be treated sufficiently, because reliable statements on productivity can only be made on the basis on a detailed quantitative analysis of a large text corpus (cf. for example Rainer 1987, Bauer 2001). As this was not possible with the resources that I had at my disposal, I am only able to give a rough valuation on the basis of tests of the reduplication types with respect to their applicability to foreign words and pseudo-words, and to the transparency of the word forms to the speakers. I draw a general distinction between lexical reduplication and productive reduplication. The lexical reduplication is completely unproductive, of course, i.e. the word forms are stored in the lexicon of the speakers and are not generated by an active reduplication process. 162

Reference to the central questions of the research on reduplication

Historically some of them may have originated in productive rules, but others, especially the echo-words, may have arisen in the language as such. The reduplication rules which are described in chapter 5.4 are all productive. But CV-imperfective reduplication in nominal word forms has very limited productivity in the modern language, whereas CV-imperfective reduplication in the tense-aspect paradigm has no restrictions at all, and as such can be considered to be the most productive reduplication type of Bikol. For plural-actor reduplication -Vr- there is a constraint with respect to nominal word forms (cf. IV.4.1.4), but within its domain plural-actor reduplication is also completely productive. Generally there can be observed a slight tendency of substitutions of full reduplications by other, borrowed, affixes, and a loss of lexical reduplications, due to the influence of Spanish and English. But this statement is only based on some unsystematic observations and information from the speakers. Scrutinizing its validity is open to further research.

2 Further perspectives The study of the reduplication types of Bikol goes into detail with many respects, but gives only a superficial treatment of some other topics. Especially those which would require a method of data collection different from the one I applied. First of all, as already mentioned in, it would be very important to test the hypothesis on the neutralization of the accent patterns of the two full reduplication types, as I suggested in 5.4.2, by additional methods, i.e. a detailed phonetic analysis. The problem is that reliable data on this topic can only be gained from spontaneous speech recordings. In order to collect the necessary amount of material on reduplication, this corpus needs to be much larger than the one I have. But any prosodic study of Bikol, as of any other Philippine language, not only on reduplication, would be very welcome, because this domain is extremely underrepresented in the research on Philippine languages. Another subject which is almost untouched so far, is the L1-acquisition of a complex reduplication system such as the one in Bikol. It would be very interesting to know for example in which order the types are acquired by children, and, especially, how they handle the ample polysemy of reduplication. Such information would not only be enriching for the knowledge of language acquisition, but would also contribute to the understanding of the processing of reduplication in general. But of course such data can only be gathered from a 163

Further perspectives

long-term study, which I was not able to realize, but which is desirable for the future. Furthermore, a serious study on productivity is missing from the research on reduplication, but, as already mentioned, this also requires large corpora. And of course a larger-scale systematic cross-linguistic study on reduplication will provide more insights into the validity of established assumptions and hypotheses. Such a study is under way in the Graz Database on reduplication project. It will also permit to verify the suggestions which I tentatively made in this study, on the basis of the knowledge of the reduplication types of one Austronesian language and a superficial comparison with those of related languages.



REFERENCES Abraham, Werner. 2005. “Intensity and diminution triggered by reduplicating morphology: Janus-faced iconicity”. In Hurch (ed.) 2005, 547-568. Adamson, Lilian and Norval Smith. 2003. "Productive derivational predicate reduplication in Sranan". In Kouwenberg (ed.) 2003, 83-92. Alderete, John, Jill Beckman, Laura Benua, Amalia Gnanadesikan, John McCarthy and Suzanne Urbanczyk. 1999. "Reduplication with Fixed Segmentism". In Linguistic Inquiry 30/3, 327-364. Allen, Robert L. 1966. The Verb System of Present-Day American English. The Hague, Paris: Mouton & Co. Anagbogu, Philip N. 1955. “The Semantics of Reduplication in Igbo”. In Journal of West African Languages 25: 43-52. Anderson, Stephen R. 1992. A-Morphous Morphology. Cambridge: Cambridge University Press. (Cambridge Studies in Lingusitics 62) Anshen, Frank and Mark Aronoff. 1981. “Morphological Productivity and Phonological Transparency”. In Canadian Journal of Linguistics 26/1, 63-72. Ansre, G. 1963. “Reduplication in Ewe”. In Journal of African Languages 2: 128-132. Aronoff, Mark. 1976. Word Formation in Generative Grammar. Cambridge: MIT Press. (Linguistic Inquiry Monographs) Atlas, Jay D. and Stephen C. Levinson. 1981. "It-clefts, informativeness, and logical form: Radical pragmatics (revised standard version)". In Peter Cole (ed.) Radical pragmatics. New York: Academic Press, 1-61. Baayen, Harald and Rochelle Lieber. 1991. “Productivity and English derivation: a corpus based study”. In Linguistics 29, 801-843. Bakker, Peter and Mikael Parkvall. 2005. "Reduplication in Pidgins and Creoles". In Hurch (ed.) 2005, 511-531. Barker, M.A.R. 1963. Klamath Dictionary. Berkeley, Los Angeles: University of California Press. (University of California Publications in Linguistics 31) Bauer, Laurie. 2001. Morphological Productivity. Cambridge: Cambridge University Press.



Becker, Thomas. 1990. Analogie und morphologische Theorie. München: Fink. (Studien zur Theoretischen Linguistik 11) Berg, Thomas. 1986. "The Monophonematic Status of Diphthongs Revisited". In Phonetica 43, 198-205. Bertinetto, Pier Marco and Denis Delfitto. 2000. "Aspect vs. Actionality: Why they should be kept apart". In Östen Dahl (ed.) Tense and Aspect in the Languages of Europe. Berlin, New York: Mouton de Gruyter, 189-225. Bezzenberger, Adalbert. 1908. “Zu den germanischen reduplizierten Präteriten”. In Zeitschrift für vergleichende Sprachforschung auf dem Gebiet der indogermanischen Sprachen 42: 383. Blake, Frank R. 1917. “Reduplication in Tagalog”. In American Journal of Philology 38 (152), 425-431. Blake, Frank R. 1925. A Grammar of the Tagalog Language, the chief native idiom of the Philippine Islands. New Haven: American Oriental Society. Blevins, Juliette. 1999. "Untangling Leti Infixation". In Oceanic Linguistics 38/2, 383-403. Bloomfield, Leonard. 1933. Language. London: Allen&Unwin. [reprinted] Blust, Robert. 1976. “Dempwolff's reduplicated monosyllables”. In Oceanic Linguistics 15, 107-130. Blust, Robert. 1991. "The greater Central Philippine Hypothesis". In Oceanic Linguistics 30/2, 73-129. Blust, Robert. 1998. "Ca-reduplication and Proto-Austronesian grammar". In Oceanic Linguistics 37, 29-64. Blust, Robert. 2001. "Historical morphology and the spirit world". In Bradshaw and L. Rehg (eds.) 2001, 15-73. Blust, Robert. 2003. "Three Notes on Early Austronesian Morphology". In Oceanic Linguistics 42/2, 438-478. Bolinger, Dwight. 1977. Meaning and Form. London, New York: Longman. (English Language Series 11) Booij, Geert, Christian Lehmann and Joachim Mugdan (eds.) 2000. Morphologie. Ein internationales Handbuch zur Flexion und Wortbildung 1. Halbband/Morphology: An International Handbook on Inflection and Word-Formation, Vol. 1. Berlin: Walter de Gruyter. 166


Bossong, Georg. 2001. "Ausdrucksmöglichkeiten für Grammatische Relationen". In Martin Haspelmath et al. (eds.) Language Typology and Language Universals: An International Handbook./Sprachtypologie und sprachliche Universalienforschung. Ein internationales Handbuch. vol. 1/1. Halbband. Berlin: de Gruyter, 657-668. Botha, Rudolf P. 1984. A Galilean analysis of Afrikaans reduplication. Stellenbosch Papers in Linguistics 13. Botha, Rudolf P. 1988. Form and Meaning in Word Formation: A Study of Afrikaans reduplication. Cambridge: Cambridge University Press. Bradshaw, John and Kenneth L. Rehg (eds.) 2001. Issues in Austronesian Morphology. A focusschrift for Byron W. Bender. Canberra: Pacific Linguistics, Brandstetter, Renward. 1916. An introduction to Indonesian linguistics. Translated by C.O. Blagden. London: The Royal Asiatic Society. Brandstetter, Renward. 1917. Die Reduplikation in den indianischen, indonesischen und indogermanischen Sprachen. Beilage zum Jahresbericht der Luzerner Kantonschule. Luzern. Broschart, Jürgen. 1987. Universalienprojekts 67.








Bruce, Les. 1984. The Alamblak Language of Papua New Guinea (East Sepik). Pacific Linguistics, C-81. Canberra: Australian National University. Brugmann, Karl. 1912/13. “Zu den reduplizierten Verbalbildungen des Indoiranischen”. In Indogermanische Forschungen: Zeitschrift für Indogermanistik und Allgemeine Sprachwissenschaft 31: 85-105. Buyne, Maria Victoria R. and Elsa Paula Yap. 1971. Cebuano Grammar Notes. Honolulu: University of Hawaii Press. (PALI. Language Texts: Philippines) Bybee, Joan, Revere D. Perkins and William Pagliuca. 1994. The Evolution of Grammar. Tense, Aspect and Modality in the Languages of the World. Chicago, London: University of Chicago Press. Carpio, Jose Maria Z. 2000. "An Kahalagahan kan Lenggwaheng Bikol". In Lobel and Tria 2000, 145-154. Chomsky, Noam and Morris Halle. 1968. The sound pattern of English. New York: Harper and Row. Comrie, Bernard. 1989. "The Phonology of Brazilian Portuguese Hypocoristic Kin Terms". In Ursula Klenk, Karl-Hermann Körner and Wolf Thümmel (eds.) Variatio Linguarum: Beiträge zu Sprachvergleich und Sprachentwicklung. Festschrift zum 60. Geburtstag von Gustav Ineichen. Wiesbaden: Steiner, 63-66. 167


Corbett, Greville G. 2000. Number. Cambridge, New York: Cambridge University Press. (Cambridge Textbooks in Linguistics) Croft, William. 2003. Typology and Universals. 2nd edition. Cambridge: Cambridge University Press (Cambridge Textbooks in Linguistics). Cusic, David D. 1981. Verbal plurality and aspect. [PhD dissertation]. Stanford University. Dahl, Otto Christian. 1976. Proto-Austronesian. 2nd rev. ed. Lund: Studentlitteratur, Curzon Press. (Scandinavian Institute of Asian Studies Monograph Series 15) Dahl, Östen. 1981. "On the Definition of the Telic-Atelic (Bounded-Unbounded) Distinction". In Philip Tedeschi and Annie Zaenen (eds.) Syntax and Semantics. Vol. 14. Tense and Aspect. New York, Academic Press, 79-90. Davis, Stuart. 1988. "On the Nature of Internal Reduplication". In Michael Hammond and Michael Noonan (eds.) Theoretical Morphology. Approaches in Modern Linguistics. San Diego: Academic Press, 305-323. Dempwolff, Otto. 1934. Vergleichende Lautlehre des austronesischen Wortschatzes 1. Induktiver Aufbau einer indonesischen Ursprache. Berlin: Reimer. (Zeitschrift für Eingeborenen-Sprachen, Beiheft 15). Dixon, Robert M.W. 1972. The Dyirbal language of North Queensland. London: Cambridge University Press. Dolinina, Inga B. 1997. "Theoretical Aspects of verbal plurality". In Xrakovskij (ed.) 1997, 483-495. Donegan, Patricia and David Stampe. 1979. "The study of natural phonology". In Daniel A. Dinnsen (ed.) Current Approaches to Phonological Theory. Bloomington: Indiana University Press, 126-173. Dressler, Wolfgang U. 1968. Studien zur verbalen Pluralität. Iterativum, Distributivum, Durativum, Intensivum in der allgemeinen Grammatik, im Lateinischen und Hethitischen. Graz, Wien, Köln: Hermann Böhlaus Nachf. (Österreichische Akademie der Wissenschaften. Sitzungsberichte 259) Dressler, Wolfgang U. 1984. "Explaining Natural Phonology". In Phonology Yearbook 1, 2951. Dressler, Wolfgang U. 1989. "Prototypical Differences between Inflection and Derivation". In Zeitschrift für Phonetik, Sprachwissenschaft und Kommunikationsforschung (ZPSK), Berlin 42/1, 3-10. Dressler, Wolfgang U., Willi Mayerthaler, Oswald Panagl and Wolfgang U. Wurzel. 1987. Leitmotifs in natural morphology. Amsterdam: Benjamins. 168


Dressler, Wolfgang U. and Lavinia Merlini Barbaresi. 1994. Morphopragmatics. Diminutives and intensifiers in Italian, German, and other languages. Berlin, New York: Mouton de Gruyter. (Trends in Linguistics Studies and Monographs 76) Dressler, Wolfgang and Mária Ladányi. 2000. “Productivity in Word Formation: A morphological approach”. In Acta Linguistica Hungarica 47, 103-144. Dressler, Wolfgang U., Katarzyna Dziubalska-Kołaczyk, Natalia Gagarina and Marianne Kilani-Schoch. 2005. "Reduplication in child language". In Hurch (ed.) 2005, 455-474. Edrial Luzares, Casilda. 1979. The morphology of selected Cebuano words. A case analysis. Canberra: Pacific Linguistics. El Zarka, Dina. 2005. "On the borderline of reduplication: Gemination and other consonant doubling in Arabic morphology". In Hurch (ed.) 2005, 369-394. El Zarka, Dina. 2006. Verbale Pluralität in vierkonsonantigen Verben des ägyptischen Arabisch. Ms. University of Graz. Fincke, Steven C. 2002. Sto.Domingo Bikol verbal portmanteau morphemes: Transitivity in conversation. [PhD dissertation]. University of California. Finer, Daniel L. 1986/87. "Reduplication and Verbal Morphology in Palauan". In The Linguistic Review 6, 99-130. Fischer, Olga, Muriel Norde and Harry Perridon. 2004. Up and down the Cline – The Nature of Grammaticalization. Amsterdam, Philadelphia: John Benjamins. (Typological Studies in Language 59) Ford, Alan and Rajendra Singh. 1991. “Propédeutique morphologique”. In Folia Linguistica 25/3-4: 549-575. French, Koleen Matsuda. 1988. Insights into Tagalog reduplication, infixation and stress from non-linear phonology. [M.A. thesis]. UMI, Ann Arbor. Gaeta, Livio and Davide Ricca. 2005. “Aspetti quantitativi della produttività morfologica”. In Tullio de Mauro and Isabella Chiari (eds.) Parole e numeri. Analisi quantitative dei fatti di lingua. Roma: Aracne, 107-124. Garrett, Andrew. 2001. "Reduplication and infixation in Yurok: Morphology, semantics, and diachrony". In International Journal of American Linguistics 67, 264-312. Gass, Susan. 1981. "Pragmatic and semantic constraints in Bikol relativization". In Linguistics 19, 1095-1106.



Gerhardt, Ludwig. 1984. "More on the verbal system of Zarek (Northern Nigeria)". In Afrika und Übersee 67, 11-30. Gil, David. 2000. "Syntactic Categories, Cross-Linguistic Variation and Universal Grammar". In Petra M. Vogel and Bernard Comrie (eds.) Empirical Approaches to Language Typology. Berlin, New York: Mouton de Gruyter, 173-216. Gil, David. 2005. “From repetition to reduplication”. In Hurch (ed.) 2005, 31-64. Givón, Talmy. 1991. "Isomorphism in the grammatical code: Cognitive and biological considerations". In Studies in Language 15/1, 85-114. Goldsmith, John A. 1976. Autosegmental phonology. [PhD dissertation] MIT. Gonda, J. 1940. "Opmerkingen over de toepassing der comparatieve methode op de indonesische talen, voornamelijk in verband met hun woordstructuur". In Bijdragen tot de taal-, land- en volkenkunde van Nederlandsch-Indië 99, 397-466. Gonda, J. 1950. "The functions of word duplication in Indonesian languages". In Lingua 2: 170-197. Goodenough, Ward H. 1963. "The long or double consonants of Trukese". In Proceedings of the Ninth Pacific Science Congress. Vol.3. Anthropology & Social Sciences. Bangkok, 77-80. Gordon, Raimond G., Jr. (ed.). 2005. Ethnologue: Languages of the World, Fifteenth edition. Dallas: SIL International. Online version: Graz Database on Reduplication: Greenberg, Joseph (ed.) 1963 [1966]. Universals of Language. 2nd ed. Cambridge, London: The M.I.T. Press. [First edition 1963] Greenberg, Joseph H. 1963 [1966]. "Some universals of grammar with particular reference to the order of meaningful elements". In Greenberg (ed.) 1966, 73-113. Grein, Christian. 1862. Ablaut, Reduplikation und secundäre Wurzeln der starken Verben im Deutschen. Cassel. Haas, Mary. 1942. “Types of Reduplication in Thai”. In Studies in Linguistics 1: 1-6. Haas, Mary. 1972. "The Expression of the Diminutive". In M.E. Smith (ed.) Studies in linguistics in honor of G.L.Trager. The Hague: Mouton, 148-152. (Janua linguarum, Series maior 52) Haeberlin, Herman. 1918. “Types of Reduplication in the Salish Dialects”. In International Journal of American Linguistics 1: 154-174. 170


Haiman, John. 1980. "The Iconicity of Grammar: Isomorphism and Motivation". In Language 56/3, 515-540. Haiman, John. 2000. "Iconicity". In Geert Booij, Christian Lehmann and Joachim Mugdan (eds.), 281-288. Harris, James and Morris Halle. 2005. “Unexpected Plural Inflections in Spanish: Reduplication and Metathesis”. In Linguistic Inquiry 36, 195-222. Harrison, Sheldon P. 1973. "Reduplication in Micronesian Languages". In University of Hawaii Working Papers in Linguistics 5, 57-92. Harrison, Sheldon P. and Salich Albert. 1977. Mokilese-English Dictionary. Honolulu: University of Hawaii Press. Haspelmath, Martin. 1993. "The diachronic externalization of inflection." In Linguistics 31/2, 279-309. Haspelmath, Martin. 1999. "Why is grammaticalization irreversible?" In Linguistics 37/6, 1043-1068. Haspelmath, Martin. 2002. Understanding Morphology. London: Arnold. (Understanding Language Series) Haspelmath, Martin. 2004. "On directionality in language change with particular reference to grammaticalization". In Fischer, Norde and Perridon (eds.) 2004, 17-44. Haspelmath, Martin. 2006. Frequency vs. iconicity in explaining grammatical asymmetries. Ms. MPI Leipzig. Haude, Katharina. 2006. A Grammar of Movima. Radboud Universiteit Nijmegen. Haugen, Jason. 2005. "Reduplicative allomorphy and language prehistory in Uto-Aztecan". In Hurch (ed.) 2005, 315-349. Hayes, Bruce and May Abad. 1989. "Reduplication and syllabification in Ilokano". In Lingua 77, 331-374. Healey, Phyllis. 1960. An Agta Grammar. Manila: Bureau of Printing. Hengeveld, Kees. 1992. Non-verbal predication. Theory, Typology, Diachrony. Berlin: Mouton de Gruyter. Hengeveld, Kees, Jan Rijkhoff and Anna Siewierska. 2004. "Parts of speech systems and word order". In Journal of Linguistics 40, 527-570.



Hestermann, Ferdinand. 1915. “Die Repetition in der Serersprache von Senegambien”. In Zeitschrift der Deutschen Morgenländischen Gesellschaft 69: 107-112. Hess, Thomas M. 1966. "Snohomish Chameleon Morphology". In International Journal of American Linguistics 32, 350-356. Himmelmann, Nikolaus. 1987. Morphosyntax und Morphologie – Die Ausrichtungsaffixe im Tagalog. München: Fink. Himmelmann, Nikolaus. 1991. The Philippine Challenge to Universal Grammar. Universität zu Köln. (Arbeitspapier Nr. 15) Himmelmann, Nikolaus. 2004. "On statives and potentives in western Austronesian (mostly Tagalog)". In Paul Law (ed.) Proceedings of Austronesian Formal Linguistic Association 11: 103-119. Himmelmann, Nikolaus. 2006. "How to miss a paradigm or two: Multifunctional ma- in Tagalog". In Felix Ameka, Alan Dench and Nick Evans (eds.). Catching Language. The Standing Challenge of Grammar Writing. Berlin: Mouton de Gruyter, 487-526. Himmelmann, Nikolaus. to appear. "Lexical categories and voice in Tagalog". In Peter Austin and Simon Musgrave (eds.) Voice and Grammatical Functions in Austronesian Languages. Stanford: CSLI. Hopper, Paul J. and Elizabeth Closs Traugott. 1993. Grammaticalization. Cambridge: University Press. (Cambridge Textbooks in Linguistics) Hume, Elizabeth. 2004. "The Indeterminacy/Attestation Model of Metathesis". In Language 80/2, 203-237. Hume, Elizabeth and Keith Johnson. 2001. "A Model of the Interplay of Speech Perception and Phonology". In Elizabeth Hume and Keith Johnson (eds.) Studies on the Interplay of Speech Perception and Phonology. The Ohio State University, 1-22. (Working Papers in Linguistics 55) Hurch, Bernhard. 1988. Über Aspiration. Ein Kapitel aus der natürlichen Phonologie. Tübingen: Narr. Hurch, Bernhard. 1998. "Optimalität und Natürlichkeit". In ZAS Papers in Linguistics 13, 115-139. Hurch, Bernhard. 2002. "Die sogenannte expletive Reduplikation im Tarahumara. Rehabilitation eines verpönten Begriffs: Euphonie". In Michael Bommes, Christina Noack, and Doris Tophinke (eds.) Sprache als Form. Festschrift für Utz Maas zum 60. Geburtstag. Wiesbaden: Westdeutscher Verlag, 51-62.



Hurch, Bernhard. 2003. On the origin of partial reduplication. [oral presentation, XVIth International Conference on Historical Linguistics, Copenhagen, Denmark] Hurch, Bernhard. 2005. "Dissimilation". In Brown Keith (ed.) Encyclopedia of Language and Linguistics. 2nd ed. London, Amsterdam, New York: Elsevier, 719-722. Hurch, Bernhard (ed.) 2005. Studies on Reduplication. Berlin, New York: Mouton de Gruyter. (Empirical Approaches to Language Typology 28) Hurch, Bernhard and Richard Rhodes (eds.). 1996. Natural Phonology: The State of the Art. Berlin, New York: Mouton de Gruyter. (Trends in Linguistics 92) Hurch, Bernhard and Veronika Mattes. 2005. "Über die Entstehung von partieller Reduplikation". In Gertraud Fenk-Oczlon and Christian Winkler (eds.) Sprache und Natürlichkeit. Gedenkband für Willi Mayerthaler. Tübingen: Narr, 137-156. Hurch, Bernhard and Veronika Mattes. to appear. "The Graz Database on Reduplication". In Faits de Langues 29. Réduplication / Redoublement. Huttar, Mary and George Huttar. 1997. "Reduplication in Ndyuka". In Arthur Spears and Donald Winford (eds.) The Structure and Status of Pidgins and Creoles: Including Selected Papers from the Meeting of the Society of Pidgin and Creole Linguistics. Amsterdam, Philadelphia: John Benjamins Publishing Company, 495-414. Inkelas, Sharon. 2005. "Morphological Doubling Theory: Evidence for morphological doubling in reduplication". In Hurch (ed.) 2005, 65-88. Inkelas, Sharon and Cheryl Zoll. 2005. Reduplication. Doubling in Morphology. Cambridge University Press. (Cambridge Studies in Linguistics 106) Jacobs, Joachim. 1983. Fokus und Skalen. Zur Syntax und Semantik der Gradpartikeln im Deutschen. Tübingen: Niemeyer. (Linguistische Arbeiten 138) Jakobson, Roman. 1963 [1966]. "Implications of language universals for linguistics". In Greenberg (ed.) 1966, 263-278. Jakobson, Roman. 1971. "Zeichen und System der Sprache". In his Selected Writings II Word and Language. The Hague, Paris: Mouton, 272-279. Jensen, Hans. 1952. “Die sprachliche Kategorie des Numerus”. In Wissenschaftliche Zeitschrift der Universität Rostock 1(2), 1-21. Jurafsky, Dan. 1993. "Universals in the semantics of the diminutive". In Joshua Guenter, Barbara Kaier and Cheryl Zoll (eds.) Proceedings of the nineteenth annual meeting of the Berkeley Linguistic Society. Berkeley: Berkeley Linguistic Society, 423-436.



Kajitani, Motomi. 2005. "Semantic Properties of Reduplication among the World's Languages". In LSO Working Papers in Linguistics 5: Proceedings of the WIGL 2005, 93106. Karstien, C. 1921. Die reduplizierten Perfekta des Nord- und Westgermanischen. Gießen: V. Münchow. Keane, Elinor. 2001. Echo words in Tamil. [PhD disseration]. Oxford: Merton College. Keane, Elinor. 2005. "Phrasal reduplication and dual description". In Hurch (ed.) 2005, 239261. Keller, Rudi. 1990. Sprachwandel: Von der unsichtbaren Hand in der Sprache. Tübingen: Francke. Key, Harold. 1965. "Some semantic functions of reduplication in various languages". In Anthropological Linguistics 7(2), 88-101. Kiparsky, Paul. 1973. "'Elsewhere” in Phonology'. In Stephen R. Anderson and Paul Kiparsky (eds.) A Festschrift for Morris Halle. New York: Holt, Rinehart and Winston, 93-106. Kiparsky, Paul. 1986. The Phonology of Reduplication. Ms. Stanford University. Kiyomi, Setsuko. 1995. "A new approach to reduplication: a semantic study of noun and verb reduplication in the Malayo-Polynesian languages". In Linguistics 33, 1145-1167. Klamer, Marian. 2002. "Ten years of synchronic Austronesian linguistics (1991-2002)". In Lingua 112, 933-965. Klamer, Marian. 2004. "Multi-categorial items as underspecified lexical entries: The case of Kambera wàngu". In Fischer, Norde and Perridon (eds.) 2004, 299-323. Kocher, Frieda. 1921. Reduplikationsbildungen im Französischen und Italienischen. H.R. Sauerländer & Co: Aarau. Kouwenberg, Silvia. 2003. "Papiamentu Reduplication". In Kouwenberg, Silvia (ed.). 2003, 161-164. Kouwenberg, Silvia (ed.). 2003. Twice as meaningful. Reduplication in Pidgins, Creoles and other contact languages. London: Battlebridge. (Westminster Creolistics Series 8) Kouwenberg, Silvia and Darlene LaCharité. 2001. "The Iconic Interpretation of Reduplication: Issues in the Study of Reduplication in Caribbean Creole Languages". In European Journal of English Studies 5/1, 59-80. Kouwenberg, Silvia and Darlene LaCharité. 2005. "Less is more: Evidence from diminutive reduplication in Carribbean Creole languages". In Hurch (ed.) 2005, 533-545. 174


Kröll, Heinz. 1991. "Beitrag zu den Reduplikationen im Portugiesischen (II)". In Lusorama 15, 25-44. Kulikov, Leonid. 2005. "Reduplication in the Vedic verb". In Hurch (ed.) 2005, 431-454. Lahiri, Aditi and William Marslen-Wilson. 1991. “The mental representation of lexical form: A phonological approach to the recognition lexicon”. In Cognition 38/3: 245-294. Lakoff, George and Mark Johnson. 1980. Metaphors we live by. Chicago, London: The University of Chicago Press. Langacker, Ronald W. 1987. Foundations of Cognitive Grammar. Vol. I. Theoretical Prerequisites. Stanford: Stanford University Press. Langdon, Margaret. 1970. A Grammar of Diegueño. Berkeley, Los Angeles, London: University of California Press. (University of Californiy Publications: Linguistics 66) Lehmann, Christian. 1984. Der Relativsatz. Typologie seiner Strukturen – Theorie seiner Funktionen – Kompendium seiner Grammatik. Tübingen: Narr. (Language Universals Series 3) Leiss, Elisabeth. 1997. "Synkretsimus und Natürlichkeit". In Folia Linguistica 31/1-2, 133160. Leiss, Elisabeth. 2005. Unterspezifikation von morphologischen Paradigmen als optimale Modellierung von Form-Inhalts-Relationen. [oral presentation, ÖLT (Austrian Linguistics Conference), Graz, Austria] Levinson, Stephen C. 2000. Presumptive meanings. The theory of generalized conversational implicature. Cambridge, London: The MIT Press. Lionnet, André. 1968. “Los intensivos en tarahumar”. In Anales del Instituto Nacional de Antropologia e Historia 19, 135-46. Lisboa, Marcos de. 1754. Vocabulario de la lengua Bicol. Manila. Lobel, Jason W. 2004. "Old Bikol -um- vs. mag- and the Loss of a Morphological Paradigm". In Oceanic Linguistics 43/2, 469-497. Lobel, Jason W. and Fr. Wilmer Joseph S. Tria. 2000. An satuyang tataramon. A study of the Bikol language. Naga City: Lobel&Tria Partnership. Lopez, Cecilio. 1950. “Reduplication in Tagalog”. In Bijdragen tot de taal-, land- en volkenkunde 106, 's-Gravenhage: Martinus Nijhoff, 151-311. 175


Lynch, John, Malcom Ross and Terry Crowley. 2002. The Oceanic Languages. Richmond: Curzon. (Curzon Language Family Series) Marantz, Alec. 1982. "Re reduplication". In Linguistic Inquiry 13, 435-482. Marle, Jaap van. 1988. „Semantics in Productivity Change“. In: Gert Booij and Jaap van Marle (eds.) Yearbook of Morphology 1988. Heidelberg: Springer, 139 – 154. Mattes, Veronika. 2002. Von der Silbe zum Segment. Die Rolle der Silbe im Spracherwerb. [M.A. thesis]. University of Graz. Mattes, Veronika. 2005. "Wortarten im Bikol". In Grazer Linguistische Studien 64, 3-27. Mattes, Veronika. 2006a. One form – opposite meanings? Attenuative and intensive interpretation of full reduplication in Bikol. [oral presentation, International Conference on Austronesian Linguistics 10, 2006, Puerto Princesa, The Philippines] Mattes, Veronika. 2006b. Looking for Parts-of-Speech in a Philippine language. Does Bikol have nouns and verbs? [oral presentation, Parts-of-Speech conference 2006, Amsterdam, The Netherlands] Mattes, Veronika and Ralf Vollmann. 2006. Lexikalische Reduplikation. [oral presentation, ÖLT (Österreichische Linguistiktagung), Klagenfurt, Austria] Mayerthaler, Willi. 1977. Studien zur theoretischen und französischen Morphologie. Reduplikation, Echowörter, morphologische Natürlichkeit, Haplologie, Produktivität, Regeltelescoping, paradigmatischer Ausgleich. Tübingen, Niemeyer. Mayerthaler, Willi. 1981. Morphologische (Linguistische Forschungen 28)




Mc Laughlin, Fiona. 2005. "Reduplication and consonant mutation in the Northern Atlantic languages". In Hurch (ed.) 2005, 111-133. McCarthy, John. 1979. Formal Problems in Semitic Phonology and Morphology. [PhD dissertation]. MIT Cambridge. McCarthy, John. 1981. „A Prosodic Theory of Nonconcatenative Morphology.“ In Linguistic Inquiry 12/3: 373-418. McCarthy, John and Alan Prince. 1986. Prosodic Morphology. Ms. University of Massachusetts and Brandeis University. McCarthy, John and Alan Prince. 1995. "Faithfulness and reduplicative identity". In Jill Beckman, Laura Dickey and Suzanne Urbanczyk (eds.) University of Massachusetts Occasional Papers in Linguistics 18: Papers in Optimality Theory. Amherst, MA: GLSA, 249-384. 176


McFarland, Curtis D. 1974. The Dialects of the Bicol Area. [PhD dissertation]. University of Yale. Miller, George A. 1978. "Semantic relations among words". In Morris Halle, Joan Bresnan and George A. Miller (eds.) Linguistic theory and psychological reality. Cambridge, London: The MIT Press, 60-118. Miller, Wick R. 1985. "Lionnet's Article on the 'Intensive' in Tarahumara." In International Journal of American Linguistics 51, 502-504. Miller, Wick R. 1996. Guarijío: gramática, textos y vocabulario. Mexico, D.F.: Universidad Nacional Autónoma de México. Mintz, Malcolm W. 1971. Bikol grammar notes. Honolulu: University of Hawaii Press. (PALI Language Texts: Philippines) Mintz, Malcolm. 2001. "Bikol". In Jane Garry and Carl Rubino (eds.) Encyclopedia of the World's Languages: Past and Present. New York, Dublin: H.W. Wilson Press, 90-95. Mintz, Malcolm. 2004. Bikol Dictionary. diksionáriong bíkol. Vol. I and II. Perth: Indonesian/Malay Texts and Resources. Mintz, Malcolm W. 2005. "Terms of Religious Adaption: The Introduction of Christianity to the Bikol Region of the Philippines". In Carl R. Galvez Rubino and Hsiu-chuan Liao (eds.) Current Issues in Philippine Linguistics and Anthropology. Parangal kay Lawrence A. Reid. Manila: The Linguistic Society of the Philippines and SIL Philippines, 167-210. Mintz, Malcolm W. and José Del Rosario Britanico. 1985. Bikol-English Dictionary. diksionariong bikol-inglés. Quezon City: New Day Publishers. Moravcsik, Edith. 1978. "Reduplicative Constructions". In Joseph Greenberg (ed.) Universals of Human Language, vol. 3, Word Structure. Stanford: Stanford University Press, 297-334. Moravcsik, Edith. 2000. "Infixation". In Geert Booij, Christian Lehmann and Joachim Mugdan (eds.), 545-552. Mosel, Ulrike and Even Hovdhaugen. 1992. Samoan reference grammar. Oslo: Scandinavian University Press. (Institute for sammenlignende kulturforskning, Series B: Skrifter, LXXXV) Naylor, Paz Buenaventura. 1980. “Linking, Relation-Marking, and Tagalog Syntax”. In Paz Buenaventura Naylor (ed.) Austronesian Syntax. Papers from the Second Eastern Conference on Austronesian Languages. Ann Arbor: Center for South and Southeast Asian Studies. (Michigan Papers on South and Southeast Asia 15)



Naylor, Paz Buenaventura. 1986. "On the semantics of reduplication". In Paul Geraghty, Lois Carrington, and S.A. Wurm (eds.) Focal I: Papers from the Fourth International Conference on Austronesian Linguistics. Canberra: Australian National University, 175-185. (Pacific Linguistic Series C No. 93) Naylor, Paz Buenaventura. 2006. Reflections on the nature of transitivity: With reference to focus and discourse pragmatics. [oral presentation, International Conference on Austronesian Linguistics 10, 2006, Puerto Princesa, The Philippines] Nevins, Andrew and Bert Vaux. 2003. "Metalinguistic, shmetalinguistic: The phonology and morphology of shm-reduplication". In Proceedings of the Chicago Linguistics Society annual meeting 39. Newman, Paul. 1990. Nominal and Verbal Plurality in Chadic. Dordrecht: Foris. (Publications in African languages and linguistics). Newmeyer, Frederick J. 1992. "Iconicity and Generative Grammar". In Language 68/4, 756796. Niepokuj, Mary. 1997. The Development of Verbal Reduplication in Indo-European. Washington: Institute for the Study of Man. (Journal of Indoeuropean Studies Monograph 24) Peirce, Charles S. 1904. "Neue Elemente". Ms. [In: Charles P. Peirce. Naturordnung und Zeichenprozess. Schriften über Semiotik und Naturphilosophie. ed. (1991) by Helmut Pape, transl. by Bertram Kienzle. Frankfurt: Suhrkamp, 339-377. (Wissenschaft 912)] Peralta Ramirez, Valentín (1991). "La reduplicación en el náhuatl de Tezcoco y sus funciones sociales". In Amerindia 16, 63-77. Pike, Kenneth. 1947. Phonemics: a technique for reducing languages to writing. Ann Arbor: The University of Michigan Press. Pott, August. 1862. Doppelung (Reduplikation, Gemination) als eines der wichtigsten Bildungsmittel der Sprache, beleuchtet aus Sprachen aller Welttheile. Lemgo, Detmold: Meyer. ( Prince and Smolensky. 1993. Optimality Theory: Constraint Interaction in Generative Grammar. Technical Report 2 of the Rutgers Center for Cognitive Science. Rutgers University. Procházka, Stephan. 1995. "Semantische Funktionen der reduplizierten Wurzeln im Arabischen". In Archiv Orientalni 63, 39-70. Pulleyblank, Douglas. 1988. "Vocalic Underspecification in Yoruba". In Linguistic Inquiry 19/2, 233-270.



Raimy, Eric. 1999. Representing Reduplication. [PhD dissertation]. University of Delaware. Raimy, Eric. 2000. The Phonology and Morphology of Reduplication. Berlin – New York: Mouton de Gruyter. (Studies in Generative Grammar 52) Rainer, Franz. 1987. “Produktivitätsbegriffe in der Wortbildungstheorie”. In Wolf Dietrich, Hans-Martin Gauger und Horst Geckeler (eds.) Grammatik und Wortbildung romanischer Sprachen. Tübingen: Narr, 187-202. Rainer, Franz. 1998. "La réduplication française du type fifille d’un point de vue diachronique". In Atti del XXI Congresso Internazionale di Linguistica e Filologia Romanza, Palermo. Niemeyer: Tübingen, 279-289. Regier, Terry. 1998. "Reduplication and the Arbitrariness of the Sign". In Gernsbacher, M. and S. Derry (eds.), Proceedings of the Twentieth Annual Conference of the Cognitive Science Society. Mahwah, NJ: Lawrence Erlbaum Associates, 887-892. Reid, Lawrence A. 1992. "On the development of the aspect system in some Philippine languages". In Oceanic Linguistics 31/1, 65-91. Reid, Lawrence A. to appear. Human Noun Pluralization in Northern Luzon Languages. Ms. Rhodes, Richard. 2004. Ongoing Sound Change in Sayula Popoluca. Ms. [42nd CAIL, Boston] Robert, Stéphane. 2005. "The challenge of polygrammaticalization for linguistic theory". In Zygmunt Frajzyngier, Adam Hodges and David S. Rood (eds.) Linguistic Diversity and Language Theories. Amsterdam, Philadelphia: John Benjamins. (Studies in Language Companion Series 72) Robins, R. H. 1959. "Nominal and Verbal Derivation in Sundanese". In Lingua 8, 337-369. Rose, Françoise. 2005. “Reduplication in Tupi-Guarani languages: Going into opposite directions”. In Hurch (ed.) 2005, 351-369. Rose, James H. 1973. “Principled Limitations on Productivity in Denominal Verbs”. In: Foundations of Language 10, 509-526. Rosen, Joan M. 1977. "The Functions of Reduplication in Indonesian". In Ignatius Suharno (ed.) Miscellaneous Studies in Indonesian and Languages in Indonesia. Part IV, Jakarta: Universitas Katolik Indonesia Atma Jaya, Lembaga Bahasa, 1-9. Ross, Malcolm. 1998. "Proto Oceanic phonology and morphology: Evidence from Taiwan". In Paul Jen-kuei Li et al. (eds.) Austronesian studies relating to Taiwan. Taipei: Academic Sinica, 727-791.



Rubino, Carl R. 2000. Ilocano Dictionary and Grammar. University of Hawai’i Press. (PALI Language Texts) Rubino, Carl R. 2001. "Iconic Morphology and Word Formation in Ilocano". In F.K. Erhard Voeltz and Christa Kilian-Hatz (eds.) Ideophones. Amsterdam: Benjamins, 303-320. Rubino, Carl R. 2005a. "Reduplication". In Martin Haspelmath, Matthew S. Dryer, David Gil and Bernard Comrie (eds.) The World Atlas of Language Structure. Oxford: University Press, 114-117. Rubino, Carl R. 2005b. "Reduplication: Form, function and distribution". In Hurch (ed.) 2005, 11-29. Sapir, Edward. 1921. Language. An Introduction to the study of Speech. San Diego, New York, London: Harcourt Brace&Company. Sasse, Hans-Jürgen. 1993. "Syntactic categories and subcategories". In J. Jacobs, A. von Stechow, W. Sternefeld and T. Vennemann (eds.) Syntax. Berlin: Mouton de Gruyter, 646686. Schachter, Paul and Fe T. Otanes. 1972. Tagalog Reference Grammar. Berkeley, Los Angeles, London: University of California Press. Schrammel, Barbara. in prep. Valency changing morphology in South-Eastern Romani dialects. [PhD dissertation]. University of Manchester. Schütz, Alfred. 1985. The Fijian language. Honolulu: University of Hawaii Press. Shaw, Patricia. 2005. "Non-adjacency in reduplication". In Hurch (ed.) 2005, 161-210. Singh, Rajendra. 2005. "Reduplication in Modern Hindi and the theory of reduplication". In Hurch (ed.) 2005, 263-281. Sohn, Ho-min. 1975. Wolaeian Reference Grammar. Honolulu: University of Hawaii Press. Spaelti, Philip. 1997. Dimensions of Variation in Multi-Pattern Reduplication. [PhD dissertation]. Santa Cruz: University of California. Spencer, Andrew. 1991. Morphological Theory. An Introduction to Word Structure in Generative Grammar. Oxford, Cambridge: Blackwell. Sperlich, Wolfgang B. 2001. "Semantic and syntactic functions of reduplication in Niuean". In John Bradshaw and Kenneth L. Rehg (eds.) 2001, 279-287. Spitzer, Leo. 1921. "Das Suffix -one im Romanischen". In E. Gamillscheg and Leo Spitzer (eds.) Beiträge zur Romanischen Wortbildungslehre. Genève: Olschki, 183-205. (Biblioteca dell' "Archivium Romanicum") 180


Starosta, Stanley. 1995. "A grammatical subgrouping of Formosan languages". In Paul Jenkuei Li et al. (eds.) Austronesian studies relating to Taiwan. Taipei: Academic Sinica, 683726. Starosta, Stanley, Andrew K. Pawley and Lawrence A. Reid. 1982. "The evolution of focus in Austronesian". In Amran Halim, Lois Carrington and Stephen A. Wurm (eds.) Papers from the Third International Conference on Austronesian Linguistics 2, 145-170. Steriade, Donca. 1988. "Reduplication and syllable transfer in Sanskrit and elsewhere". In Phonology 5, 73-155. Trosdal, Mimi B. 1990. Formal-Functional Cebuano-English Dictionary. Cebu City. Uhlenbeck, E.M. 1978. Studies in Javanese Morphology. The Hague: Martinus Nijhoff. (Koninklijk Instituut for Taal-, Land- en Volkenkunde: Translation Series 19) Ultan, Russell. 1975. "Infixes and their Origins". In Hansjakob Seiler (ed.) Linguistic Workshop III. Arbeiten des Kölner Universalienprojekts 1974. München: Fink, 157-205. (Structura 9) Urbanczyk, Suzanne. 2001. Patterns of reduplication in Lushootseed. New York, London: Garland Publishing. Uspensky, Boris A. 1972. "Subsystems in language, their interrelations and their correlated universals". Linguistics: An International Review 88, 53-71. Vollmann, Ralf. 2006. Wortbildung im Tibetischen. Ms. University of Graz. Vennemann, Theo. 1988. Preference Laws for Syllable Structure and the Explanation of Sound Change. Berlin, New York, Amsterdam: Mouton de Gruyter. Voort, Hein van der. 2003. "Reduplication of Person Markers in Kwaza". In Acta Linguistica Hafniensia 35: 65-94. Walker, Dale F. 1976. Grammar of the Lampung Language. Jakarta: Badan Penyelenggara Seri NUSA. (Linguistic Studies in Indonesian and Languages of Indonesia) Wilbur, Ronnie. 1973. The Phonology of Reduplication. Bloomington: Indiana University Linguistics Club. Wiltshire, Caroline and Alec Marantz. 2000. “Reduplication”. In Geert Booij, Christian Lehmann and Joachim Mugdan (eds.), 557-567. Wolfenden, Elmer P. 1971. Hiligaynon reference grammar. Honolulu: University of Hawaii Press. (Pali Language texts: Philippines) 181


Wood, Francis A. 1895. Germanic Studies II: Reduplicating Verbs in Germanic. Chicago. Wood, Esther and Andrew Garrett. 2002. "The semantics of Yurok Intensive infixation". In Proceedings of WAIL 4. UC Santa Barbara Papers in Linguistics. Wundt, Wilhelm. 1900. Völkerpsychologie : eine Untersuchung der Entwicklungsgesetze von Sprache, Mythus und Sitte. Leipzig: Engelmann. Wurzel, Wolfgang Ullrich. 1984. Flexionsmorphologie und Natürlichkeit. Ein Beitrag zur morphologischen Theorienbildung. Berlin: Akademischer Verlag. (Studia Grammatica 21) Xrakovskij, Viktor. 1997. "Semantic Types of the Plurality of Situations and their Natural Classification". In Xrakovskij (ed.) 1997, 3-64. Xrakovskij, Viktor (ed.) 1997. Typology of Iterative Constructions. München: Lincom. (Studies in Theoretical Linguistics 4) Zeitoun, Elisabeth and Lillian M. Huang. 2000. "Concerning ka-, an overlooked marker of verbal derivation in Formosan languages". In Oceanic Linguistics 39, 391-414.


Appendix A: Maps


Map 1: The Philippines: Indication of the languages taken into consideration


Appendix A: Maps

Map 2: Bikol dialect map (cf. Fincke 2002: 297): Indication of the places of language recordings (Names of Provinces in Italics)


Appendix B: Glossing abbreviations and symbols








locative voice






argument marker


not begun


actor voice














predicate base


conveyance voice










predicate marker




proper noun




politeness marker




















patient voice














immediateness marker














specific article










undergoer voice


locative - morpheme boundary

~ reduplication

= clitic

* reconstructed or unattested form

{} infix


Appendix C: Content of the dialogues, poems and stories of the corpus

APPENDIX C: CONTENT OF THE DIALOGUES, POEMS AND STORIES OF THE CORPUS Story telling enot na tawo 'The first human beings': The story of the creation of the first man and woman, based on an old Philippine belief. Before the appearance of the first human beings, the world consisted of only the sea and the sky. A bird was flying around in the sky. She teased the sea which therefore threw stones at her. The stones created a piece of land, where a bamboo started to grow. The bird, who was thirsty, opened a piece of the bamboo, hoping to find drinking water inside. Instead, she found two creatures inside, which were the first man and the first woman of the human race. merr_asuwang "Asuwang" is the name for a female ghost or a witch, which can adapt the appearance of a human being. It is a figure of Bikol mythology and very present in the people's beliefs and fears. This text is an anecdote about the appearance of an “asuwang”: A couple is riding on a motor bike very early in the morning, on a lonely street. In front of them there suddenly appears an old woman, moving slowly by hovering above the street. Although the motor bike is riding at a high speed, the couple is not able to reach the old woman. valentine's day This is an anecdote about the fear of "asuwangs" (cf. [m_asuwang]). A couple decides to spend the night of the Valentine's Day at a Beach Resort, where they go by motor bike. It is already dark and the road is deserted. In order to reach the beach, they have to pass a cemetery. The wife, sitting in the back, becomes very frightened, because she is convinced that spirits ("asuwangs") who live in the cemetery are flying above her, and she even can hear them. Therefore she tightly wraps her arms around her husband, whom she considers not frightened, praying intensely to God for protection. Suddenly she hears her husband praying even louder. So they arrive at the beach, praying, and full of fear, instead of spending a romantic Valentine's evening. 186

Appendix C: Content of the dialogues, poems and stories of the corpus

paul's stories Three "cock-and-bull stories": In the first joke, a Philippine guerrilla soldier received a pocket lighter from General McArthur. Later the soldier was shot by a Japanese patrol, but he dived and was saved at another beach. Two years later he found his pocket lighter on the beach, and it still worked. But it broke later when a friend spit on it. In the second joke, Philippine prisoners of war had to help at the pineapple harvest in Hawaii. While there, one of them lost his wedding ring. Some years later, back in the Philippines, his wife opened a can of pineapples and found his wedding ring in it. In the third joke, a fisherman caught some flying fish and transported them home. The dead fish seemed to be asleep inside the boat. Just before reaching the coast, the boat bumped against a big rock, the fish woke up and jumped back into the sea. ibalong "Ibalong" is a heroic epic of the Bikol region. It is a story about the three heroes Baltog, Handyong, and Bantong. These three heroes fight monsters and forces of nature and bring peace, agriculture and laws to the Bikol people. During this period the mountains, hills, valleys, beaches and islands of the Bikol region are shaped. The original epic consists of sixty verses. This text is a short oral narration from memory. Conversations magana 'Healthy appetite' is a conversation of three women during lunch. The main topic is the food (a soup and a meat dish), and the plan to organize fresh coconut. Included is a short dialogue about a new housing estate, built in the neighbourhood. bikol express A dialogue of two female friends, while cooking a spicy meat dish called "Bikol Express". They comment their actions, and exchange knowledge on the way of preparation.


Appendix C: Content of the dialogues, poems and stories of the corpus

bisita 'Visitors': Recorded while the family, who is hosting me, is having lunch, and several people drop in. First, two men approach. The housewife recognizes them as agents of a company selling electric appliances, knowing that they come to collect the money due for her washing machine. They invite themselves for lunch. Shortly after, a good friend of the family appears who has not turn up for a while. The family introduce me to the visitors, and explain to them, that I am in the Philippines in order to learn Bikol. Therefore they insist on speaking in Bikol, and not in English, despite the presence of a foreigner. agom 'Husband': An old woman, who is a widow and a mother of eleven children, tries to convince me to marry, preferably a Philippine man, by enumerating the advantages of having a husband. As the main advantage she underlines the fact that one never needs to be cold, because there is always somebody to embrace. magluto 'Cooking': Two friends discuss how to prepare the most delicious fried rice. p_asuwang 'Spirits': An old woman explains the nature of an "asuwang". It is a spirit without a body, which can kidnap people in order to appear in their bodies. Consequently, one never can be sure if a person is a human being or rather a spirit using the body in order to influence people in a certain (usually negative) way. pilipinas 'The Philippines' is a conversation of six people, talking about the Philippines in a slightly self-ironic way. The topics are religion, overpopulation, corruption, and young women marrying rich white men.


Appendix C: Content of the dialogues, poems and stories of the corpus

Poems (rawit-dawit) Tolong Bangging Bulanon, by Hanley Maldo [r-d: Bulanon] 'Three loony nights': A romantic description of loony nights, where the nature dances and the animals are happy, only the busy people in the cities do not notice the exceptional beauty of these nights. Basábas sa mahiwas na nátad, by Carlos "Itos" Briones [r-d: basabas] ( 'Cutting weed in the wide surroundings': Describes the pleasure of the meditative activity of field work. Romdom ko, by Nephtaly Jel B.Bontor [r-d: romdom] 'I remember': Memories of the childhood, which means a playful and small and manageable world, which therefore is happy and protected (as opposed to the world of adults). An lada, by Jerry E. Adrados [r-d: lada] 'Chilli': An ironic description of the Bikol people being fond of Chilli, in every dish and in every situation, with all its positive and negative consequences. Masaen ka pa? Makaraon na., by Jerry E. Adrados [r-d: kaon] 'Where are you still? Let's eat!': A paean of praise for the delicious and diversified Bikol dishes.

The texts Angelina, Paurog and Bunga are part of Jason Lobels text corpus, available on his Bikol CD-Rom Reference Set.


Appendix D: List of Bikol reduplications

APPENDIX D: LIST OF BIKOL REDUPLICATIONS Bisyllabic reduplicated stems lexeme

Mintz and Del Rosario Britanico (1985)

fieldwork Mattes (2005/2006)


dapdáp [MDL]

Indian coral tree


dagdag [MDL]

to add


ga'gá' [MDL]

to to boil herbs, leaves, etc. for the purpose of drinking the liquid, to brew






gakgák [MDL]

to walk in loose, flowing clothes, to walk with wings held out from the body


gamgám [MDL]


animal (bird)


abrasion, to scratch, to mark the surface worn out (shoes), rough skin, scratched movement/ varnish surface


raw starch, traditionally collected from arrowroot or cassava

starch flour



to brush against



gigís [MDL]

pinky little finger, little toe

body part

gígit [MDL]

to touch or do with only the tips of the fingers



squeeze one's way through

gisgís [MDL]

to return rice which still retains part of the husk for a second pounding

dog's skin disease (loss of hair, itchy)


gúgo' [MDL]

kind of tree

 and: herbal shampoo


gu'gó' [MDL]

to shake; to raffle off

to cope with everything (oneself)


gukgók [MDL]

to keep a secret


1. refuse a requests 2. to slice without lifting the knife

to kill by cutting the head

2. movement

gusgós / guságos

1. to rub 2. old, said in anger

gusgós, 2.

1. movement

gutgót [MDL]

1. to cut by drawing a knife back and forth like a saw 2. to get a rope burn

1. movement

maya bird (similar to a sparrow); “tiririt” (song of this bird)



Appendix D: List of Bikol reduplications

guygóy [MDL]

to cook greens, such as taro leaves



hadhád [MDL]

to clean or hollow out a reed or other skin disease (pigmentary abnormality; tube-like structure so that it may be used white, itchy) as a siphon,...


ha'há' [MDL]

to shake within a container



hakhák [MDL]

to gobble up, to eat quickly and voraciously


halhál [MDL]

to gasp, pant

hashás [MDL]

to spear or shoot hidden beneath the water, or obscured in a thicket


hathát [MDL]

to thin out seedlings



hawháw [MDL]

a braggart, boastful, bombastic



hayháy [MDL]

to exude pus or blood (a wound, abscess)


hidhíd [MDL]

a ritual conducted by the balyána ... (Bikol mythology)


hinghíng -> ingíng

to whisper

noun (whisper) nag-hi-hinghíng 'is whispering'


hiphíp [MDL]

to bribe, to coerce





hudhód [MDL]

to sink into soft ground, to plant or press into soft earth or mud

hughóg [MDL]


shake out


hu'hó' [MDL]

to shake within a container

empty crab


hunhón [MDL]

to select abaca fibers, separating first the longest, ...

dry cooking of fish (opposed to fish in broth)


i-rára' [MDL]

descriptive of words that are stinging, biting, sarcastic; to offend or shame s.o. with such words

is'ís (arc.) [MDL]

avaricious, mercenary


instigate s.o.

it'ít [MDL] to nibble a bit of to decide if it is worth eating, usually applied to animals, but applicable to people as well


diminutive plural

Appendix D: List of Bikol reduplications

kabkáb [MDL]

hand fan; to fan


kadkád [MDL]

to dig a hole with the hand or a stick



a rake



1. to have the legs spread apart 2. to be in a rush, to be preoccupied

kalkál [MDL]

to scratch the scab off a sore or wound

and: digging, investigating (fig. and concrete)



1. to grope, to feel one's way 2. to appropriate s.o. else's property




to walk with the legs spread apart, to walk with the arms out, as if is tucked under the armpits

also: space in between


kangkóng [MDL]

leafy vegetable (vine, possessing edible  hollow stems and leaves, growing in mud or stagnant ponds; Ipomea reptans aquatica)


1. mang- -an: (slang) to force attentions 2. on a girl 2. mag- -on: to grope for or feel for where it is not easily seen (as in the dark, underwater); to fumble 3. mag- -an: to frisk, to search for by frisking


1. fast, rapid, speedy, swift, quick 2. speedster, speed demon (fast, with a Spanish agentive ending: kaskaséro



to feel or grope for; to paw


to scratch the ground with the hand, foot, paw


kidkíd [MDL]

to pull weeds from the rice field

scratching (dogs/cats)



to be surprised

to scare s.o. Nag-highig mo ako pag-abot mo. 'Your arrival scared me.' Nag-highig ka? Did I scare you?


ki'kí' [MDL]

hop, hopscotch

and: jumping on one leg


kíkig [MDL]

water snake

kikík [MDL]

to squeak (mice, rats)



animal bird (believe: when you hear it singing, animal s.o. is dying)


Appendix D: List of Bikol reduplications


to extort, to blackmail

1. to nosh 2. to take from s.o.


to scratch (as a dog does when he has fleas)

remove from the shell (by scratching)


kimkím (or: 1. to bribe (sl-) ikím) 2. to carry tucked under the arm

holding very tight (fig. and concrete)


to pick at one's food, showing little interest in eating, to eat with just the tips of the fingers, to eat slowly and fastidiously



kitkít [MDL]

a small scab, to pick or scape off with a finger nail


kubkób [MDL]

1. to place the arms around a seated 1. child while standing at his back in order (protect, cover) to warm or protect him 2. to mate (fowl)

kudkód; an to grate coconut; grated coconut kinudkód


ku'kó' [MDL]


body part

kukó(d) [MDL]

the nails of animals such as the water buffalo, pig

and: nails of humans

body part

kukók [MDL]

the clucking of hens when calling their young or other hens



to cuddle or cradle, to have s.o. sit on one's lap

kumkóm [MDL]

to fold the arms across the chest

make a fist; be stingy


1. to shrink 2. to stuff a hole


kungkóng [MDL]

to walk hunched over

kupkóp [MDL]

describing hair that is short or plastered sheltering down, a crew cut

kuskós [MDL]

to call cats

to rub

sound / movement


a pit, hole

digging with the fingers






labláb [MDL]

to drink excessively

 also: laklak


ladlád [MDL]

to unfold or unfurl

to make it know, “outcoming”



Appendix D: List of Bikol reduplications


to dismantle, to take apart, to demolish



dirt in the creases of the skin (of the neck, arms, legs)


1. to create the world and its creatures 2. creative (arc-) [MDL]

laláso' [MDL]

caterpillar (hairy, causing a burning sensation when touched)

scattered wounds


lamlám [MDL]

to take more than agreed upon

a sermon


lanlán [MDL]

to separate into groups according to unknown type



-on: to slice or cut meat away from the  to peel away a thin layer bone -an: to remove the outer portions of wood or bamboo with a knife or bolo, to shave the bark off wood or bamboo,...



to split, rip, to tear away

cut (e.g. slit one's wrists)


loose, baggy; dangling,...

and: fish species







dangling or hanging loosely




1. to avoid, bypass, go around, detour around 2. to go to a place unseen, unnoticed


lislís / luslós

curled up, upturned (as a lip), blown up (as a skirt)


to season over an open fire (banana leaves used for wrapping rice, sugar cane, wood to enable it to be straightened)

, soften by heat (or water)



to grow worse, to become more grave (an illness), to deteriorate (a condition)




noodle dish (served either dry or in a soup, containing a sauce with ground tinapá)


luklók [MDL]

to put a whole handful of food in the mouth (as a child might do)




hidden away, out-of-the-way, hard to find, secluded, obscure

, the side of a mobile for example, which cannot be reached (e.g. the side of a bed or cupboard touching the wall)

luklók [MDL]

the side of a house


Appendix D: List of Bikol reduplications


maghing-, hing- -on: to shed fur (as a dog); to change skin (as a snake); to molt (as a bird)

no: hiluno

lu'ló' [MDL]

to soften bark by soaking it in water to facilitate the stripping of fiber (usually the bark of the ábaca, malub'gó)

a nickname


doleful, gloomy, melancholy, morose, dreary, drab, dull, overcast, cloudy

lunlón [MDL]

to clean weeds and grasses from the edge of a cultivated field, throwing these weeds and grasses toward the centre of the field



1. dangling, hanging by a thread, ... 2. flabby, flaccid


madmád / ma'bád [MDL]

to arrange, to place in order


malmál da'íng [MDL]

of no use or value


ma'má' [MDL]

to maintain or hold (as beliefs); to be convinced of, ...

no; baby talk for playing "Carabaw"

mangmáng illiterate, uneducated



acquainted, familiar with; to know or be  acquainted with

mikmík [MDL]

to step on or crush with the foot



milmíl [MDL]

to get bruised or dented by falling




breast (human); to suck milk from a breast or a bottle

baby talk for mother's breast; suck: sopsop

baby talk



eat like a bird



worn thin, worn out



1. overused, worn out, worn down, ... 2. to wipe up the floor with s.o. in fighting







mukmók [MDL]

1. to beat s.o. on the head until he collapses 2. to grind or chop finely

day dreaming, close the eyes


1. to suck the fingers; to suck 2. to do slowly and carefully



Appendix D: List of Bikol reduplications


1. describing s.o. who does not immediately answer what is asked of him; tongue-tied 2. (fig-) stupid







bump with the head



describing dirt on the face

bump with a body part



endearment for mother / aunt / ...

hypocoristi c


endearment for small monkeys

name for boys

hypocoristi c

mungmóng to get everything go your way

lack of common sense

musmós [MDL]

to slowly rot or decay


naknák [MDL]

to grow larger (a wound)




title for a mother, aunt or godmother

, but not for "mother"

hypocoristi c


pus (Eiter)


name for a young girl

name for the youngest daughter

hypocoristi c


name for a young boy

name for the youngest son

hypocoristi c


gnat, gnats (found swarming around ripening fruits)

plural, animal

nunó' / nánok

sound, deep (sleep!)




to run (the nose)


to scour, scrub; to rub with a cloth


to be exactly what one wants or desires [MDL]


ngákngák [MDL]

bird of prey (large)

laughter (also: ngarakngak)



the sound of birds such as the kite or eagle; the bellowing of the water buffalo; the cackle of certain types of laughter, ...




to bite or gnaw on

to rip with the teeth



ngawngáw to prattle, chatter, jabber [MDL]



Appendix D: List of Bikol reduplications

ngawtngáw to swallow one's words t [MDL]


ngingí' [MDL]

to cry (babies)

slang for cry


ngipngíp (ngípon: tooth)

to bite, to chew, to gnaw on

no: ngatngat



descriptive of the darkness of night

anger deep inside


ngudngód [MDL]

to dull a knife blade by striking it against a store

to scrub (angry, aggressive)



to sob, to sob over

to cry


ngupngóp / to chew or gnaw on tough or fibrous unknown ngutngót things (as a water buffalo chewing on its tethering rope)


ngusngós [MDL]

to whine (a dog)




1. to be blown away by the wind 2. to drag anchor (a boat driven by the wind)

 napadpad siya sa Austria = he ended up in Austria (unpolite)

padpád [MDL]

to cut off the top of the coconut




to shake out (as a rug); ...



wing (as of a bird)

and: clap



stupid, dull


to drive in (as a post into the ground)


paypáy (mang-)

to fan

and: to beckon


prostitute, whore, harlot

panggáng [MDL]

to dry-fry in a kwáli' producing a result similar to toasting, rather than frying



worn thin due to long use (referring only to pots, kettles, etc.)



pangpáng / river bank; bluff; pampán



bamboo cot or bed

pápan [MDL]

upper crosspiece with a loom to which the threads are are attached



a name commonly used to refer to address a father, uncle or grandfather

hypocoristi c

paspás [MDL]

to purify gold in a crucible

by scrubbing; and: move faster (also: kaskas)



Appendix D: List of Bikol reduplications


a method of threshing in which the rice is beaten with sticks


paspás (> fast, quick, rapid, ... (speedster, speed paspaséro) demon)



to unravel thread

stick (Tagalog)

paypáy (mang-, -an)

to fan


paypáy to beckon to; to motion to s.o. to come (mag-, -on)

pidpíd [MDL]

a small bamboo enclosure placed in a field for he purpose of snaring birds or trapping rats



small fields set aside for planting of rice a game seeds, serving as a seed bed

pingpíng [MDL]

to be filled to overflowing / to fill to overflowing

aiming at, to shoot at

pípi' [MDL]

to sit with the legs crossed, the knees together and raised toward the chest, and the skirt pulled firmly down over the legs (women)

female genitals (polite term)

pi'pí' [MDL]

to wash clothes

compress, which looses thereby its original shape (e.g. crumpled clothes in a suitcase)

pispís [MDL]

the flower of sugar cane or reed grasses sideburns; hairline above the ear (men and women)


worn down or out, thin; depleted, unknown diminished (as a harvest); defoliated; ...


matted down (as wet fur, feature)

sound of the horn of the car



to whip lightly, to spank s.o.




fast, speedy, to speed up, ...



pudpód [MDL]

to make fish balls



pudpód / puspós

worn down or worn out; overused

pugpóg [MDL]

to shed (feathers, fur)



to beat, hit, strike s.o. or with a stick

1. beat s.o. with a stick on the head; also: 2. prostitute 3. material for weaving clothes



body part

Appendix D: List of Bikol reduplications

pumpóm / punpón

to collect with the hands (things left to dry in the sun, fish from dry fields)


a string of, a bundle of


pungpóng [MDL]

to close off the mouth of a canal

to stop a stream (blood, water, ...)

pungpóng [MDL]



pupó' [MDL]

a hot-tempered, vengeful spirit who, by touching the head of a child with its hand, causes the child to grow weaker and weaker until it dies; [Bikol mythology]



to wash the behind after defecating

wash the genitals (children)


puspás (!!) rice porridge (commonly given to the babies or the sick) with native chicken



to beat down, to hit or strike




worn out or worn down, overused


to use up or spend everything




crew cut (as the hair); cut short, into pieces

cutting/preparing of the firewood



the sound of beep-beep, honk-honk, ...

of the train/horn


puypóy [MDL]

a hen with small tail feathers which face unknown downward



to cut the stalks of rice, grass, etc. with a knife; to harvest rice in this manner

slang for arson



referring to fallen crumbs, particles of food, grains of rice, dirt, hair, ...

carelessly making gaps, where they should not be (e.g. in the floor)


rakrák [MDL]

to break, smash or rip in anger




poisonous, to become infected with poison, ...


rára [MDL]

to weave mats, baskets, ..., from palm leaves



to lap up, to drink (animals), fowl, ...

da'ra' (i.e. a carabaw)


1. wasteful, flighty, mildly - disturb s.o. irresponsible; restless, ... - to spoil -> waste, destroy 2. to deride, mock, ridicule, scoff, to defile, ... 3. to cast a spell over, to fall ill due to a an evil spell



Appendix D: List of Bikol reduplications

raydáy / rayráy

to fray; to become unstrung (as beads), ...


ribríb [MDL]

to chip a piece off, to get chipped


rimrím [MDL]

to drink (birds, fowl)

pig's loosing of appetite


to sprinkle (as one might do to clothes before ironing)

female nick name


decaying or rotting teeth

push-start a car

rugróg / ragrág

to fall (hair)

rugróg, rorog (also: falling scurf)


to feel depressed or dispirited

not used with humans: chicken's illness emotion (passive behavior); also: ruyroy


fan palm (tree possessing fronds used for the thatching of houses)

see rukrok


to graze, to graze one (as a particular type of grass)

animals eating of grass (slang: ngabngab)


to cut weeds or grass just below the surface of the soil, not necessarily removing the root stock; to cut grass very short

full-length (dress); sink

sadsád [MDL]

to perform dance steps

, at ati-ati-han festival, in Bisaya, in honor of St. Niño (January 1-15)


saksák / taksák / tasák

to stab

 saksak: stab repeatedly (ta(k)sak: stab once)



to plug in; to insert, to plug or s.o. insert into

salsál [MDL]

to thin or flatten out metal by pressing or beating


 to accumulate wealth by somewhat ruthless means; to grab money or land; to commander, to seize or confiscate, ...

sansán [MDL]

to squeeze or push into a small opening (such as a gunpowder into the barrel of a gun)



to grope for



to stuff into the mouth with force

also: sa'sa'


disobedient; to insist on having one's own way

bad order


fish (small, silver colored, disk-shaped,  saltwater)






movement plural


Appendix D: List of Bikol reduplications


to lop off, to cut off the rough edges of wood, bamboo, ...

sasá' [MDL]

to blunt a point or tip

to enjoy


to crush or smash (large items such as metal drums, cars)



to prepare fish by rubbing salt into the flesh until the flesh becomes soft



gossipy, a gossip (Spanish: chacharear) 


to wash (as fish or meat before cooking); to rinse (as rice before boiling), ...



to dip (as food into a sauce)

sawsáw (slang)

to defeat s.o. in a game with a big difference in score


sasáy / saláysay

to narrate or chronicle, to tell a story to,  ...


sen sen, Chinese breath freshener


sibsíb [MDL]

to eat that which falls from the table or falls under the house (animals)

also: sabsab; eating of the grass


sidsíd / saridsíd

to feel for (with the feet) underwater, sand, mud

sidsid:  saridsid: noise of dragging



to drag the feet, to shuffle the feet




to feel for lice, fleas, sand, etc., in the hair, fur, feathers, etc. of a particular person, animal or bird

 and: siksik, soksok: squeeze


to preen or smooth the feathers (birds, fowl)


simsím [MDL]

a sharp point found at the prow of certain boats ...

also: samsam (s.o.)


to feed on or eat (geese!)


ring (as worn on the finger)



blame (Tagalog)


to rip or split open; to slash


to call one's attention by hissing (Spanish: chichear)


to fall forward



to sear, singe

to burn


subsób [MDL]

to be standing in a lot of water (transplanted rice)








Appendix D: List of Bikol reduplications


(fig-) “immersed in work, study”

sudsód [MDL]

to contradict or disagree with s.o.

jump on one foot


sudsód [MDL]

to be overloaded with cargo (a boat)




the point of a bladed weapon; the cutting edge of a knife, plow or other bladed instrument


sugsóg [MDL]

describing or s.o. that continually causes harm or destruction (as a wild boar entering the rice fields, pirates raiding a town)



never late, always on time (in doing

very regular, strong

suksók / su'sú'

to insert into a narrow slit or opening


to shuffle cards


to be remorseful

and: brainwash

sumsóm-an roasted water buffalo or boar meat eaten as a climax to the hálya ritual [Bikol mythology]

movement mythology


to eat while drinking beer, wine, liquor


to chase fish toward shallow water where they can easily be caught

to follow, fetch

sungsóng [MDL]

to go against the current or into the wind

wade in the water; and: interfering in someone else's affairs/quarrels


to clog or stop up



to sip or suck, to draw liquid through a straw, to puff on a cigarette


to cram, crowd, to force one's way into or push one's way through ...


to snuggle, nestle, nuzzle

breast, shell



to suck

to suck



to foment, to incite s.o. to act in a particular way



to insert, to plug in, ...

gossip; to suck in


to cut bamboo from or rattan; to cut the  fins of fish, the wings of fowl; to hamstering, cut the tendon in the hollow of the knee



movement, sound


Appendix D: List of Bikol reduplications

tabtáb (-on)

fins of a fish


body part


chopped, minced




riddled, cut in many places

fish filled with salt




 (scatter as in body painting)


tiring (a long trip); to be tired or bushed shake (e.g. on a holey road) from a long journey

tagtág [MDL]

to share out



to tap, rap; to dislodge by tapping or rapping

shake out

takták [MDL]

to spread out nets for fishing or hunting -


to get ejected; to get thrown (as a kicked also: badbad rock)

taltál [MDL]

to release or set free from prison


taltál (slang) (maka-)

to end up


tantán [MDL]

to do with great caution or and: leave alone circumspection (as eating fish slowly so Tantan-an mo ako! = Leave me alone! that one does not swallow a bone; walking carefully in the dark so that one does not trip over


the clanging sound of a bell

and: tarangtang  and: rough surface

tangtáng [MDL]

to remove a cupping glass



to tap with the open palm


to fray, unravel


doorway, lintel; door



a title for a father, uncle, or godfather




 (every boy has to be circumcised by his 12th birthday )


tattoo, brand, label, trademark, ...



a line, row



tíbtib cutting grass, etc.

tibtíb [MDL]

bolo, said in anger




movement movement, plural



hypocoristi c

Appendix D: List of Bikol reduplications


an impromptu rhyme

 (the rhyme is important, not the content)


a toast


tiktík [MDL]

to dislodge by tapping a container gently with the fingers

movement, sound


a call for chicks




secret agent, spy, detective


to taste with just the tip of the tongue or lightly with the lips


a tinkling sound, the clinking or ringing broom stick (Tagalog) sound when metal hits metal


a dent, a notch, ...



to drip

and: baby talk for male genitals


a hollowed coconut shell used as a dipper or as a container for water



a thumping sound (as when hitting s.o. on the back)



tubtób [MDL]

to crop or cut the hair



tubtób [MDL]

to pay all of one's debts


tudtód [MDL]

to sock or punch




music, to play a music instrument


tuktók [MDL]

to peck food




to chop into tiny bits



to knock on


tuktók [MDL]

referring to that which is find in the head of a fish

tip of a mountain or a roof; slang for mind


to learn, to grasp or get the hang of



well-behaved or well-mannered


to cook a dish called tinumtoman



out, dead (a fire)


tumtóm (fig-)

wet, soaked: Tumtóm sa tingángis. 'Soaked with tears.'


one of the Easter Sunday processions held in Naga City

take off, lowering hangig


movement sound




Appendix D: List of Bikol reduplications

tungtóng [MDL]

to tap (so as to dislodge stuck inside)




to step on; to stomp on


tustós [MDL]

drooping, hanging down, extremely slack




referring to well-fried or crisply fried food



tustós (Tagalog)

to pay for someone, to support s.o.






an intimate friend




referring to the slight jerk on the fishing  line when a fish begins to bite

tuytóy [MDL]

to cross a river by means of a bridge, to unknown ford a river

um'óm / um'ón

to stuff into the mouth


umóm / humóm

to immerse in water


ung'óng [MDL]

to eat directly from the pot or serving plate, or drink from the container in which it is served, and not from one's own plate or glass

 also: um'óm (only drink, not eat)


to slide down a little bit (as when lying and: shrinking on a mat or in a bed in order to be more comfortable)



us'ós / as'ás to ooze (as pus from a wound)


ut'ót / huthót / putpót

to use up or spend everything, to loose everything

ut'ót and putpót: empty; huthót: empty movement


to scatter (as grains), to leak (as rice from a sack)



rice species



to shake out ( arug, clothes to remove dust or lint)



to disembowel, eviscerate, gut

to ooze out

wakwák [MDL]

bird (said to walk for one month on one  foot, and for the next month on the other)



baggy, loose






Appendix D: List of Bikol reduplications


having a tongue hanging out (as a dog that is tired)

, and speaking without sense tongue hanging out: also diwal-diwál

waswás [MDL]

to lighten the cargo of a boat during a storm

1. empty 2. be in arrears (in a game, competition)

waswás [MDL]

to drain off water or remove mud from a  well for the purpose of cleaning the well



to remove a thorn or splinter (with a needle or pin or by opening the area with the point of a knife)



wáwa' [MDL]

to hang out (as the intestines of a unknown wounded animal); to pop or buldge (the eyes of one who is ill)


wa'wá' / wagwág

loose fitting, slack

and: empty out by turning over and shaking



to shake excess water from; to shake down a thermometer, to shake a pen to see if it has ink in it, to shake the penis after urinating to remove the last few drops of urine, to jiggle

to besprinkle (e.g. the ironing)


wikwík [MDL]

bird of prey

sound of the bird

animal, sound


describing s.o. with a hanging lower lip 




to slip


yabyáb [MDL]

to drag a wing (birds, fowl)



to grate (sweet potato)

 (e.g. grate leaves from a stipe)

yangyáng [MDL]

to go astray, to become delinquent




movement, plural

also: lamlam to take a position against s.o.


to fan s.o. with a hand fan

also: paypay

yawyáw / yawíyaw

anything used to chase away flies or birds


tired, fatigued, bushed, having a weak feeling in the limbs


baby sitter, governess



sagging, drooping, hanging open or hanging loose





Appendix D: List of Bikol reduplications


to grate (sweet potato)

 (e.g. grate leaves from a stipe)

movement, plural


a yoyo

movement, plural

yudyód [MDL]

to be suspended in the air, to be hanging pulling with force freely



to shake (as a tree)




yunyón [MDL]

to take a position against s.o.

no: yamyam/lamlam

yupyóp (slang)

to smoke

and: inhale


to jump or bounce up and down



Lexical partial reduplication lexeme

Mintz and Del Rosario Britanico (1985)

fieldwork Mattes (2005/2006)


alimpu~púro / ampu~púro

the part at the top of the head from which the hair appears to spiral out in different directions

arim(pu)púro fontanel

body part




animal (insect)

alinta~táw / alinaw~náw [MDL]

center of the eye containing the iris and the puppil


body part

alu~luypán (aluhípan, ulaypán)



animal (insect)

alu~luntí / ula~lantí



animal (insect)



amamansít cricket

animal (insect)

ama~mátak scorpion (amátak) [MDL]

today: scorpion

animal (insect)


mana~nanggál witch, who can divide her body into two parts


witch (Bikol mythology)


Appendix D: List of Bikol reduplications


vampire (Bikol mythology)

(nagdu~)duno witch, who visits dying persons with the intention to get their spirit by sucking their blood with a long tongue



echo, reverberation





animal (insect)

1. anta~taró 2. anta~taro-on

caterpillar (sp.: large) to be eaten by...

1. anta~taró 2. unknown

animal (insect)

anu~núhot [MDL]


very tiny centipede

animal (insect)

anu~núngkot [MDL]





ape (Bikol mythology)




an affliction characterized by a swollen lymph gland, usually in the armpit or groin

Unknown (bagá' 'lung')



kind of grass

long grass, “carabow grass”



to bewitch, charm, enchant, put a spell on

an extraordinary long chin

body part (aberrant)


sound of a rooster crowing




ripe coconut with uneven coating of unknown meat and little water



cassava cake

or: rice cake



having the legs spread apart

Nagtutukaw siya bikangkáng. 'She body is sitting with her legs spread apart.' position (pej!)

bu~bu'a [MDL]

a puckered swelling occurring on the genitals of some women


bu~bugaw [MDL]

death knell



bu~bugwangon [MDL]

the stomach of animals, fish


body part

bu~bulat [MDL] to travel early in the morning




red. unknown (buláte 'worm')

animal (insect)

bu~bungaw [MDL]

hernia, rupture



bu~butkan [MDL]



body part



distributed, spread widely



Appendix D: List of Bikol reduplications

bula~lakáw [MDL]

meteor, shooting star


shooting star





to rummage through




fine details



tadpole; fish-species

very small, blowed up fish, poisonous

animal (fish)


frequently, often

gradual, slow dayáday niyang piglingawan su saiyang problema 'he slowly, gradually forgot his problem' (stylistically high expression)

plural, diminutive

dayang~dáng [MDL]

kind of plant




hell (Bikol mythology)


animal (insect)


to search with the hands in the sand for clams, mussels; to paw


animal (insect)



today: lobster

animal (sea)



gasá~gas [MDL] rain storms which come from the north in December girá~ray

again, over again, ...


to rub the body against (to scratch)



gisá~gis ; gis-gis skin disease of dogs (itchy)

plural / disease

wild grass


hagabháb [MDL]

sound made by the flapping wings of unknown large birds in flight

hagáhag / hagáhap




whizzing sound

rumor (< hinghing whisper)



buzz, drone, hum, whir

sound from an cave, echo, murmur from a contained place



to stretch a rope between two points hang the washing on the line above the ground halayháy-an 'clothesline'


to rub gently, to caress, fondle, ...

diminutive plural

h-ar~abáhab [MDL]

illness characterized by fever

stomach acid




Appendix D: List of Bikol reduplications


- descriptive of clothes which which big spaces in between (i.e. teeth), are much too large for the person implies: things in a row wearing them - separate (opposite of touching; in close proximity


warm and humid, muggy, sultry



painful, itchy or stinging sensation of the skin

slight hurt due to the newness of a wound; fresh wound is exposed to air, water, medicine, ...

sensation (pain)


- to pain of a lush, blow, burn - breeze

bearable pain

sensation (pain)


fresh, refreshing, cool

comfort, ease, when relieved of a problem, success Hayáhay ang pagmati kang ina na maray na su aki niyang naghilang. 'The mother is relieved after the recovery of her child.'

sensation (emotion)


kind of soup

Chinese soup

food (soup)



to loll (e.g. in a chair)


small bamboo branches which are 1. a small bamboo, which spreads cut and used for fish corrals, fences, widely etc. 2. fancy properties



Adam's apple

coughing sound



product, goods, wares

business (Tagalog)

kalamkám [MDL]

ticklish, to feel ticklish


sensation (emotion)

kalingkíng / kinalingkíng

sweet potatoe sticks coated in a batter made from flour and fried

interwoven shape (the dish is prepared exactly in this shape)


karangkáng [MDL]

describing the open hand with fingers spread apart

sitting with the legs spread apart, walking with the legs apart, walking on the tiptoes


karaskás [MDL]

a rattan band worn around the neck or wrist as a sign of mourning

unknown or old

karigkíg [MDL]

to walk quickly taking small steps



karúkad [MDL]

to search for by turning a container completely upside down



karuskós [MDL]

to roll up the sleeves or trouser legs; kaluskós to lift the skirt off the ground


astray, off course, ...

to get lost; today: nalagalag especially referring to gods or spirits, that are responsible for getting lost




Appendix D: List of Bikol reduplications


1. enchanted, bewitched,... unknown 2. dance (ceremonial or worship, danced by 4 to 8 couples) 3. Lagayláy: a song and dance drama associated with the feast of finding the cross (May 3),...


man, boy, male,...

lalála / lálaw

(arc-) to mourn the dead of ...

laláw 1. wilde (animals) 2. “good-for-nothing” (persons)

an mga na-lalagdá'

contents, table of contents

it's written (The fate is sealed.) lexicalized derivation from lagda '(lit-) to publish or to print; signature/destiny' (Tagalog)

lalánat [MDL]

to follow in single file behind s.o.



the extra part of skin at the rear of a (Tagalog) chicken from which the tail feathers today: buntód grow


a bruise, a black and blue mark


face, visage



to gargle, to wash out by gargling (as  bits of food)

litgít [MDL]

music instrument (similar to a violin)


referring to a land, territory, or a anywhere, distant region, usually far and unfamiliar to Sain kan lupálop nagduman? the speaker 'Where have you been?' (implicit: long and far away) (lupá land, property (Tagalog) Bikol: dagá)


tired, withered

weak, beaten up, effect of desperation



tree (Hopea mindanensis)




convalescence, recuperation

condition of a patient just / not yet fully recovered

 laki unknown láki [MDL] monster, mixture of man laki-láki 'pride' pig-laki-laki niya ang power 'he is and horse, ... flaunting his power' Pig-la-laki-láki ko ang agom ko police. 'I take advantage of my husband's being a policeman.'

also laslas; playing the litgit

body part

body part


nayuknók [MDL] tiny; in tiny pieces; ...




body part



Appendix D: List of Bikol reduplications

ngarungóg [MDL]

the sound of one crying in fear or with low, heavy sobs




sound of a starting motor


wooden sound



the sound of flapping wings


rustling sound (such as that made by sound from leaves, cloth cloth or canvas blown by the wind)



bird (similar to a swallow, but larger)

animal (bird)


a knocking sound; the sound such as  that made by an ax when chopping wood


fish (usually dried)

sound of the train, blowing the horn animal (fish)


humble, to humble s.o. before

pakúmbabá' act of humility

pala'pá' / pla'pá'

the full branch of the coconut or other palm trees, frond, the full leaf of the banana plant

pala'pá' 'leaves of coconut'

palás-pagás / pulós-pugós

to move carelessly without particular and: sound of restlessness attention to where one is going or (animal/person) what one is doing

movement / sound


pampano (fish) Spanish

animal (fish)


fin, flipper

body part

paró-pagulóng / puró-pagulóng

winged beans



payagpág [MDL]

describing thinning hair or s.o. with thinning hair

to change (positively) Digdi kami sa lugar na-mayagpág. 'This is the place where we made the difference.' pagpág 'shake out'


to roll between the palms of the to rub the eyes hand or the fingers



1. to twist or curl 2. (fig) clinging, hugging (as wet clothes on the body)


ra'akrák [MDL]

a cracking sound such as that made  by a breaking branch or a tree falling today: ragakrak in the forest

1. rice bird 2. threads over use

to twist




ragabnáb [MDL] to break (the voice)




the crackling of the fire



ragadnád [MDL] poor quality sugar cane



Appendix D: List of Bikol reduplications


the ra-ta-tat-tat sound of a machine gun; the clacking sound of a noisemaker



- ragasnás - ragasrás

- a rustling sound (such as that made ragasnás sound of a crawling snake sound by starched skirts); a swishing sound (such as that made by long grass in the wind) - a scraping or rubbing sound; the sound of sanding with sandpaper

ragawráw / ragawdáw

the sound of water being poured into unknown a container, the sound of one liquid being poured into another


ragaydáy [MDL] the sound hacking with a knife

Camarines Norte dialect

ragindín [MDL]

to carry to the edges of a field the rubbish cleaned from the center

unknown, but known as a family name


a rustling sound (such as that made by leaves); a swishing sound (such as that made by long grass in a breeze)

discrete movement of animals

sound / movement


the sound of tearing (such as a cloth), or ripping (such as a paper)

 snapping sound (breaking of cloth e.g.)



swamp reed (parts of which are used sound of blowing of the wind in the plant/ in the weaving of mats and baskets) swamp reed sound


distant sound that seems to intensify


sound of an approaching / starting motor


ragukrók [MDL] the sound of loud snoring

collywobbles ragngak sound of snoring





ramismís [MDL] tasteless, insipid (food, drink), ...

taste, sweet Ramismis ang tubó'. 'The sugar cane tastes sweet.'

sensation (taste)

rapárap [MDL]

to wander far and wide in search of

feel one's way in the darkness, clouded eye (arap blind)

raráma' [MDL]

to speak ill of s.o.

even Rarama ang bunga kan mangga kaya ma-gayon hiling-on. 'The fruits on the mango tree are evenly spread, therefore it is nice to look at.'

raráng [MDL]

a cut piece of sugar cane, ready for eating, ...

tree which looses its leaves in a certain season


a roaring sound (such as that of a waterfall or the pounding surf)

a buzzing sound



Appendix D: List of Bikol reduplications

rimírim [MDL]

to hold ill-will toward s.o.

the loosing of the appetite of the pig

riwáriw / ruwáriw


ruwáriw a good-for-nothing riwáriw go to many places without good purposes


cloudy, overcast; describing a time when either the sun or the moon is obscured

rumiróm also: dark because of a shadow, unsteady light

animal (bird); plural

sagadsád [MDL] to clean the weeds from fields, to clean the fields of weeds




plunk, the sound of a loose guitar string; the dull sound of a cracked bell




a screeching sound, a hissing or fizzing sound, ...




the sound of a boat cutting through the water

sound of reeds in the wind


saguksók [MDL] the sound produced by suction or sucking (such as when sucking the marrow from bones)




sound of a particular bird and also animal the name of the bird (a small bird in (bird); the rice fields) sound

bird (red, worm eating)

sagunsón [MDL] to go in search of one who has and: a person who is accompanied previously gone in search of another closest to the goal and has not returned as expected Ang bisita s-in-angunsón/pigsangunsón ninda. 'They accompany the visitors.' sagutsót

the patter of running feet; a slurping calling sound, through a whistle sound

salagság [MDL]

rice that grows thin and sparse in the simplex: wall of bamboo fields red.: manner of flooring the house with bamboo slices

saligsíg / sisíg

to shake the winnowing basket in order to separate small grains and dirt from the larger grains; to sift

process of getting rid of the hull of the grain


salingdíng [MDL]

to suffer for another reason; to sacrifice so that another person may go free


sensation (emotion)

salinggógon [MDL]

tree (used by smiths as fuel for their forges)




pancake (native, cooked from a batter made without eggs)





Appendix D: List of Bikol reduplications


a splinter

 Na-salugsog-an ako kaya makulog. 'I have a splinter, therefore it hurts.'

salungsóng [MDL]

a small rivulet or stream that flows into a larger watercourse


sanipsíp [MDL]

describing difficult to find due to its small size (such as a needle in a haystack)

anything that has been sharpened; to patch

sari'áyay [MDL] to stumble when walking (as one drunk)

suru-sarayad 'stumble, unsteady walking'


saridsíd / sidsíd

to feel for underwater or in the sand with the feet

sidsíd 'movement of a diver' saridsíd 'movement, a stroke of swimming (on the surface)'


saridsíd [MDL]

to stub the toe on




a sucker or shoot which grows from the base of a rice plant




to chip off




pick up odd bits of information unknown about s.o. or (as by overhearing part of a conversation, seeing in passing, etc.)

saruksók [MDL]

leather from the hide of the water buffalo worn across the chest and shoulders for added protection during combat”


sarumsóm [MDL]

to penetrate, to become absorbed (a liquid)

more common: laylay-an 'spilled and absorbed liquid'






sarupúsop [MDL]

a sucking sound (such as that made by a child nursing or sucking on a nipple)



sarúro' [MDL]

aqueduct, canal

sagúrong 'conveyance, conducts water'

sawáwa [MDL]




by chance, by luck, a lucky shot found unexpectedly, by chance (e.g. a big fish, gold)


to stagger when walking; to walk unsteadily


fish species

galunggong (short: gigi) 'small fish, animal fish of the poor' (fish)



abnormal / disease


Appendix D: List of Bikol reduplications


to follow closely the progress of; to follow closely in the tracks of

continuous attendance to (e.g. a TV series)


to say derogatory in order to stop an argument

act of blocking people, harassment, embarrassment


to block a shot (in basketball)



su-súman [MDL] old couplets or romances

red.: unknown


a rapping or pecking sound; the sound of two pieces of wood being knocked together

sound of metal

taga-ma-súso [MDL]

midwife (taga: agent; súso: breast)


tagamtám [MDL]

herb (medicinal)




twang, plunk (the sound of a loose guitar string)




a ticking sound

whip, rood



barren, infertile, unproductive

plant, which is infertile



a jingling or tinkling sound (of coins, sound of e.g. bamboo a bell)





kind of stain, spotty (due to not washed clothes)


sound of a fruit, tapped with the sound hand to determine, whether it's ripe

tagudtód [MDL]

packed (as the soil on a well-traveled unknown dirt road)

tagustós [MDL]

a creaking sound (such as that made by the supports of a house when too many people are inside)


talastás (lit-)

to find out about

advertisement (Tagalog)

talá'-tagá' [MDL]

to hack, to slash; to make cuts in the unknown side of a tree for the purpose of collecting resin or sap

movement, plural


spine (of humans, animals; of a leaf) bagbung

body part

taludtód [MDL]

the ties used for lashing the awning called kayáng to a boat


taluntón [MDL]

to lift by placing rope or rattan beneath the object, and then pulling upward on the rope


taluptóp [MDL]

to cover (a jar with a cap, a basket with a cloth, etc.)




Appendix D: List of Bikol reduplications


referring to the flow of water or other liquid down a vine or branch

taguytoy 'flow of water, slow, trickle'


a makeshift bridge, a small bridge



a small mythological forest creature said to lead people astray; when it laughs, its lips open to cover its whole face

fearful figure with big mouth, the mythology lips can cover the whole face, when it laughs (no gender)


coconut tree (dwarf)



a travelling bag made of woven reeds

 container, purse (specific use for nuts (to chew))



roll up one's sleeves


woodpecker (bird)

animal (bird)

tariktík (fig-)

quarrelsome, easily piqued, vexed



tarimukmók [MDL]

to break into small pieces, to get smashed to smithereens




duck species


animal (bird)


to pivot, revolve, rotate, spin, turn around


movement / plural

tariwtíw [MDL]

to run with great haste (as when running away from s.o. or when running after s.o.)



tatagangtáng [MDL]

resonant, to resound (the voice)



tataguktók [MDL]

tsk-tsk, a clicking sound made with the tongue on the teeth



tatamó' [MDL]

to reprove, censure, to give s.o. a talking to



proficient, knowledgeable; to know,  (lexicalized derivation from ta'o 'to fathom give')

taya'tá' [MDL]

to change one's clothes every day




process of sculpturing

tinumtóman (tumtóm: to cook tinumtóman)

dish (made with the clam called bibí, tinumtóm wiping a liquid with an containing bamboo shoots, coconut absorbent milk, and seasoned with langkawás)

tinutóng a beverage made from the scorched (tutóng: layer of rice found at the bottom of scorched, seared, the rice pot singed)


 rice coffee tinutongán 'procedure of cooking with coconut (burning the meat of the coconut, then extracting the juice)'



plural food

food / drink

Appendix D: List of Bikol reduplications

tingáting [MDL]

a poor season; a poor harvest, scarcity in the yield of rice, drought

 crisis, drought


(billiards) a pole used to support the cue stick in place of the hand when attempting shots that are hard to reach

crutches and: Tiririt nang maya. 'The song of the maya bird (rice bird).'


titibák [MDL]


elephantitis (filariasis)


tiyáya' (naka-)

prone, supine, lying on the back


tutu'pán [MDL]


unknown tutu'pak 'going to erode, due to rain'

uróg-udóg / huróg-hudóg

a thundering or rumbling sound





waguswós [MDL] a swishing sound (such as that made sound of rushing waters in heavy by a whip or a switch of wood), ... volume wakáwak [MDL]

to let out more of the anchor rope or unknown chain (enabling a boat to drift further from its anchor)


wide open (as a door)

wangíwang [MDL]

gaping (a hole), to widen out (a hole) unknown


to brandish, wield, to wave (as a flag)

watáwat [MDL]

to be found over a large area or - flag (Tagalog) territory, to extend over or be strung - kind of tree (its blossoms have over a lot of ground clearing and disinfecting ingredients)


a mechanical swing ride at a carnival shake off urine after piddling

yagingyíng [MDL]

a clinking sound, a sound that resounds



ti-wangwáng naka-ti-wangwáng 'unintentionally left wide open (should be closed)'

waving, e.g. wave the weapon / arms



movement plant


Appendix D: List of Bikol reduplications

Lexical full reduplication lexeme

Mintz and Del Rosario Britanico (1985)

fieldwork Mattes (2005/2006)



a hindrance (a person who is always unknown in the way)


aloof, arrogant, haughty, pompous,... proud (pej.)



to fuss over, fawn over, show special wound, twined attention toward



respect, recognition

consideration, appreciation Napagal ko, ta ako kaya lang alang-álang ki Veronika. 'I am tired, but because of Veronika (I came).'



amulet, charm, fetish, talisman = antíng

amulet, talisman

apit-apit [only MDL]

waver, fall



arot-arot [only MDL]

very old (age)




to be carried away by the current = atóng [MDL]




to think deeply about, contemplate, to analyze




sound made by water as it enters through a large hole



badal-badal [only MDL]



plural/ movement

baid-baid [only MDL]

kind of tree


plant (tree)


gizzard, gullet of animals

body part


to keep feelings, anger, pent up inside; hide true feelings



to contemplate, speculate about


to gallivant, to roam around with nothing to do

easy going, lucky


shanty, shack, hovel

baróng-bárong 'shanty'


a thin membrane, amniotic sac



kind of tree




small, oblong-shaped deep-fried crisps made of flour and sugar

food, plural



Appendix D: List of Bikol reduplications


to bubble up, to gurgle, make a gargling sound

spasm in the stomach Nag-bu-burok-burok ang tulak ko. 'My stomach is churning.'

plural, sound


sea cucumber, beche-de-mer (butó' penis)




describing hanging (as a flabby not a very formal / a funny person breast) (positive)



to roam about, often said in anger, ... 




animal (insect), sound


to accompany / escort for a short distance

finger-pointing, accusing



name for S ekís-ékis 'zigzag'


a rattan chain used for tying monkeys

ka-galáng-gálang: respectable




gatáng-gátang [MDL]

to divide into portions, things that unknown will later be sold for the value of one gánta of rice; gátang [MDL] to buy from the hacienda; old measure (by a wooden container)




historically referring to s.o. with one  shoulder higher than the other who walks tilted to one side


white pyrites commonly found in mines

hoping/sensing for the coming of a long expected event


to wobble, to be lose (as a chair leg, a tooth)




small flies

plural, dimin., animal (insects)


kind of tree (Nopea philippinensis) gisók kind of tree (Shorea gisok)




describing the stillness and silence of the night

sound of a loose thing



animal (insect) plural

product of imagination, illusion


movement, disease

Appendix D: List of Bikol reduplications


to be almost the same age

 a group of people of the same age


tremolo; gobble



a temple built of bamboo and coconut fronds used for the celebration of prayers to the gugúrang (Bikol mythology) guláng age




kind of snail



cute and cuddly

 the feeling that you have when you see a very cute baby

smallness, emotion


to complain or crumble about, to nag,...

a mess



a mess, messy, rumpled



to mumble




trembling sound





worn to threads (clothes); disrupted (e.g. paper)


to shake

disrespectful addressing, provocation



kind of wasp


animal (insect)


to go for a good time, to enjoy oneself

 (from engl. happy)



1. describing or s.o. different lord it over, having control over from the way remembered in (landlord over a territory; gangs previous experience over a quarter in the city) 2. almost, on the verge of happening


feigning disinterest or dislike

hesitance Helé-héle bago, pero kere. 'He is hesitating but he likes it.'


referring to the movement of fish as they swim in the water




spasm, tic, twitch

(h)ibót-(h)íbot 'burning, pulsating pain'



to manipulate, to change around

trying to estimate


to move in the wind (trees)



to slighter, to wiggle, writhe,...

(h)ikól-(h)ikól 'writhe (like a worm)' movement




Appendix D: List of Bikol reduplications



very slow movement



to giggle, to snicker, snigger,...


movement, plural, sound


to regain one's composure, to calm down or relax



to chatter (teeth)


hulád-hulád [MDL]

to twitch, to barely move, referring to the limbs or body of s.o. hulád to wait

hulád-húlad 'laze around'


specific dish

hulóg-húlog 'dish with banana and root crabs mixed in coconut melk, sugar (afternoon snack)'


to look for attention, to call for attention



to be loose (teeth)




to groan, moan,...




restless, unable to stay put in one place

sound of shaking



to browse around; to walk around and look at things




to walk heel to toe = ikíd





going tours, loops (íkot-Jeep)



to be about to do




glow of a dying amber


to search for things that are needed or lacking

1. entering a room without greeting 2. estimate (amounts)


to whimper

very very low sounds


tree (small, used as a border for plots  of land, its seeds woven into bags and necklaces) Leucaena glauca; ípil: tree (tall, producing high quality timber used for posts and beams) Intsia bijuga


to secure or get things in preparation  for; to gather things together; to pack; to prepare


slowly, gradually; to go slowly with; unknown to do at a leisurely pace


sound, movement



sound plant (dimin.)


Appendix D: List of Bikol reduplications


to incite a fighting cock to fight by placing it near or pushing it toward its opponent

way of preparing the cock for the fight


kubá-kubá, kudá-kudá 'to throb (as the heart); to palpitate'

only kabá-kabá: 'anxiety'


to take things easy



rattling or clanking sound; to rattle, clank

 sound because of broken or loose; also metaphorical: old, useless stuff


plant disease (coconut)


to look around for

kalág-kálag 'familiarize oneself with', considering the superstitious believes kalág soul, spirit, ghost; Dai kang kalág. 'There is no compassion.'


to dangle, to hang freely, to swing back and forth = kálay




vine (located in the forest, climbing by means of tendrils, producing greenish-white flowers and a darkpurple fruit; Cayrata trifolia)




to do carelessly and in haste, to sound of active business move quickly and without direction


to flounder, to move clumsily and without proper balance

to succeed, but with big effort and difficulties


in a hurry or rush (kaphag as synonyme)

 also: doing things aimlessly



doing in the free time (interrupting it whenever there is a more important thing to do)


sound of gurgling; sound of liquid description of s.o. with a wagging bubbling through the narrow neck of tongue (s.o. who speak nonsense) a bottle

sound, movement

karé-karé / karíkarí

dish (consisting of oxtail or cow's leg cooked with long beans, banana heart, eggplant, white reddish, seasoned with onion, garlic and seeds, and stewed in a sauce containing ground peanuts and powdered toasted rice) mag-karí-karí 'to cook'



 karé native restaurant (informal, for simple people) = kari-han





Appendix D: List of Bikol reduplications


to look for with great care

to rope Karáp-karáp siya para sa paghánap kang pang-tuition kang aki niya. He did everything to find a way paying the school fees for his children.


to bob on the surface of the water



kawál-kawál / kiwál-kiwál, tawál-tawá

to hang loosely and move back and forth (as a loose trouser leg when s.o. walks)



to be suspended in the air dangling (also without sound)



to scratch the ground (chickens, dogs) = kayás

kayás red. unknown



to walk or move forward with short jerky movements (as if a part of the plants or skirt is caught in the crack of the backside) kayós to fornicate

scratching, also with sexual connotation

ki'ál-ki'ál / ki'áyki'áy

to walk in such a way so that the only ki'áy-ki'áy body is thrown off balance (such as a person with one leg shorter than the other might walk), to wobble (as a loose wheel)



to tremble



to waddle

waving of round hips



to wiggle (small insects, worms

involuntary movements of the muscles



puppet of a ventriloquist








walking with the heaps waggling



flickering, twinkling, shimmering, blinking,...

 e.g. dying amber

plural, movement


to keep secret

murmuring, speak to oneself (usually disappointed)



to move the lips (as when reading to mumbling o.s. or praying)



describing the movement of water as mix with the fingers it moves from one side of a lake, pond, canal or container to the other, as might be caused by a earthquake



to quiver, shake, shiver, shudder




to tear into stripes

 pick to pieces



Appendix D: List of Bikol reduplications


to shiver, shake




to walk with tiny and suggestive steps due to having the legs restricted by a tight dress (women)




descriptive of the work involved in the raising of children

1. mentally ill 2. a vamp


to wag the tail


spasm, twitch


wrinkles in the skin, to wrinkle (th skin)



to eat quickly



inconsistent, variant, ...



to creep or crawl on the skin (as an kamang/kamang-kamáng insect), to walk the fingers across the skin)


to be ashame or embarrassed



to be shaken (as water in a glass)

more than the capacity (for example movement, people in a room) plural


to tremble, to tremor, vibrate




to kick the legs in all directions during a tantrum




bird (similar to a turtle-dove)

(rare/endangered bird; appears also in Bikol mythology)

animal (bird)


regretful, disconsolate



bite-size snacks, such as crisps, potato chips, peanuts, etc.; to pick up, touch and examine things with the finger, to tinker with








movement movement



plural movement



head shaking 1. negation 2. illness


labád-labád [MDL]

to be bold, rash

to fly without direction (insects)


labó-lábo / rubó- 1. to work without let-up, to work only: labó-lábo rúbo fast and: additional roof at the house 2. a street fight, a free-for-all, mêlée, (e.g. above the window) rumble, brawl




disease (resulting in the swelling of the face and body)



Appendix D: List of Bikol reduplications

lagák-lagák [MDL]

sound made when swallowing liquids




tramp, vagabond, vagrant, bum, ...


lamó'-lamó' [MDL]

traces of blood, hair or feathers that remain on a knife blade after the knife has been used

dirty face of children (from food)

langá-lánga [MDL]

audacious, insolent, forward; to show disrespect toward, to be insolent with

to participate in a communication without being invited (old term)


to guzzle or gulp down



to slip through without hitting anything, to avoid getting hit magláos-láos kan urúlay 'to quibble, not to come to the point'


to stagger when walking




to fawn over, to show undue attention toward

beanpole; also: seesaw


Laság-laság [MDL]

bridge of the nose


body part





to distribute or apportion



lató-láto [MDL]

to repeatedly hit, beat or punch



lawá-láwa [MDL]

to think over carefully

cotton candy






layáng-layáng [MDL]

a makeshift roof or covering

kind of a bird

animal (bird)

limót-limót [MDL]

to pout



lináb-lináb [MDL]

sparkling, shining




cunning in a devious sort of way, ...

having relaxation

lipíd-lipíd [MDL] twisted (a tree, branch)


sound, movement


lisáy-lisáy [MDL]

1. to have a restless night's sleep, to 1. alimpasay toss and turn in one's sleep 2. to be unstable and move (such as a log or bamboo pole when one attempts to walk on it)


lisók-lisók [MDL]

to be in a state of dread or fear, to be unknown in a state of shock



Appendix D: List of Bikol reduplications

lúba-lúba [MDL] mask, to wear a mask (arc-)



plant (possessing leaves which may be eaten as a vegetable)

lumód-lumód [MDL]

to have great pity for, to be saddened lumód by s.o.'s absence or adverse situation



to dangle, to move gently from side to side (as leaves or branches)



luyóng-lúyong [MDL]

to sway or swing from side to side




to ask for mainly for the purpose of seeing if it will be given



 referring to variegated shades of color, particularly as changes in hue occur when an object is struck by light from different angles


to be aware of, to understand


muhá'-múha' [MDL]

to feel as if dirt or dust has fallen into or near the eyes, and show this by shaking the head or rubbing the eyes with the hand



feeble, weak (as after an illness)

s.o. who does not think, just follow

muláng-múlang [MDL]

to be indecisive



to fumble or muff





mungáw-mungáw describing s.o. who is still sleepy after waking up; dazed, groggy, stunned

 = muraw-muraw (also after a blow on the head or an accident)


to examine one's conscience; to meditate or reflect on

feeling with regret, hurt feeling



to wail while reading the Pasión

biblical song (which is sung nonstop for 24 hours on Ash Wednesday)


nawó'-náwo' [MDL]

to suspect s.o.


ngahól-ngahól [MDL]

to chew food poorly due to having no teeth


ngála-ngála [MDL]

to cry out, to call or shout

unknown ngalá-ngalá 'palate' ngala'-ngala' 'speaking without direction'


plural, sound

Appendix D: List of Bikol reduplications


body joints unknown (fig-) magku'á an kangaló'-ngaló'an 'wait for the right moment; look for the weak points'

body part

ngamá'-ngamá' (magka-) [MDL]

to neglect, to carry out a task poorly due to lack of concentration


ngará'-ngára' [MDL]

to shout to warn s.o. of approaching danger




an expression indicating that one wishes to do, but cannot

resemble Si V. ngari-ngari-ng Mount Mayon magayon. 'V. has similarity with Mout Mayon.'


ngaró'-ngaró' [MDL]

to eat quickly and with gusto


ngasál-ngasál [MDL]

to respond rudely or gruffly to one who asks for or tells you to do


ngayó'-ngáyo' / mayó'-máyo'

to beseech, implore

pray, reflect (old term)


referring to the up and down movement of the jaws when chewing” (see: ngípon; ngipngíp)


ngu'áb-ngu'áb / nguráb-nguráb [MDL]

to mumble, to talk to oneself, to nguráb-nguráb or ngurob-ngurob babble so as not to be understood (as when talking in one's sleep)



the sound of a growling animal




dragonfly (small) pádi' priest (Sp-: padre)

animal (insect)






to dangle





a booming sound such as that of thunder


to walk back and forth; to pass by a  place often, to pace, to strut, swagger

pasád-pasád na [MDL]

filled with rice (baskets)


patter (as the sound of rain on a roof) 


payók-payók [MDL]

tall (people)




to shake the head from side to side indicating negation

kiling-kiling or kuyong-kuyong



movement sound movement



Appendix D: List of Bikol reduplications



shaking before dying (s.o.)


pirí'-pirí' [MDL]

to form small bubbles when boiling (as rice)




ripped to bits; squashed to bits; unknown shredded / to squash, rip, shred pisáng [MDL] to cut sweet potatoes, to give s.o. a portion of cut sweet potato


pitó-píto (sl-)

the sound made by the last drops when urinating




fingerling, young fish


animal (fish)






to hang the head or nod the head in sleepiness or weariness

ngongo 'speak through the nose'




'suffocation of the car'


pusót-pusót [MDL]

filled with fish to the point where the unknown fish can hardly move (traps)



to paddle with the hands and feet



rabá-rába [MDL]

to glitter (as gold); to sparkle, glisten; to give off light (as a fire)


plural, light

ragá-rága [MDL]

a decoration placed around the edges unknown of woven mats

ragás-ragás [MDL]

gritty, sandy

ragáy-ragáy [MDL]

the sound of falling from above unknown and scattering on the ground

ramák-ramák [MDL]

a cracking sound (such as breaking unknown open a nut, stepping on branches), ...



to speak quickly and with irritation



rasí'-rási' [MDL] starkly white, extremely fair




to go around town, to stroll, walk around


plural (distr.)

rawá-ráwa [MDL]

gold filigree work


rawá-ráwa [MDL]

to eat as a matter of courtesy or etiquette and not because one is hungry


rigák-rigák [MDL]

the sound of a croaking frog





Appendix D: List of Bikol reduplications

rigó-rigó [MDL]

to say loud and fast, mumbling in such a way as to not make s.o. understood



ri'ík-ri'ík [MDL] the sound of a stick cracking



rimáng-rimáng [MDL]

to feign, to pretend to so



rimá-rimá [MDL]

to be dirty, to be covered in dirt or mud



grotesque, hideous, horrible, lurid, morbid, ...



murmurs, the sound of voices carried ragubnob on the wind, ...


rimóng-rimóng [MDL]

the sound of many church bells ringing at the same time


riníb-riníb [MDL]

the crackling sound of a fire, to make unknown this sound


riník-riník [MDL]

the patter of rain on the roof


ritá-ríta [MDL]

to say in a gruff or irritable way unknown

sound, emotion

rití'-rití' [MDL]

cockroach (in the fields)


animal (insect)


to spy on


rubó-rubó [MDL]

to bicker, argue, quarrel (two or more people)




buzzing sound (mosquito)

bicker, quarrel rubó-rúbo/labó-lábo work without letup, work quickly


rumá'-rúma' [MDL]

describing a mixture of different kinds of rice in the grain bin



rutób-rutób [MDL]

the sound of gun or rifle fir, the popping or bursting sound as when things are burned in a fire



rusóp-rusóp [MDL]

the sound of many fish moving or jumping together in the water




to blur out in anger


emotion, sound

ruták-ruták/ rutók-rutók

the cracking sound of body joints




the crackling sound of a fire, to make this sound


Appendix D: List of Bikol reduplications

rutós-rutós [MDL]

the sound associated with the straining wooden joints which are lashed together indicating that they are weakening and may eventually break


sa'án-sa'án (magpa-, ipa-)

to look for an excuse (for a quarrel, to leave a particular place)

'anywhere' (Tagalog)


to pretend, to bluff, feign, pose


sakól-sakól [MDL]

rough (as a surface)




to weave bamboo strips, rattan, palm  fronds

plural, movement

saní-saní [MDL]

to take or finish everything; to leave unknown nothing for anyone else



to lisp



kite (American; ordinary kite with a tail)


corner shop




sawí'-sáwi' [MDL]

the sound of air passing through the lips as when one whispers or reads quietly



sawóng-sawóng/ sahóng-sahóng

to disturb s.o. by coming and going; to kibitz



bird (inhabiting the forests, whose call was once seen as an omen)


animal (bird)

sayág-sayág [MDL]

fine and neat in appearance, neatly attired



to glide




cormorant (bird)

kind of insect

animal (bird/ insect)

silí-silí (slang)

descriptive of people, animals, vehicles, etc. that move quickly, weaving in and out


to sniff (animals such as dogs)



sipí-sipí [MDL]

bashful, embarrassed



to swish back and forth (as water in a bathtub or a glass)




the sound of simmering




fish (tiny, used in the making of fish  meals); siróm ant (small, red, biting)



animal (fish)

Appendix D: List of Bikol reduplications


to shower attention on; to fuss over







winds that are said to blow for nine days



siyám-siyám (slang)

descriptive of the sound when one starts to urinate




to rush, to hasten in doing








the sound made by bubbles breaking the surface of the water; the sound made by a drowning man; glug-glug


the sound of gasping or panting; the sound of gurgling



to stack (as boxes); to pile one on top of the other

ta'ád-ta'ád [MDL]

to speak one's mind




tabó-tábo [MDL] to fly about in the wind (cinders, dust)



sweet (made from a mixture of flour  and mashed sweet potato, formed into balls and fried, usually served skewered on a stick) tabóg tree (producing a fig-like fruit)


tagáy-tagáy [MDL]

a sandbank extending from the shore unknown

taká-taká (Tagalog)

surprising, amazing

takóy-takóy [MDL]

to tremble, shake


taló'-taló'/ tiló'-tiló'

loose as a post in the ground, a knife unknown handle)


to pout

 movement/form of the lips


tanák-tanák [MDL]

referring to the appearance of tiny thread of water formed when drops of water fall in rapid succession


movement, plural


castor oil plant (Ricinus communis)



tangás-tangás [MDL]

to carry out household duties


tangóg-tangóg [MDL]

to shake (the head, due to age, illness)





Appendix D: List of Bikol reduplications


shortly, in a little while


to tap gently; to pat on the back (as for comfort or congratulations)


expression, visage unknown taróng [MDL] a black or white mark or patch

tawá-táwa [MDL]

orchid (Grammatophyllum scriptum) unknown



tibís-tibís [MDL] to drip rapidly or repeatedly



rice bran extract

brand of a vitamin

ti'ám-ti'ám [MDL]

to be left with a bad taste in the mouth



to move the lips (as when reading silently to oneself)

and: fingerfood

movement / plural


to shower, drizzle


tingá-tínga [MDL]

to act happy or merry due to the effects of an alcoholic beverage

tingá'-tínga' short-winded

tirá'-tíra' [MDL]

to ease (a heavy rain); to stop raining after a heavy shower or downpour; to clear up

kind of candy (sugar and flour)

food/ plural

titá'-titá' [MDL]





babbling, blah-blah

speaking without direction


tiwá'-tiwá' [MDL]

ripped, rent, torn (clothes)


tubóg-tubóg [MDL]


kind of food (cue with cassava, sweet potatoe or rice paste, fried)


tubó'-túbo' [MDL]

tree (used for making charcoal)

tubó' 'sugar cane' (*tubó'-túbo') túbo' 'profit / increase' túbo'-túbo' 'gain a lot of profit' túbo 'pipe' - tubo-túbo 'small/imitation of a pipe'


tubó'-túbo' (magka-)

to be almost the same age (two people)



false testimony, perjury




prolong, merge, compose

tukóng-tukóng [MDL]

very tall (a person)


a human-like creature which can stretch or shrink to take the shape of whatever it is next to [Bikol mythology]





Appendix D: List of Bikol reduplications


eating places where prepared food is restaurant for the simple people displayed in glass showcases; one points to the food one wants, hence, the name (Tag. turo' 'to point')



to pick on s.o.; to insult, to abuse

tuyá'-tuyá' pick on s.o. seriously uyát-uyát pick on s.o. joking



 (Naga dialect)


to scorn, disdain, to take s.o.'s name in vain

sweet rice dish


upá-úpa [MDL]

bird (the size of a hen)


animal (bird)

uród-uród [MDL]

to tickle a child in order to make him unknown laugh



to moan, groan



uróg-uróg [MDL]

to compose or recite couplets, verses, romances


urók-urók [MDL] to spill (water when it suddenly gushes forth as it is being poured from one container to another)



to lope

 (Bisayan dialect)


usáb-usáb [MDL]

covetous, greedy

animal's chewing


utá-utá [MDL]

to take or pick things up one at a time



utáy-utáy [MDL] to play joke or trick on

a little amount (Bisayan)


mosquito larva

gully ball


to make a fool of, to mimic, mock, ... 

animal (insect)

utóy-utóy [MDL] hair at the nape of the neck


body part


mint (used for treating wounds)



uyát-uyát/ tuyá'-tuyá'

to pick on s.o., to insult, abuse

wa'íng-wa'íng [MDL]

to say in an annoyed or irritated crazy manner



a squall

animal (sea animal)

sea animal (similar to a snake)


Appendix D: List of Bikol reduplications

wangís-wangís [MDL]

to look at s.o. with irritation or  (informal term) contempt while muttering under the breath

wasáng-wasáng [MDL]

to attack violently (animals), to do in a furious rush (humans)

wasá-wása [MDL]

to flow or pour out (rice from a sack, unknown blood from a wound)


wikí-wikí [MDL]

to shake the hand (as after touching unknown hot, after getting a finger caught in the door)



hearsay, rumors




to walk around without purpose or direction



yabó-yabó [MDL]

to get a whiff of; to catch the scent of


yakáb-yakáb [MDL]

to open and close the mouth (a crocodile when eating)

(not used)



the sound of running footsteps




very fat



yuwó'-yuwó' [MDL]

to be annoyed or fed up at seeing done slowly or apathetically



fieldwork Mattes (2005/2006)


disorganized, without direction


Echo words 1. with initial consonant mutation lexeme

Mintz and Del Rosario Britanico (1985)


the tip of a grain of rice



shaking / trembling of an animal / person before dying



mahogany (tree)


nginí'-kíni (mang-) [MDL]

to shudder



rapák-dapák [MDL]

the sound of running footsteps


ra'án-da'án [MDL]

to hold a grudge against s.o.



rawíg-dáwig [MDL]

to hang suspended in the air, swaying from side to side




Appendix D: List of Bikol reduplications


unnecessary talk, uncalled for verbiage or chatter; describing the speech of one who is verbose, ...

poetic speech, poem


ri'íg-di'íg [MDL]

to say harshly or gruffy


sound / emotion

ripáy-dípay / to be groggy (referring to difficulty rapáy-dápay 'to be drunk' rapáy-dápay / in walking straight); to wobble, reel rupóy-dúpoy


riwág-díwag [MDL]

to move from side to side, ...




to regain one's composure, to calm down or relax

ru'áy-dú'ay [MDL]

1. to stagger, totter; to sway from side to side 2. to beat around the bush; to speak in a roundabout way



rukáy-dúkay [MDL]

to search or rummage through



to walk unsteadily



descriptive of a field plowed or planted in a haphazard way, and not in discernible rows; also of the sea when the tide runs against the prevailing wind

 and: state of rotting (old, fragile, decaying) also: person who is weak --> sufferings, can be easily broken (emotionally) by difficulties

rupót-dúpot [MDL]

clever, ingenious, talented


ripáy-dípay / to be groggy (referring to difficulty rupóy-dupóy / in walking straight) rapáy-dápay


ruwág-duwág [MDL]

reasonable, moderate, sufficient, ...


ruwóngduwóng [MDL]

to nod in weariness or sleepiness (the head)



 because of load of fruits (ruyóng bent; the branches of trees usually when laden with fruit)


ruyón-dúyon / to move (the branches of trees as ruyóngwhen caught in a breeze) duyóng [MDL]



Appendix D: List of Bikol reduplications

2. with internal consonant mutation baráy-basáy

to go back and forth, go in and out (annoying movement, for example performed by a child in order to attract attention)

1. walking idly, just anywhere without goal 2. go in and out (without purpose)



to boil, be boiling, to bubble (e.g. soup, porridge, lava, ...)

movement, sound


o gurgle, gush out (e.g. a sparkling source)

movement, sound

burík-butík [MDL]

full of spots, dots



sound of falling water


garánggasáng [MDL]

1. jagged 2. coarse (i.e. what remains in the sieve)



rough (surface)


giríng-gitíng [MDL]

a scalloped edge or border

 also: being sensitive


rubbing (e.g. the nose, out of annoyance)

plural, movement

gurónggusóng guró'-gudó' [MDL]

tremolo, gobble

sound, movement


to roar with laughter



haráp-hasáp [MDL]

rough, to become rough


hirík-hitík [MDL]

to giggle, to snicker, snigger,...

sound, movement

huríng-hudíng rumor

by rumor



kaláng-kagáng jangle

sound similar to kalíng-kagíng





jingling sound (as of coins); tinkling  sound (as of a small bell)



rattling sound; scratchy sound



dull sound (wood, carton, ...)




sound of falling water

kalóng-kagóng [MDL] kaló'-kagó'

to shake (as water in a glass, a milkshake or cocktail)


Appendix D: List of Bikol reduplications

1. kalít-kagít 2. kalít-kalít

 rattling sound; scratchy sound vine (located in the forest, climbing by means of tendrils, producing greenish-white flowers and a darkpurple fruit; Cayrata trifolia)



to shake (as water in a glass, a milkshake or cocktail)



1.restless, fidgety 2. wavering, wobbling 3. sound of feet on a wooden floor

movement, sound


bird (small with a gray black and white breast)

animal (bird)

karóg-kadóg [MDL]

the sound of running feet

sound (shaking)


karós-kadós [MDL]

to do hurriedly and without much care

movement in haste


kipáw-kipáw [MDL]

to have only the upper part of the head visible (as when walking through tall grass)


kiráy-kisáy / kuráy-kusáy / kisáy-kisáy

to kick the feet out repeatedly (as a child throwing a tantrum)


kiríg-kidíg [MDL]

to take short, quick steps

shivering (because of cold)


kirí-kisí [MDL]


grind, rub (eyes, feet, etc.)



crackling or sizzling sound




kutáb-kutáb to throb or beat (the heart), to palpitate

heart palpitation (stronger than normal) movement (intensive)


to lift the skirt by tucking the hem in to kick the feet out repeatedly at the waist (as in preparation for wading in water)


kurób-kurób / kurób-kutób

the sound of muffled roaring such as only kurób-kutób that made by a car engine being started; the sound of distant thunder


kuróg-kudóg [MDL]

the sound of stamping feet, a drum roll


nervous, trembling

kuróng-kutóng to shake (the head due to age or [MDL] illness)

unknown kuyong-kuyong: head shaking 1. negation 2. illness


clucking, the call of a hen to her chicks


to scrub clothes



Appendix D: List of Bikol reduplications

palángpagáng [MDL]

to awaken in fear or anxiety



palás-pagás / pulós-pugós

to move carelessly without particular attention to where one is going or what one is doing


palód-pagód [MDL]

the sound made by running feet

sound, movement


to stamp the feet repeatedly (as when a child throws a tantrum)



a pattering sound (such as that made thunder claps, lightening, sound of by rain on a roof) guns


paráng-patáng present; before or in front of [MDL]


parík-patík [MDL]

robust (children)


a galloping sound


sound of fire crackers


parok-patok piráw-pitáw

to blink the eyes (so as to remove that has fallen into them)



to spatter, splatter, to flip (as a fish  out of water); to shake dry (as a wet dog); etc. ...


piríng-pitíng [MDL]

to shake the head (as when indicating negation)

getting it self loose


piríng-pitíng [MDL]

to twirl or spin around




to roll between the palms of the  hand or the fingers






poróng-potóng / paratangpatang

to lament, resisting commands


a galloping sound




to be irritated or annoyed


purók-pusók /burók-busók

sound of falling water




sounds of explosives



cranky, grouchy, grumpy, irate, ...

 annoyance, exaggeration, emotional explosion (pusóng heart)


Appendix D: List of Bikol reduplications

rabáy-dábay [MDL]

to walk around from place to place, accomplishing nothing




interfering, butting in a conversation

saláy-sagáy [MDL]

describing a woman's voice when she yells in an argument

talá'-tagá' [MDL]

to hack, to slash; to make cuts in the  side of a tree for the purpose of collecting resin or sap


uróg-udóg / huróg-hudóg

a thundering or rumbling sound


 waríng-wasíng waving wuróng-wusóng tear one's hair


waríng-wasíng 1. to bite meat away from the bone, / wuróngto hold in the mouth and shake wusóng it vigorously from side to side (dogs) 2. to strike out in all directions



Productive partial reduplication simplex




Reduplicated word form ba~batód [MDL]

Mintz and Del Rosario Britanico (1985)

fieldwork Mattes (2005/2006)


to give s.o. a heavy blow with a fist, stick, or stone

and: b-in-ató ' hit by a stone'

CV-:deriv. (impfv.?)

barikíg describing a feeling b-ar~arikíg of awkwardness shown by a particular posture or stance; crooked, e.g. a picture on the wall



-Vr-: plural


referring usually to a b-ar~árot woman who lets herself go, not taking care of her figure or looks, ...


barót 'sloppy dresser, bad manner of dressing' --> bar~arót (pl)

-Vr-: plural


back to front; turned bi~birík-an around, inside out

a large spindle used for the spinning of cotton; knob, dial

and: to be turned; turning the wheel, furl

CV- (-an): deriv.




a hen that produces many eggs


CV- (-an): deriv.



nag-b-ur~uso sinda


they dived

-Vr-: plural


Appendix D: List of Bikol reduplications


double, twice as much

pag-do~dóble reduplication


to grief, lament, suffer


grief, lament, sorrow going through the punishment



du~duwá = duwá-dúwa

only two

du~duwá 'only two' CV-: numeral duwá-dúwa '2 each' limitation duduwa na sana ang bado ko 'I have only two dresses left'




a household spirit carried around on the person, capable of granting the owner's requests (Bikol mythology)

gu~guráng 'ancient, antique'

gumód- to complain or gúmod crumble about, to nag, ...


complaints, grumble gubód = g-ur-ubód 'involved in an intrigued problem; unkempt hair; confused thoughts'

-Vr-: plural

hánting to hunt, go hunting, h-in-anting 'he was searched'


referring to groups roaming around looking for a fight

-Vr-: plural


to quarrel with one another; to have a misunderstanding; also: the quarrel (“may iwal sinda”)


argument, the same, but 'seems -Vr-: plural controversy, quarrel, to be ongoing' ... ma:-iriwal (fut.)


to put or place


container, receptible -

lampaw to jump over [MDL] r-al~ampaw



CV-: deriv. CV-: deriv.

CV-: deriv.

CV-: deriv.

su sira nag-Vr-: plural of ralampaw su sira actors 'the fish jumped out'


to tether, hitch, to tie la-lawíg-an an animal to ... (lit-)


CV-(-an): deriv.



only five


CV-: numeral limitation

lukón [MDL]

to conceal or hide, to lu-lukn-án / deny the existence lukón-lukón of

the curvature of the hollow of the knee

CV- (-an): deriv.

li-lima (limalima)


Appendix D: List of Bikol reduplications


chicken, fowl; ma-manók manokón an matá 'describing eyes that stare somewhat cross-eyed'

describing s.o. who is night blind

farm of chickens/ CV-: deriv. chickens in abundance/owing a lot of chicken nag-ma-manokmánok = particles flying out of a fire ma-manok-ón = seemingly crosseyed (“village term”)


to breed, reproduce (animals)




learned, educated

nu-nu'ód [MDL]

to imitate words or actions

simplex: 


to be surprised


amazing, astonishing 



as a verb: pointing with the lips as a noun: form of the lips while pointing

CV- (-on): deriv.

no: lips (mouth: ngimot)

ngu-ngusó'-on describing s.o. with a large mouth; mag-on: to reproach, admonish


- clapping - wing

p-ar-akpák [MDL]

the sail-like protuberance on the back of certain lizards

clapping sound plural

-Vr-: plural


broken into pieces, fragmented, shattered, smashed


fragile; mag-: to break (as waves on the shore); mag- / -on: to shatter, smash, fragment

 also ma-pasá'-on

CV-: deriv.

pinggán a plate


cupboard, pantry

unknown cupboard: aparador

CV-: deriv.



dismissal (as from school at the end of the day)

home coming of -Vr-: plural family members who live not at home (e.g. students), for a family festivity) Fiesta na naman, sinda ma-pu-pu-rulí-an. 'There is a family celebration again, they are coming home!'


winged beans

to return, go home

pagulón roller (farm g implement)


CV-: impfv.?

full: plural

Appendix D: List of Bikol reduplications


noise; clamor, din, hubbub

riribók [MDL]

to do slyly or by stealth

pl. of noise




soaking wet

rupít-rupít ka! = 'you full: 1. are soaking wet' intensive, 2. or: a little bit wet diminutive (soaking wet: rupitón-on)


soup, broth, the water of coconuts

s-ar-abáwanan (fig-)

the focus or center of attention



property, own


to fend for oneself, to live independently or have an independent means of living


to depart, to leave the station (trains), to replace, to change, ...

saró-salída / saró-s-aralída

to vary, to fluctuate, being replaced to alternate, to work continuously (e.g. a in shifts, etc. guard of a movement), replace each other, people taking turns



saonly one saró'/saró'-sá ro


CV-: numeral limitation





nine each

tig-CV-: numeral distributive



si-siyám / siyám-siyám

only nine


CV-: numeral limitation


(be) alone


solitary, alone

state of being alone; ang pag-so-solo ko dahil sa mag-susurat 'I am alone, because I am writing.'


surrender, bow down

su-súko [MDL]

to cover up for by lying; to deny actually done



to separate


separation, divergence, segregation, ...

referring to a discussion, s-ur-úway argument

-Vr-: plural





sleeping of many

-Vr-: plural



ta-taná' [MDL]

to taste


CV-: impfv.


Curu-/ -Vr-: plural

Appendix D: List of Bikol reduplications


to laugh

ta-táwa' [MDL]

to laugh at

Fut. of to laugh


man, human being

tatáwo [MDL]

wooden figures or statues made in honor of the dead




tig-t-ur-uló tu-tuló / tulótuló

three each to divide into three, to send three at a time



accustomed with, familiar with

tu-tu'ód [MDL]

to be standing, to Fut. of simplex, will CV-: impfv. stand, to be on one's obey, believe feet

úlay (maki-, ma- -ka, ka- -on)

to converse with, to speak to or talk to, to discuss with s.o., ...

- ur-úlay (pag-, ka-, -) - mag-ur-úlay - úlay urúlay

- conversation - to come to an agreement - agreement, compromise, deal, understanding, concord

- urúlay plural -Vr-: plural; (formal talk among full: elders) diminution - mag-ur-úlay to come and talk - ulay-úlay formal talk between small groups (important topic); only in rural contexts, elsewhere: meeting, forum, etc.


to applaud for, to clap for; beat with the fist


applause, clapping

plural, fist fight

-Vr-: plural



tig-w-ar-aló wa-waló / waló-wálo

eight each, only eight


CV-: numeral limitation


gone, nothing



warara' = they are gone (s.o., that is searched for)


to dissipate, wreck

w-ar-arák / wasák

scattered, strewn around, smugged, smeared



CV-: impfv.

CV-: numral limitation

-Vr-: plural

Appendix D: List of Bikol reduplications

Full (and Curu-) reduplication simplex

abót [MDL]


to arrive, reach

reduplicated word form 1. abót-abót 2. abót-ábot

Mintz and Del Rosario Britanico (1985)

fieldwork Mattes (2005/2006)


1. ill-humored, illnatured, ill-tempered 2. intermittent, recurrent, spasmodic, sporadic

abót-ábot: 1.occasionally attacking, 2. sporadic, come and go abot-ábot ang higus 'inconsistent production'

dimin. plural

ági [MDL]

prints, tracks, line, to pass, to go a particular way; drop by, come, pass by


story, event agí-ági kan buhay 'biography, autobiography'

life history, story agí-ági = abot-ábot

lexica-lized (plural)

agóm [MDL]

spouse, mate, husband, wife


common-law husband / wife

just playing a husband now: living partner, ka-living

dimin. (imitat.)




every day, daily

Curu-: plural

álon [MDL]

(small) wave


- that flows with the wind, gentle waves



lead; ámo 'boss' (sp.)


to act like a leader

amó-amó-han 'pretend to be the boss'

dimin. (imitat.)

ánas [MDL]

all, completely


all in all

anás-anás 1. most 2. all (overgeneralized)

1. dim. 2. int.

anó [MDL]


anó-áno [MDL]

related to each other anó-anó 'what' (when plural a plural answer is expected) Ano-anó ang k-in-ara-on mo? 'What (pl.) did you bring?'

ánod [MDL]

to bob on the surface anód-ánod of the water; float with the current



describing a person who has no own opinion (a rubber stamp / yes-man)


Appendix D: List of Bikol reduplications

anóm [MDL]


1. anómanóm [MDL] 2. a-anóm / anóm-ánom

1. divide into six, six 1. six each 1. plural 2. at a time, six by six or int. 2. only six 2. six, which is more than appropriate (e.g. Anom-anóm ang agom ko. 'I have six husbands')

apát [MDL]


1. apát-apát [MDL] 2. apát-ápat

1. divide into four, etc. 2. only four

cf. anóm

1. plural 2. int.

aki [MDL]


aki-aki [only MDL]

pretend to be child, treat s.o. as a child


dimin. (imit.)


ouch, cry in pain


describing s.o. who always shouts ouch

interjection, which inter-jection expresses that the speaker feels himself addressed by a critique

aro [MDL]

console a crying child

aro-aro [MDL]

caress a child


dim. plural?

átom [MDL]

conscientious, to concentrate on, to give intensive care on; pejorative: to be always interfering


a busybody (=para-átom)

plural (habit.)


new, change


to be in a state of change


full/ Curu-: plural (cont.)

bágay [MDL]

1. things, stuff, item, bagáy-bágay matter 2. to fit well

odds and ends, bits and pieces, things

and: ideas



hut, house



red. unknown, interpretation: small

Curu-: dimin.


naive or curious


describing s.o. who is curious


plural (habit.)


to go berserk, run buróamok; balingháwon bad temper, blowing ones top

describing s.o. who is balingháw

bara-balingháw = Curu-: plural habitually bad (habit) temper/blowing ones top



to come and go, back and forth,...

 plural Balik-bálik ka sa Pilipinas. 'You have been to the Philippines more than once.'



Appendix D: List of Bikol reduplications

balós [MDL]

to take revenge on, to get even with, ...; mabalos 'thank you'



to reciprocate


to go across, go over buró-balyó to the other side, to move, to transfer,...

to take turns, work in shifts

buro-balyó/barabalyó person who changes money / to keep on changing

Curu-: plural (cont./habit.)


evening, night


every evening

Curu-: plural


to divide in half, to halve, bisect, to share, ...

1. b-ar-angá'bánga' 2. bangá' -bánga'

1. to divide into 1. share equally (pl.) plural pieces, to section, 2. one half each segment, subdivide 2. half-and-half, 5050%


to hinder


to fail the hands or wave the hands about, to swim freestyle


trifle with


helter-skelter, out of barág-barág 'chaos' place

bári' [MDL]

break, fracture


disjointed, describing that bends at the joints, ...






Curu-: dim. (imit.)


to read


to skim or thumb through

to read just for pleasure, in the leisure time


basá' (Tag.)

wet, damp, soggy


soaking wet, sodden basá'-basá'


describing thoughtless or unplanned; básang 'inexact opinion / assumption'


describing desecration, to treat intens.? without substance or without respect worth, profane, vain Dai mo pag-basángbasáng-on an ngaran kan diyos. 'Don't put the name of the Lord in vane context.'




very full



whenever, as long as; just


just like that, so easily

easy to get, cheap, worthless


1. without value lexicalized Bará-bára ang isip mo. 'Your way of thinking is indifferent.'; barabára-ng decision 'an indifferent decision' 2. to block s.o. (in a conversation) plural?



Appendix D: List of Bikol reduplications




plant, with small bean-like pods

batók [MDL]

to bark at; bátok


describing a dog that  is always barking


báwi' [MDL]

1. to take back what báwi'-báwi' you give, or go back on what you say [MDL] 2. to exorcise, to drive out evil spirits

describing s.o. who takes back what he gives; an Indian giver

continuous giving and taking


bílog [MDL]

1. circle, disk, ring, sphere 2. frozen, clotted MDL: whole thing

hoop, small balls, bearings; very plump or round

bilóg-bilóg 1. several rounded things 2. semi-round

1. plural 2. dimin.

bítay [MDL]

to hang s.o. to death; bitáy-bítay to hang (from the [MDL] top)

to hold on to and swing back and forth that is hanging plural? from the top (e.g. the mosquito net)

bitín [MDL]

1. to suspend or hang s.o. up; 2. too short (e.g. a skirt)


to swing back and forth

bitín-bitín intens.? 1. which is too short, but only by a hairbreadth 2. suspenseful 3. state of 'you want to have more, but it's finished' (food, story, that ends at the point where it gets really interesting)


complete, entire; different things which are kept together (e.g. the content of a handbag)


 to do everything associated with a particular act or give everything associated with a particular object





ground lizard

unknown, interpretation: toy crocodile; personification of crocodile

Curu-: dimin. (imit.)


to return, to go or come back


to come and go, to go back and forth

come and go continuously

Curu-: plural


to tie a knot in; to untie, ... also: to unit


knotted up

b-ur~ugkós: many things/everything united

Curu-: plural




crazy (but only joking)




an imitation of a banana

dimin. (imit.)

Appendix D: List of Bikol reduplications

búhay [MDL]



personal experiences way of life MDL: tierra suelta



hill, mountain


knoll, small hill

and: na-bu-bukid-búkid ' is curling up'



node (as in bamboo), bukó-bukó knot (in wood)

joints (of the body, ankle)



bump, lump


a large bump or lump

1. a small bump 2. many lumps

1. dim. 2. int.;pl.





almost round


búlan [MDL]

moon, month


1. every month, 1.  monthly 2. now: moon-moon 2. an extra or reserve (!) player, describing s.o. who participated in a game just to even up the sides

1. distr. 2. lexicalized




to feign deafness

a little bit deaf, hard of hearing bungóg-bungóg-an 'pretending to be deaf'

dim. (imit.)


1. whisper 2. whirl water movement; also: moody


to whisper

ang tubig buróngbúrong 'whirlpool water' Burong-búrong ang uta' ko. 'I am confused.' buróng-buróng-an 'whirlpool'



to push


drawer with wheels, so that it can be pushed easily nag-bu-busol-busolan 'pushing continously' busól-busól ang lamesa 'the table is moveable'





to feign blindness

a little bit blind butá-butá-han 'pretending to be blind'

dim. (imit.)

dálan [MDL]

path, trail



a temporary way, path



Appendix D: List of Bikol reduplications

dámay [MDL]


damáy-dámay synonym


accumulation of damót-dámot to save, accumulate many things; savings


audible, to hear

1. dangógdangóg 2. dangógdángog

1. hearsay, innuendo both: hearsay 2. clearly audible


beaten, defeated, lost; looser


to bully, boss around, to discriminate or be prejudiced against

underdog; s.o. who plural makes things to (habit.) satisfy others (a husband for example)

dará [MDL]

to bring, take, to carry


baggage, luggage


a bit, drop, fraction

1. dikít-dikít 2. dikít-díkit

1. little by little 2. very little or few

diit-diit 'little by 1. distr. little', 'very little' 2. dimin. (no difference by the accent)

díla' [MDL]



1. tongue of the shoe 1. everything that has 1. dim. 2. fried, tonguethe shape of a tongue 2. pl. shaped sweet made 2. blow job movem. of glutinious rice flour 3. MDL: wiggle the tongue


poke one's tongue out at s.o.

diwal-diwal [MDL]

MDL: poke one's tongue out at s.o.

speaking without direction


double, twice as much


repetitive/to reduplicate

doble-doble = dorodoble = d-or~oble 'many times over'

Curu-: plural





somewhat wet, soaking wet



to shed, molt, to change skin or feathers


to accompany/escort finger-pointing, for a short distance accusing


breeze, wind


to take in some air, to refresh oneself

dúwa [MDL]


1. duwá-duwá 2. du~duwá = duwá-dúwa 3. duwá-dúwa

1. to divide into two two each 2. only two 3. to hesitate over, be dubious about,...


giving each other sympathy damay-dámay kita 'we support each other'


a little bit of saving

dimin. dimin. (2. intens.)

lex. (pl.)

plural (distr.)

Appendix D: List of Bikol reduplications


cradle, hammock


to rock or swing in a  hammock

gabós [MDL]

all, everybody


all in all, altogether

gapó' [MDL]

stone, rock, boulder


pebble, small stones small stones in the rice

dimin. (pl.)


brim, edge, border, ...


almost, to be on the verge



to do, work on, to perform


 actions, behavior, conduct, customs, ...


like, as


cheerful, pleasant, lively, vivacious

to imitate, to mock s.o.


clew, tuft



clew, tuft, which can dimin. still be unraveled




bumpy in texture

(sound of flapping?)

guló (Tag.)



cute and cuddly

cuddle (babies)


not, to dislike, to refuse


to feign dislike

unsure, refusing, but wanting it

dimin. (imit.)

hagás [MDL]

1. frail 2. greedy


describing s.o. who is careless in dress, speech or actions

strong wanting, graving



to wait


to mill about

1. wait for a long time 2. wait just a short time

1. intens. 2. dim.


to mix


a refreshing snack, mixing consisting of various ingredients


to inhale, exhale, breathe


to pant, to be out of breath, be winded





play house

temporary shelter; or: dimin. initial stage of a (imit.) settlement of a newly married couple

háwa' [MDL]

scarcely, hardly

hawá'-háwa' [MDL]

too much, unnecessarily excessive

contaminating; spread of contamination


a look


describing seen occasionally glancing 1. dimin. very clearly 2. int.?


and: sum up pl., intens. everything G-in-abós-gabós ko ang presyo 'I sum up the prize.'




plural (lexicalized)

Appendix D: List of Bikol reduplications


treat very carefully e.g. carefully shell prawns with the fingers

himáy-hímay [MDL]

to rest after eating or = hímay after a period of also: in an abstract work, to relax sense (to think about very carefully)


compassion, pity, woe


mercy, pity



to share


to divide land

to divide equally, fair


press (birth)





to move to a new place


nomadic, describing unknown s.o. who moves from place to place

plural (habit.)


opinion, judgement, huná'-húna' view; intuition, though

to contemplate

a guess



hard working, industrious


to consider, think about, reflect on





a wide grin or smile

hiyom-hiyom 'smile'





gentle sound of the plural breeze (cont.) huyop-huyop-an cave 'name of a cave where the wind resonance can be heard'


dog, canine


(fig-) to trail along behind s/o; to act like a dog following his master

dimin. (imit.)


tail íkog


to tail, to follow closely behind




first in a series; initial


pioneer (ka- -i)




scarce; to need, require






wiggle the bottom



wiggle the bottom (playing children)

dimin. (imit.)



Appendix D: List of Bikol reduplications


to avoid Nag-iwós siya sa peligro. 'She avoided risks.'

iwós-iwós [MDL]

to be unable to sleep avoidance, act of due to fears or avoiding apprehensions Nag-gibo siya kaiyan para sa pag-iwosíwos sa problema. 'She did it in order to avoid problems.'


that, those; to fornicate; a woman who is easy (vulgar)


to come and go, to go back and forth

simplex: dem. pronoun iyán-íyan = balikbálik



hunchback (congenital deformation)



kubá-kúba 'bad posture, bowed back (temporary hunchback)'

dimin. (imit.)


two-toned, half-and- kabánghalf (color, kabáng contents); mismatched colors


uneven, washed-out colors, due to stain


horse and: flat iron

karo~kabáyo play horse, rocking kuru-kabayo horse; mule, donkey, toy, small horse sea horse

Curu-: dimin. (imit.)


to attach, fasten; to clamp; to fix or affix; 2. concubine


to join or hold hands to grid very in a group strong (e.g. the handhold in the Jeepney)

plural; inten-sive


canal, ditch, trench, gutter in bowling


small ditch or canal,  furrow, groove, rut; channel, drain, tread of a tire



a temporary shelter, kará-kamálig a hut or shed consisting of a roof / karóand posts, but no kamálig walls, built in the fields or along the coast to protect harvested crops, boats, etc.


kamálig: hut Curu-: kará-kamálig / kuru- dim.(imit.) kamalig: very small and highly temporary, provisional shelter (only for harvest time, or only for the wedding party)

Appendix D: List of Bikol reduplications


we (excl.)


among ourselves

 intens.? Kami-kami lang – kamo-kamo man. 'We mind our business – do you mind your own business!' and: invite somebody, with a gesture (informal and very personal) Pig-kami-kami niya su mga tawo na magboto. 'He was drawing people to vote.'


you (pl.)


among yourselves

 see kami-kami




kamot-kamot- án

- imitating - kleptomaniac - mischief

dim. (imit., pej.)

(ka-) niyá

3. pers. sg.


to each his own; every man for himself

 intens.? kanyá-kanyáng báyad 'Dutch treat; to work on each man for himself'


crazy (seriously ill)


crazy, mad

only joking, not serious (used to address nice and funny persons)



bed; camp bed (< Spanish) (soft bed: kama)


small bed

kuru-katre: “imitation of a bed”

Curu-: dimin. (imit.)


tilted or inclined to one side; uneven



kibáng-kíbang 'move plural from side to side careen (a platform)'


Appendix D: List of Bikol reduplications


mouse, rat, rodent


1. carnival game (in circus which a mouse or guinea pig is let loose near a ring of rotating, numbered houses 2. the lump caused by a sharp blow to the muscle 3. (slang) describing a man who tries to start up relationships with women in a hit or miss kind of way, never showing particular interest or seriousness



to stand or walk on tiptoe


to walk heel to toe

kuru-kintid walk on the tiptoes; not fully tiptoe

Curu-: dimin.


sway, uneven line


referring to the swaying motion of the hips when walking




closed (door, wound, ...) also: sit with closed legs


to sit with the hands kipí'-kípi = not clasped around the completely closed, knees, usually as a knock-knees sign of shyness (women)



deformation of the eyes

kiráp-kiráp [MDL]

to blink, to flicker

kiráp-kiráp = seemingly a deformation



we (incl.)


among ourselves

 or: referring to a limited number of persons

intens. or: limit.


imbalance, shake

kiwág-kiwág [MDL]

to wag the tail (dogs)  diminu-tive and: shake a little bit plural (but repeatedly)


nervous Na-kubog siya dahilan sa istorya. 'He is nervous because of this affair.'


to shiver, to tremble




Appendix D: List of Bikol reduplications


a fence, to fence



 fence (low, mainly decorative, as might be used around a flower bed)


lacking, short (as of kulángmoney, food, time); kuláng under- (as in undervalued), deficient, inadequate, insufficient, wanting

feeble minded, retarded, an idiot



crawl (baby)

kunáp-kunáp [MDL]

to creep or crawl on the skin (as an insect)

kanap-kanap 'crawling'




to flutter, to flap

kupád-kupád 'to flutter, to flap' ma-kupád-kúpad seemingly slow





ka-kupu-kupu 'person plural who embraces' (habit.)

kupáy [MDL]

to swim by paddling kupáy-kupáy with the hands (in [MDL] the water) kampay

to motion with the hands, to beckon

kampay-kampay 'paddle in the water'


to obey, to comply with

obedient, mindful, headful



excessive, too much, labí-labí overdone (as anger, a response)

very much overdone, simplex unknown intens., pej. outrageous red:  pej., person who does the things in an inappropriate manner




a little bit broken



rotten or decayed ladó'-ladó' coconut, mag- to rot; mad (not seriously)

crazy, nuts (slang)

seriously mad


lágo [MDL]

to buy a full ration of at one time

lagó-lagó [MDL]

to spend unwisely, to unknown give things away in too generous a fashion



to walk, to stride


to go for a stroll or walk



extra fat found around the stomach and waist area


wattle, flesh hanging  = luru-lambí down from the throat or head of fowl





plural (habit.)


Appendix D: List of Bikol reduplications

(ha-) langkáw





highly ambitious h-ar~alangkaw plural

inside la'óg-la'óg (fig-) work, school, classes

entirely inside la'óg-la'óg sa ... 'appropriate for, just right for...'

la'óg-la'óg 'a little bit intens., inside' (just over the dimin. border, almost outside)




referring to the actions of the hands when trying to pick up hot and not get burned

scattered/a little wound

diminu-tive plural


buoyant; to float or be floating 1. ... on the surface 2. obvious, clearly visible


clearly seen or visible

keeps on staying at the surface

plural (cont.)





small cans






many bruises



growing of fire

layáb-layáb [MDL]

a flame (long, tongue-like), mag-: to dance (flames)

intensive growing of inten-sive fire


1. to fly, to take off layóg-layóg (as an airplane) an mga ... 2. a witch that flies upright with its arms outstretched and its eyes gazing at the full moon; etc. [Bikol mythology]

flying insects or bugs


to walk around, to rove, wander, to meander, to surround


to walk around, to go  for a drive around



lidóng-lídong a hoop



shaky, unbalanced, unstable, unsteady


ligó' [MDL]

to shake in the hands ligó'-lígo' (as dice before rolling)

to rock back and forth on the water (as a buoy)

 to wobble, to be unsteady or unstable (keep on to be unstable)


Curu-: ?


plural (cont.) plural (habit.)

Appendix D: List of Bikol reduplications


to recover

likád-likád [MDL]

referring to the rolling of logs or bamboo as one attempts to walk on them

a little recover




limá-líma / li- only five lima

cf. anom

plural (distr.)

limán [MDL]

to defer, to put off, limán-líman to allow to run down without taking any action


turn the head

lingág-lingág to turn the head looking from one place to another



lingó'-lingó' [MDL]

to turn the head from , and side to side 1. a little bit crazy 2. always crazy

dimin., plural


frown face - reject for hurting s.o.

li'ód-li'ód [MDL]

to glare at



to move, to transfer to a new house


nomadic, describing  s.o. who moves from place to place

plural (habit.)


bunch of money





(sl-) crazy, nuts

lipóng-lipóng synonyme


fiery, glimmering (the eyes)



simplex: splitter, astray red.: keep on


to go crazy, nuts


crazy, nuts, off, cracked, a fool


1. to pound lubák-lubák (bananas, root crops such as the sweet potato) 2. road hole

rough (as a bumpy road), potholed

full of road holes

to ignore what is procrastinating happening, to pay no attention to what is said half turning

- s.o. is hurting - the frowning of the face (repeated)

a lot of bunches of money



plural; dimin.

plural (cont.)


Appendix D: List of Bikol reduplications


1. the backside, lubót-lubót behind, buttocks, [MDL] rear end, rump, seat, anus, rectum, ...; the bottom, the last, sa lubot kan train = the last wagon 2. to talk about people behind their backs, to backbite

to indicate a little bit back displeasure or dislike (impolite usage)



a rash; small insect bite; bump



ludóg-ludóg 'a little swelling' ludog-lúdog 'crazy'






1. a little wounded 2. a lot of wounds

1. dimin. 2. plural/intens ive




crazy woman


lukón [MDL]

to conceal or hide, lukón-lukón / the curvature of the to deny the existence luluknán hollow of the knee of





every Monday

plural (distr.)


to fly (not actively)


to glide in flight, to hover, to soar



deficient air



a little bit deficient air


“quiet” (as a command)


to tell s.o. to be quiet unknown [MDL]




slow, careful; slowly, carefully


to breed, reproduce (animals); copulate (animals)


to dote on

repetitive copulation plural



Muró-Martes every Tuesday


1. to feel (as a matí'-màti' pain); to sense; to feel sick 2. to hear





Curu-: plural

a slight illness

 2. usually heard, unverified news


every Wednesday

Curu-: plural


Appendix D: List of Bikol reduplications


muslim (sp.)


drama (which depicts the struggle between the Moslems and the Christians, in favour of the Christians)

 (political lexica-lized connotation) 2. Show (e.g. the government has a program but doesn't tell the truth about it)


finger, fingers; toe, toes

muró'-muró' [MDL]

(to talk hesitatingly about due to confusion or lack of information)

every finger




muyá-múya [MDL]

happy, content

like a little (dim.)



curious in an nalí'-nalí' innocent or ignorant sort of way; also used to describe people who keep drawing attention to obvious or known; obsessed (pos. and neg.)

describing s.o. who is nalí'

somewhat obsessed

plural (habit.); dimin.




too much, excessive; nganá'-nganá describing one who super-excessive overdone, inordinate ' does things to excess


ngutá' / nguyá'

to chew, to masticate (animals)

ngutá'ngutá' / ngatá'nguyá' / nguyá'nguayá'


plural (cont.)


point with the lips

nguwónguwó [MDL]

 to speak in a confused and roundabout manner, never making a point or reaching a conclusion


a pair


1. to pair many things together 2. card game

namás-namás excessive, too much; 1. a little naughty overdone in action or 2.  response; undisciplined, spoiled


continuously chewing

1. dimin. 2. intens.


1. more than one pair plural 2. color-coordinated look (clothes, accessoires, etc.) 3. card game (with several pairs of cards)

Appendix D: List of Bikol reduplications


hit (corruption)

pakángpakáng [MDL]

(describing s.o. excess. / dim. naked from the waist down)

intens; dimin.


shovel, spade


framework, scaffolding, trellis, lattice



carried by the wind

palíd-palíd [MDL]

(silver or gold embroidery)

palid-pálid swaying also fig. (unstable political opinion)


pole, mast (sp-) old: now: pole


spindle (sp-: stick)

piece of wood, used to wring the laundry

dimin.; lexical.


just for or just pará'-pára' because of; only for or only because of; erase, fade

or the first time

erasable, again and again






heavy wind



rice still growing in the fields; rice before milling

paróy-pároy [MDL]

weak with hunger or illness



part, portion, section, segment





to pass (as pasá-pása 'please pass the salt') (sp-: pasar) 2. to throw toward to

passing (as from one  person to the other)



pass, passage, passageway; step, pace


to pace back and forth (sp-)


paták [MDL]

split, cleft



a lot of dripping



dumb, stupid, dense, patál-patál dull, slow (sp-: fatal)

very stupid

a little bit stupid

intens; dimin.


tortoise, big turtle (small turtle: bako)

parópawikánon / pawikánon

a crybaby

*paro-p. Curu-: puru-pawikán dimin. imitation of a tortoise (imit.) puru-pawikán-on personification of a tortoise: symbol of crying


hut, shelter, cabin (traditionally constructed in the fields where one is working)



smaller than a hut, not for living in


dividing into parts


Appendix D: List of Bikol reduplications

ping-gán a plate

puró-pinggán jellyfish (large, unknown, whitish, translucent) interpretation: toy plate





to return or go home pulí'-púli' (Naga)

to commute; to unknown travel back and forth (as to school, work)



pungágpungág [MDL]

to awake in fear



stunted, dwarfed, squat, stubby

puróg-puróg [MDL]

to be wet with drizzle, mist, spray





a little bit destroyed re-raba' 'entirely destroyed'


scattering, falling of radág-radág ripe fruit [MDL]

the sound mad by plural (left over, fruits when they fall scattered, fruits of from a tree which is the trees) shaken



smeared (as food in the face, ink on a sheet, etc.)


1. describing s.o. intensive, plural of who mumbles, to the simplex mutter, to mumble 2. describing s.o. with smeared on his face, particularly food around the mouth

plural, intensive


bad, lousy, unfavorable, corrupt; destroy


(sl-) crazy, mad, nuts a little bit destroyed ra'ót-rá'ot very bad, in extremely poor condition



demolished, destroyed; dilapidated, rundown

rapák-rapák [MDL]

the sound made by breaking bamboo, reeds

a little bit rarapak 'intensive'


ratób [MDL]

to gnash (the teeth)

ratób-ratób [MDL]

the sound made by gnawing, making such a sound



risík [MDL]

to become heated to risík-risík the point where it [MDL] can be soldered or welded

crisp, as a new dollar unknown bill



Curu-: dim.(imit.)

herbal medicine plural (mixed in a ratio 7x7) (lexicalized)




Appendix D: List of Bikol reduplications


crazy, insane, mad, a rungáwfool; only with rúngaw respect to animals


a little bit crazy


topsy-turvy, disorganized


a little bit topsy-turvy dimin.

rúpok [MDL]

to smash to rupók-rupók pieces or fragments; weak (because of age or illness)


somewhat weak



to do together; to do at the same time


concurrent, simultaneously, synchronized,

together, more than two; all together


sagot (Tag.)



to answer back, to be (Tagalog) sassy





if I am not mistaken; a lot of mistakes all things considered



to depart, to leave the station (trains), to replace, to change, ...

saró-salída / saró-s-aralída

to vary, to fluctuate, being replaced to alternate, to work continuously (e.g. a in shifts, etc. guard of a movement), replace each other, people taking turns

Curu-/ -Vr-: plural


to catch fish with the saló-sálo hands; to support from below with the hands

a game of catch




to venerate, worship; sambá-sámba praying mantis to adore (religious context)





sarósantóhan / santó-santíto

sanctimonious; to pretend to be good or saintly

suru-santó-han / santó-santíto

Curu-: dim. (imit.)


to chew



keep on chewing (e.g. a bubble gum)

plural (cont.)


the sound made sapák-sapák when animals chew; drub


many drubs





foot of a sewing machine

suru-sapátos 'it's being worn'

Curu-: dimin. (imit.)


to keep, to retain



keep over a long period of time

plural (cont.)




Appendix D: List of Bikol reduplications


one, single; alone, by oneself


to do one person at a time, saró'-sáro' 'only one; all alone'


1. to articulate, enunciate, pronounce; 2. prenuptial ceremony (not performed any longer)


1. clearly seen or understood; pronounced, prominent, obvious

2. interpretation: 1. intens., 2. sayod-ceremony with dimin. very few people




every week


Curu-: plural


to split or cleave; cracking of a tree


to split or cleave

slight cracking of a tree



boat (without balance poles)


vinta (Philippine sailboat)

a small boat



to move or step back; to retreat or withdraw


to vacillate

to step back a little



to be busy or occupied


very busy or occupied, engaged, hectic


sígay [MDL]

to approach each sigáy-sígay other ready to fight (two fighting cocks), referring to cocks and men

describing a male who does to attract women

1. to court s.o. 2. to approach each other a little bit (in order to fight)



okay, all right, sure


 continuing, persisting, relentless, steady, uninterrupted

plural (cont.)




(pseudo, bogus, spurious; a sham; a person of no consequence or importance, a nobody)



curve, bend, elbow

sikó'-síko' / sikót-síkot

zig-zag, having many curves or bends





to nibble on (as sneeze slowly, a candy bar); to walk slightly around while eating



pedal, ride on a bicycle


Appendix D: List of Bikol reduplications



siro-singsing (suro-...)

ringlets, rings (as for suro-singsing curtains) 1. a kind of a ring (similar) 2. to wear a ring Suru-singsing ko su singsing ni Harry. 'I am wearing Harry's finger ring.'

Curu-: dimin. (imit.)


the fork of a tree, insertion; groin


crotch, groin

1. groin 2. alternate







plural (cont.)


delectable, delicious, luscious


(fig-) (magpa-) to enjoy oneself, to have fun, to have a good time

enjoying, without considering if there are any consequences


thin-skinned, easily sirí'-sirí' angered; illhumored, ill-natured

hotheaded, easily angered





1. only nine cf. anom (magpag-, pag- -on) to divide into nine, to send nine at a time 2. lexicalized: “until the cows come home” (i.e. never) An tesis mo abot-on ki siyam-siyam bago ma-tapus. 'My thesis will not be finished until the cows come home.'

1. limit. 2. lexicalized


chick; chicken embryo


an embryo




to be alone, to go somewhere or do alone


all alone

live alone, as a single intens.?


alike (Sorsogon)


very much alike




egg (Naga)





to crawl through a tight space, corner


out of the way, remote, removed, sequestered



Appendix D: List of Bikol reduplications


a leaf from the sumbá-súmba a kite with such a anáhaw or coconut leaf (Spanish: palm placed on the zumbar: to hum, edge of a kite wing buz) so that it will make a humming noise in the wind

a small leaf


be fed up with


- in abundance


next, follow


chronological, consecutive, sequential

sunód-súnod 'obeying'


a letter, mail; surát-surát writing, penmanship

to scribble, to scrawl - vandalism (daub dimin. - pretend to write (as children do)


pouring of vine



to hand round a glass plural of wine


eyelid, fix a pair



always stick together plural takob-tákob 'fixing of the cover'


accurate, correct, right


just right, ideal, perfect

just enough




to prance around on;  to stamp the feet up and down on, to trample on




taná'-tána' [MDL]

to do slowly, gradually; to work on bit by bit

taste a little bit



to give, issue, impart; to hand in, render, submit

ta'ó-ta'ó (mag- -an)

to reciprocate; (pa-an / -rV- -an): mutual, reciprocal, interdependence, mutuality, reciprocity

pa-ta'o-ta'o 'exchange'




ta'ón-ta'ón = every year; annual, káda ta'ón annually mang-: to complete one year, to reach one's birthday

plural (distr.)


completed, finished, sa tapósin the end, finally, terminated tápos / sa ka- lastly, ultimately tapós-tapós-i istóriang tapóstápos 'short story/non-serialized story'

tapos-tápos 'completely finished'






Appendix D: List of Bikol reduplications


hello, what's up?


a greeting, salutation; an introduction, prologue



to stub



continuous stubbing



to dance tarók-tarók and: planting of rice


lopsided, tilted, tibáng-tibáng synonyme inclined to one side, uneven, not level

a little bit tilted



bent, curved


having many bends or curves, crooked, sinuous, tortuous


tilíng [MDL]

a high voice, soprano; sound of a small bell and: crazy


a small bell; a tinkling sound

ringing of the bell (e.g. the ice cream man)




fish species

row, paddle


tína' [MDL]

to dye or tint black


plant (possessing leave from which a black dye may be optained)


lexical. deriva-tion


1. all 2. to proceed


1. all, everything

2. do without stopping

1. intens. 2. plural (cont.)


lesbian, homosexual, toró-tómboy gay (woman)

a tomboy

seems to be lesbian, but is it nor really

Curu-: dimin. (imit.)


1. to wallow in the mud and: plating of gold

tubóg-tubóg [MDL]

deeply embedded (an 1. dim. arrow, a lance or spear)


sibling, brother or sister

paka-t-urfraternity úgang-túgang



tukáb-tukáb [MDL]

to palpitate; to throb , and: flipping


a covering, pick up


a flap

to prance about as if ta-tarok-tárok 'not dancing seriously dancing'


túgang-túgang 'stepbrother or stepsister, adopted brother or sister' paka-turúgangtugang [old term] > today: pagiging magturugang

flapping of a cover, blown by the wind


1. intens. or dimin. dimin. (imit.)

Curu-: plural

Appendix D: List of Bikol reduplications


a bridge, a span


a small or makeshift  bridge


immediately, promptly; outright


instantly, at once


to approach s.o. to tumóy-túmoy talk over a question; (makipag-) to interrupt a conversation

to make a small talk, unknown to kiitz; to chatter



slang for 'appear' (after a long time)

tungábtungáb [MDL]

(to move about (fish repeatedly disappear buried in the mud)) and appear (habit)



to guess, to predict or prophesy 1. accustom 2. obey, believe

tu'ód-tu'ód [MDL]

- to do with care guess somehow and confidence - (pagka-) manner, disposition



to memorize; to commit to memory


suspicious; a suspect memorize, but not carefully / seriously




tupí-tupí [MDL]

embroidery with 1. a little bit fold gold or silver thread 2. fold intensively

1. dim. 2. int.


a measurement tupóngequivalent to ten tupóng arm lengths on each side; exactly equivalent in size

(a human-like not exactly creature which can equivalent, but only stretch or shrink to more or less take the shape of whatever it is next to [Bikol mythology])


tupós [MDL]

1. to squeeze out all tupós-tupós the juice; [MDL] 2. burnt

to see a project through to its end

2. somewhat burnt



(sl-) crazy





(sl-) crazy




sleep, slumber


mimosa (plant possessing leaves which close up when touched) turóg-túrog 'to act like one is sleeping, to feign sleep'

1. turog-túrog: 1. lex. deriv. mimosa 2. dimin. 2. turog-turóg: to act (imit.) like sleeping


a drop, a trickle


to drizzle, shower




consumed, exhausted, finished

ubós-úbos biyáya'

nothing left in reserve

, and ubos-úbos 'consuming at once'




Appendix D: List of Bikol reduplications

úlay to converse with, to (maki-, speak to or talk to, ma- -ka, to discuss with s.o. ka- -on)

uláy-úlay (pag-)




1. uló-uló 2. uló-úlo [MDL]

1. tadpole, mosquito 1.  larvae 2. unknown 2. feathers placed at 3. glans the prow of the boat called barángay

3. lex. imit.?


the distance between urú-ultan two things, e.g. with planting (usually: pag-ultan-an)

alternating, staggered


Curu-: plural


boil rice to its appropriate consistency



1. boil rice on small flame 2. stay in bed after wakened up

1. dimin.


pig, hog, swine



piggybank; s.o. who is pudgy



make a hollow-back urad-urad


a light hollow-back



extra skills


- to make a show of one's own skills - upset

usíp (maka-)

to converse, to have usíp-úsip a conversation, secrets, gossip

story, topic of conversation

a little bit secret



1. clownish 2. to push


1. someone who is clownish or who clowns around

2. push and pull

1. dimin. 2. pl.


to amble, to swing the arms when walking

wakóy-wákoy lanky




1. waló-waló 2. waló-wálo / wawaló

plural (distr.)


to dissipate, scattered

warák-warák to fall from above [MDL] and scatter on the ground

 plural




(Naga dialect)

1. to divide into eight, to send eight at a time 2. only eight






chapter i: introduction - Reduplication

INTRODUCTION CHAPTER I: INTRODUCTION This dissertation is a study of reduplication in general and a description of the types of reduplication in Biko...

2MB Sizes 9 Downloads 65 Views

Recommend Documents

chapter i introduction - Shodhganga
proclaimed the unity of Indonesia as "satu nusa, satu bangsa dan satu bahasa " (one land, one nation and one .... book,

published in 1988 in United State of America by St. Martin,s Press. The feature of the novel is cannibalistic serial kil

chapter-i introduction - Shodhganga
of socio-political and economic background is sought to be assessed to grasp the nature and type of administration of a

CHAPTER I Introduction - Shodhganga
In this process the city of Goa was made to evolve as the core centre, where the essential .... of the Portuguese imperi

chapter i introduction chapter i introduction - Digilib UIN Sunan Ampel
students can access the book in the form of e-book or it can be downloaded for all students free. This way ... 1 Republi

1970s, not a few those who show the spirit of freedom embodied in the. 3 Maman S Mahayana, Angkatan 70-An: Kembali Ke Tr

chapter - i introduction - Shodhganga
Jhumpa Lahiri is one of the distinguished short-story writes of. Indian diaspora .... This Blessed House is the home sha

Chapter I Introduction - Shodhganga
waves of democratization have occurred in the modem world. 1. The first ... According to Huntington the second wave of d

DATA 1 Teacher : Dia merasa apa? Student : Enjoyed. Teacher : Excited nya Enjoyed. I enjoyed it a lot. Itu kenapa ada ka

2 He is an expert commentator of al- Qur a>n which is famous with his Tafsi>r Al-Mishbah. Besides with Tafsi>r Al- Mishb