Publishing Partner: Cambridge University Press CUP Extra Publisher Login

New from Cambridge University Press!


Revitalizing Endangered Languages

Edited by Justyna Olko & Julia Sallabank

Revitalizing Endangered Languages "This guidebook provides ideas and strategies, as well as some background, to help with the effective revitalization of endangered languages. It covers a broad scope of themes including effective planning, benefits, wellbeing, economic aspects, attitudes and ideologies."

New from Wiley!


We Have a New Site!

With the help of your donations we have been making good progress on designing and launching our new website! Check it out at!
***We are still in our beta stages for the new site--if you have any feedback, be sure to let us know at***

Review of  Annotated Texts in Beṭṭa Kurumba

Reviewer: Sanford B Steever
Book Title: Annotated Texts in Beṭṭa Kurumba
Book Author: Gail Coelho
Publisher: Brill
Linguistic Field(s): Language Documentation
Text/Corpus Linguistics
Subject Language(s): Kurumba, Betta
Issue Number: 30.3689

Discuss this Review
Help on Posting

Based on the author’s quarter-century study of the language, Annotated Texts in Beṭṭa Kurumba (ATBK) is the most comprehensive and authoritative treatment of the Beṭṭa Kurumba (BK) language ever to appear. BK is a non-literary South Dravidian language spoken in the Waynad-Nilgiris micro-linguistic area of South India, a mountainous region that includes Ālu-Kuṟumba, Badaga, Irula, Kota and Toda, among other languages. One defining isogloss of the micro-area is use of the Dravidian etymon *kāṭu ‘wasteland, wilderness’ (DEDR 1438) to mean ‘field’, more precisely ‘swidden’, i.e. a field created by burning wasteland. The Kolami word vēgaṭ- ‘field’ illustrates how this semantic transition occurred: it a compound of vē- ‘burn, be scorched’ (DEDR 5517) and kāṭu ‘wasteland’. Interestingly, the Kongunadu dialect of Tamil, which abuts this micro-area from the east, also uses kāṭu in this meaning, as in the recent novel Mātorupākan ‘One Part Woman’.

Contacts with the three surrounding literary languages Kannada, Malayalam and Tamil have led to a degree of bilingualism in the BK speech community. One text (p. 462) records that the ancestral god of a certain BK phratry speaks Malayalam! In another (pp. 251-55), the protagonist sings songs in Kannada. Former administrative languages such as Urdu and English have also left their imprint on the BK lexicon.

With approximately 32,000 speakers spread over three states, BK does not appear endangered in the near term. However, maintenance of the various non-official languages is a low priority of state administrations, so BK remains subject to external pressures that may alter the language and its use. Coelho’s informants, for example, live in a hamlet within the Theppakkadu National Park, where some of their traditional hunting-gathering economic activities have been curtailed, and replaced by new ones promoted by the government.

ATBK is, to my knowledge, the only collection of South Dravidian texts to appear with interlinearized glossing. Reading texts in Ālu Kuṟumba, Kota and Toda generally requires one to skip back and forth between the originals and their translations, attempt to match up the original words with their translation, and guess at their morphological and syntactic analysis.

The Introduction (Chapter 1) contains an overview of BK, including a grammatical sketch and information about the physical, historical and social context. The sketch will enable readers to interpret the glossed sentences in the various texts. Phonology consists of an inventory of phonemes, certain phonotactic constraints and remarks on syllable and word structure. BK morphology is agglutinating and exclusively suffixal. A primary distinction is made between nouns and verbs; adjectives constitute a minor category. BK also has a set of postclitic particles, which generally deal with discourse and pragmatic phenomena.

Nouns are marked for case: nominative (unmarked), accusative, genitive, dative, comitative and locative. The cases are distributed differently between human and non-human nouns, e.g., while non-human locatives simply take the case marker, human locatives require a postpositional phrase, often involving insertion of a human body part to which the locative case marker is then suffixed, e.g. pəṇəkkən benntl nambikki ‘trust in a woman’ [woman body-LOC trust] (p. 275). BK preserves a Proto-Dravidian distinction in the first person plural between exclusive naŋgǝ ‘we (not you)’ and inclusive yaŋgǝ ‘we (and you)’ forms, but has lost gender oppositions in all its pro-forms. This latter development has implications about how speakers keep track of referents in a discourse.

The word-formation rule for verbs is complex, with as many as 11 internal morpheme boundaries (p. 25), making the language look polysynthetic. It may be simplified by reanalyzing the theme formatives (TF) as automatic adjustments to following derivational (DS) or inflectional suffixes (IS). In effect, the TFs create stems to which a DS or IS can be suffixed. TFs lack distinctive meaning, being the residue of earlier morphology. For example, the transitional verb form kel-t-əŋa ‘upon reading’ is glossed as read-PST-TRNL with three morphemes and two morpheme boundaries, but there is no non-past alternative *kel-p-əŋa to contrast with, so the –t- here is not a full-fledged morpheme with past tense meaning.

Some of this complexity arises from the fact that certain verb forms are contractions of earlier compound verb constructions, having undergone univerbation, so that an old auxiliary verbs are now bound morphemes (p. 34). Most are related to independent verbs. Only the morpheme –ūr- ‘to halve’ has no independent use; however, it is likely cognate with Kota ūry- ‘throw’, which serves as an intensive auxiliary verb in that language, e.g., en āḷ idn tavarčṯ ūrykō ‘my husband killed it off’ [I-GEN husband it-ACC kill-CNJ throw-PST-3] (Emeneau 1944-46, Text 21.127).

As elsewhere in Dravidian, BK distinguishes between finite and nonfinite verbs (see Steever 1988), with finite verbs restricted to certain positions, e.g. the final verb in a sentence or before the quotative verb ān- ‘say’. BK has elaborated its set of nonfinite forms to convey a variety of aspectual and temporal uses; it has at least ten (pp. 35-7). Nonfinite forms are used in forming coordinate, subordinate and cosubordinate (verb-chaining) structures.

The typical finite form (apart from hortative and imperative forms) consists of a verb base, tense marker and subject-verb agreement (SVA) marker, with a typical nominative-accusative system of alignment. Interestingly, BK has two sets of SVA markers. Set I contrasts only singular –ǝdǝ versus plural –ǝgǝ. Set II, with two subsets, makes further distinctions. Set IIa has /-i/ for 1S, 2S, /-a/ for 3S and /-o/ for plural; Set IIb /-iya/ for 1S, /-i/ for 2S, /-a/ for 3S and /-iyo/ for plural. Set I appears when the action is backgrounded, while Set II is used to mark ‘heated action’ (p. 86). Though Dravidian dative subjects do not usually trigger SVA, Coelho shows that in some BK dative-subject constructions, /-i/ marks agreement with a dative subject.

Coelho draws attention to the following paradigm: navǝ ǝḍtǝdǝ ili ‘I did not take it’ [I take-PST-S be.NEG.1S], naŋgǝ ǝḍtǝgǝ ilo [we take-PST-PL be.NEG.PL], navǝ ǝḍpǝdǝ ili ‘I do not take it’ [I take-NPST-S be.NEG.1S], naŋgǝ ǝḍpǝgǝ ilo [we take-NPST-PL be.NEG.PL]. These compound verb constructions are serial verbs (in the sense of Steever 1988): the preceding main verb, which may occur independently as a finite verb, shows number agreement with the following finite auxiliary. Some similarities may be drawn with Malayalam past and present negative compound verbs, which also consist of two forms each of which is independently finite. The BK and Malayalam forms could be independent innovations or, perhaps, restructured retentions of an earlier negative serial verb formation. The gradual loss of person-number-gender marking of the proto-serial verb leads to the BK paradigms, its complete loss to the Malayalam.

ATBK also sheds light on a set of formulaic compounds consisting of two finite verb forms, e.g. nəyra barəy ‘come look’ [see-IMP come-IMP-S]. The second verb in these formulas is restricted to ‘come’ or ‘go’. Other South Dravidian languages have similar formulas, e.g. Kota tinkōm vām ‘let’s eat’ [eat-HORT come-IMP], but Coelho is the first to draw attention to this pattern in the modern South Dravidian (SDr) languages.

While typical of a SDr language, BK has a number of unusual features morphosyntactic features. It lacks echo-compound formation with ki- ~ gi-, e.g. Tamil oṭampu kiṭampu ‘body and the like’, Kannada prīti gīti ‘love and such’, Kota kūv gīv ‘all sorts of food’, Ālu Kuṟumba bēci gīci ‘cooking and such preparations’. It also lacks obvious, widespread reflexes of the coordinating clitics *=ō ‘any, or, whether’ and *=um ‘all, and’, which function as quantifiers and conjunctions. This throws some of the burden of conjunction onto the processes of simple juxtaposition and compounding.

The texts include folktales and dialogues, but the two genres are not entirely distinct. On the one hand, the telling of BK folktales requires audience participation in the form of backchanneling; as Coelho notes (p. 59), “Storytelling is generally a dialogic event… ” On the other, the interlocutor in the dialogues provides minimal input, encouraging the informant to continue at length. All the texts are preceded by their full English translations.

The section on folktales (Chapters 2-6) includes a Prelude and four tales: Turban Maker (289+ sentences); Fish Prince (80+ sentences); Offended Daughter (130+ sentences); and Prince who Subdivided Himself (99+ sentences). The section on dialogues (Chapters 7-10) include a Prelude (311-64); Aspects of Community Life (325+ sentences); Legends about Deities (298+ sentences); and Legends about Ancestors (409+ sentences). Included in both the discussion and the texts of the dialogues is information about kinship, e.g. what constitutes marriage, legitimate and otherwise. This will help readers contextualize the motif of incestuous marriage in ‘Offended Daughter’ and ‘Prince who Subdivided Himself’.

The content of these tales and dialogues abound in non-BK themes, e.g. they contain references to princes and princesses, even though BK society has nothing comparable to royalty or kingship. This is not surprising: many societies, having transitioned to democracies, persist in telling stories about royals and royalty. Another example of assimilating external themes and culture is a folktale (p. 301) that includes a motif of writing down a history of events, even though BK is not itself a written language. (Some speakers may have a degree of literacy in such languages as Tamil.)

Coelho shows how BK storytellers weave this external content into a distinctive BK narrative style. One narrative device is the alternation between the two sets of SVA markers noted above. Another is apocope (pp. 61-63): in speech, all words except the last in an extended phrase lose their final vowels (most words are vowel-final in citation form). The scope of apocope often coincides with grammatical units such as sentences or with narrative periods. Further instances of BK narrative style are discussed below.

The book is rounded out by a Glossary of BK words (613-38), References and an Index.

There are very few printing errors in this book; most are self-correcting. However, the transposition of ‘former’ and ‘latter’ on page 86 gives the impression that perfective aspect is used for backgrounding and imperfective aspect for foregrounding, when the opposite is meant. Apocope is absent from the Index. The phrase ‘light-verb jingle’ for ‘light-verb jungle’ in the References (p. 639) gave me a chuckle.


Not only does ATBK admirably succeed in its primary goal (p. 78) of providing material for grammatical, particularly syntactic analysis, it also offers new material for comparative linguists, anthropologists, sociologists and folklorists.

One immediate, important result is that comparison of BK texts and grammar with those from Ālu Kuṟumba (Kapp 1982), reveals that BK and Ālu Kuṟumba are two distinct languages, not dialects of one.

A comparative linguist might provide the DEDR numbers for most of the entries in the Glossary. The Glossary will also furnish material for a dictionary of the Nilgiris micro-area lexicon, long a desideratum. For example, BK kisəl ‘be able’ is one of a very few cognates of the Old Tamil auxiliary kil- ‘be able’, which became the present tense marker in Modern Tamil.

A specialist in folklore will be able to identify the motifs used in the various texts. For example, ‘Turban Maker’ contains the motif ‘love through the sight of a hair of unknown woman’ which prompts the search for that woman (Thompson 1946:498). It is adapted into different South Dravidian oral literatures in different ways: in a folktale in Kota (Emeneau 1944-46) and Ālu Kuṟumba (Kapp: 1982:257) the hair is found inside a fish caught in a stream, in the BK folktale it is found woven into the fabric of a turban cloth. The absence of echo formation means that BK is unlikely to borrow a motif based on this grammatical process; Kota does have such a grammatical process and uses it as a motif (Emeneau 1944-46, Story 19).

The following are some grammatical observations and suggestions I have gleaned from my reading of the texts; they do not exhaust the potential riches to be found there. Further, they could not be imagined without the rich collection of texts in ATBK. Coehlo has made a major contribution to Dravidian linguistics simply by enabling scholars to make them (thanks, again, to those interlinearized texts).

The texts reveal that marking in BK may mark honorification, e.g. igə ēn māḍṇō nārāyən? What did he, (the god) Narayan, do?’ [they-PL what do-PRF-PL Narayan]. BK also has collective plurals, e.g., gəṇḍāla ubbaru [husband both.HUM] means ‘husband and wife’, not ‘two husbands’; pəṇḍə ərru āḷu [wife two person] means ‘husband and wife’, not ‘two wives’; and ammən-rər-ka [father-PL-DAT] means ‘to the father’s people, household’, not ‘to the fathers’.

As typical of Dravidian, polarity is a verbal category. The grammar offers speakers two ways expressing negation, through a single verb, e.g. dārə banrlo ‘no-one came’ [who come-PST-NEG-PL], or a compound verb, e.g. kə̄ṭəgə ilo ‘(they) did not ask’ [ask-PST-PL be.NEG-PL]. A closer study of the texts may reveal whether these are free variants or contextually conditioned. The texts appear to contain other auxiliary compound verbs, such as a prospective future tense, e.g. toḍḷa pə̄səŋ ‘upon going to touch’ [touch-INF go-TRNL] (p. 514).

No reader of BK texts can fail to notice how frequently expressions may be repeated, sometimes by simple reduplication but other times with an elaboration of the base of the repeated expression. Some instances are attributed to false starts (and are noted as such in the texts), but these patterns are too pervasive not to be systematic. Further research is needed to determine whether such repetition is due to grammar, narrative style or both. The varieties of repetition include the following.

BK has a Tail-To-Head device: the final verb of one sentence is repeated at the beginning of the next, usually in a nonfinite form. This helps to ensure coherence between parts of the narrative, e.g., gǝṇḍāḷǝ ḏressǝ ōṭǝya bǝyrnu uṭṭǝdǝ. uṭṭaṭu idǝ pǝ̄sǝdǝ ‘Getting men’s clothing, she wore them. Wearing/having worn (them), she went’ [man-GEN clothing all get-ACP wear-PST-S wear-CMP go-PST-S] (p. 188). In ‘Fish Prince’ it occurs 20 times in 80 sentences; in ‘Prince Who Subdivided Himself’, it occurs 41 times in 99 sentences. This device appears in folktales and dialogues. It occurs throughout the nonliterary languages from Kota to Kuvi to Kurux, but is rare in written texts of the literary languages. Nonetheless, it may be a source of the poetic device known in Tamil as antāti ‘end-beginning’, where that last word of one stanza is repeated at the onset of the next.

Simple reduplication may signal such notions as intensity, e.g. təbbə təbbə ‘very close’ from təbbə ‘nearby’ or quantification dārə dārə ‘whoever, who all’ from dārə ‘who’. In a few cases, it signals addition, e.g. ə̄ḷu ə̄ḷu pānāku moḷi ‘seven seven, i.e. fourteen, cubits’ [seven seven fourteen cubit] (p. 162). Doubling of verb roots, one in finite one in non-finite form, signals emphasis, e.g. naḍidənu naḍida ‘he did walk’ [walk-ACP walk-PST-3S]. Parallels abound in other languages, e.g. Kota arčak arčvē ‘I really did know’ [know-CNJ know-PST-1S] (Text 11.53).

Examples of repeated and elaborated NPs are startling in the Dravidian context. Consider the following example (p. 62): [[vanrǝ mutki kīrl], [kaḷi oḍkǝ mutki kīrl], [pǝynigivǝ mutki kīrl] idenu] ‘living in an old woman’s house, in an old sweeper woman’s house, in the house of an old working woman’ [one old.woman house-LOC floor sweep-RC old.woman house-LOC old.woman house-LOC live-ACP]. Here three NPs with the same head noun (and, presumably, referent) precede the clause-final verb: the first consists of a numeral + N; the second two, a relative clause and head N.

Another example from page 64: [[alli tana] [tan māmǝnrǝ kīrl tana] idǝnaḍdǝ] ‘she stayed there itself, (at) her father-in-law’s house itself’ [there EMP self-GEN house-LOC stay-PST.PRF-S]. The first NP has a locative adverb and emphatic particle, the second two genitives preceding a locative case-marked noun and emphatic particle. In many cases, the first NP is represented by a bare noun or a numeral and bare noun, while the second NP contains other modifiers, such as demonstratives, adjectives and relative clauses. The general progression (there are exceptions) is from general to specific. Such doubling appears occasionally in Kota, e.g. nin pādtn ām nimd kunǰ nimn kubiṭr vadēmē ‘we, your children, keep worshipping you, your feet’ [you-GEN foot-ACC we-NOM you-GEN child you-ACC worship-CNJ come-PRS-1PEX]. It is not, however, nearly as exuberant or widespread as in BK. In the 370 sentences of the first two texts in ATBK, there are at least 106 instances of repeated and expanded NPs. At first glance, BK NPs tend not to overburden their head nouns with long strings of modifiers, distributing them instead over several NPs with the same head. This is likely a stylistic tendency, not a grammatical rule.

The rhetorical device of chiasmus often draws attention to a focal point in a narrative. For example, on page 229 … kūṯnu attōḍḍə. atti kūṯnaḍədə ‘sitting she cried. Crying, she sat’ [sit-ACP cry-PST-PROG-S cry-STAT sit-STAT-S] signals an emotional turning point in the story. The third- and second-to-last sentences of the same story (p. 236) have one chiastic structure embedded in another, to nail the landing, so to speak, and end the story: [[porāmaynu ōṭa … [batkisəgə ubəru. ubəru batkisəŋa] ōṭəya āməŋa poṭṭkissə] … nala batkisəgə] ‘all (the others), being jealous, the two of them prospered. The two of them prospering, while all the others were jealous … they prospered nicely’ [ all prosper-PST-PL both both prosper-TRNL all=DISTR they-DAT jealously nicely prosper-PST-PL]. As a poetic device chiasmus is much used in Toda song language (Emeneau 1971), but in BK texts it may serve a grammatical function. The doubling and chiasmus of a pair of nouns gives rise to a reciprocal construction, e.g. pəṇḍə gəṇḍāḷə gəṇḍāḷə pəṇḍə samədāni mayrnu ‘husband and wife agreeing with each other’ [wife husband husband wife acceptance do-ACP] (pp 167-68).

A brief, non-BK aside. In ‘Offended Daughter’ (pp. 251-255), the heroine sits on a tree branch and sings, nān(u) barula ‘I will not come’ in response to entreaties to come down. The song is supposed to be in Kannada. While nān(u) is clearly the Kannada first person singular pronoun, I suspect that barula is a BK rendition of the older Kannada baral olle ‘I do not want to come, will not come’ [come-INF like-NEG-1S] rather than baral-illa ‘I did/do not come’ [come-INF be-NEG-NPST].

The texts in ATBK are only a part of those that Coelho has collected over the past quarter century. Dravidian scholars will applaud her admirable achievement and look forward to the future publication of additional texts and grammatical analysis of this fascinating, understudied language. ATBK is the foundational study on which all subsequent study of BK will stand.


I served on the author’s dissertation committee at the University of Texas.


ACP – accompanying event; CLF – classifier; CMP – completed event; CNJ –conjunctive form; DAT – dative case; DEDR – A Dravidian etymological dictionary,2nd edition; DISTR – distributive; DS – derivative suffix; GEN –genitive case; HORT – hortative; HUM –human; IMP – imperative; INF – infinitive IS – inflectional suffix; LOC – locative; NEG – negative; NPST – nonpast; PRF – perfect; PROG – progressive; PST – past ; STAT – stative; TF- theme formative; TRNL – transitional event.


Burrow, Thomas and Murray Emeneau. 1984. A Dravidian etymological dictionary, 2nd edition. Oxford: Clarendon Press.

Emeneau, Murray B. 1944-1946. Kota texts (Cols. 1-4), University of California Publications in Linguistics 2. Berkeley: University of California Press.

Emeneau, Murray B. 1971. Toda songs. Oxford: Clarendon Press.

Kapp, Dieter B. 1982. Ālu-Kuṟumbaru Nāyan: Die Sprache der Ālu-Kuṟumbas (Beiträge zur Kenntnis südasiatischer Sprachen und Literaturen 7). Wiesbaden: Otto Harrassowtiz.

Murukan, Perumal. 2012. Mātorupākan. Kalachuvadu Publications.

Steever, Sanford. 1988. The Serial Verb Formation in the Dravidian Languages. Delhi: Motilal Banarsidass.

Thompson, Stith. 1946. The Foktale. New York: Holt, Rinehart and Winston.
Sanford Steever is an independent scholar, specializing in the study of the Dravidian language family. His areas of interest include syntax, historical linguistics and the grammar of Tamil.

Format: Electronic
ISBN-13: 9789004378247
Prices: U.S. $ 180
Format: Hardback
ISBN-13: 9789004378223
Pages: 652
Prices: U.S. $ 198