An Outline Grammar of the Havyaka Dialect
Author: D. N. Shankara Bhat Year: 1971 Publisher: Deccan College Postgraduate and Research Institute, Poona Series: Linguistic Survey of India Series, No. 5 (Code: LSI 151) Language: English (academic) Subject: Havyaka Kannada dialect Source: DjVu scan (archive.org)
DjVu source:
20-havyaka-outline-grammar-djvu.md
Book Overview
An Outline Grammar of the Havyaka Dialect is D. N. Shankara Bhat’s 1971 academic monograph on Havyaka (also called Havigannada, ಹವ್ಯಕ ಕನ್ನಡ), the variety of Kannada spoken by the Havyaka Brahmin community of the two Kanara districts of coastal Karnataka and the Western Ghats. The description is based specifically on the variety prevalent in South Kanara (Puttur Taluk). It was published as the fifth volume in the Deccan College’s Linguistic Survey of India series and was issued in the same year as Bhat’s companion monograph on the Koraga language.
The book is a structural-descriptive grammar written in the tradition of American descriptive linguistics. It systematically covers phonology, nominal morphology (including derivation, plural marking, and case), pronouns, verbal morphology (including derivation, four finite paradigms, personal suffixes, and non-finite forms), sample sentences, an elicited narrative, and a classified vocabulary. The presentation was designed to facilitate comparison with other monographs in the same series. OCR artifacts are present in the digitized DjVu source.
This work is significant as one of the first rigorous academic descriptions of a coastal Kannada dialect, challenging the assumption — widespread among Dravidian scholars at the time — that coastal Kannada was simply identical to or a minor variant of the standard written dialect. Bhat demonstrates that the coastal dialects have a distinct history of at least one thousand years of separate development, preserving archaic tense categories (future and negative) lost in non-coastal dialects, and lacking the vowel-raising change (e > i, o > u) that defines non-coastal varieties. The fieldwork conducted for this grammar fed directly into Bhat’s subsequent comparative and typological work on South Dravidian languages.
Table of Contents
| Section | Page |
|---|---|
| Introduction | 6 |
| Phonology | 7 |
| Nouns: Derivation | 8 |
| Plural | 15 |
| Case | 17 |
| Pronouns | 21 |
| Verb: Derivation | 24 |
| Sample Paradigm | 29 |
| Past Tense | 32 |
| Personal Suffixes | 36 |
| Other Finite Suffixes (Future, Negative) | 39 |
| Non-Finite Suffixes | 45 |
| The root iru ‘to be’ | 49 |
| The root aradi | 50 |
| Sentences | 51 |
| Story | 58 |
| Vocabulary | 67 |
Introduction
Bhat opens by situating Havyaka within the unusually dense linguistic ecology of coastal Karnataka. The coastal belt is described as a “treasure-house of languages and dialects” — in addition to Kannada and its variants, the area hosts several dialects of Konkani, Marathi, Tulu, Malayalam, and a few tribal languages including Koraga (the subject of his companion 1971 volume). Social (caste) distinctions are reflected by sharp dialectal variation even within the same language.
Bhat distinguishes two types of Kannada in the coastal region. The first is a cluster of native coastal dialects — including Havyaka, Kota, Halakki, Gauda, and Kumbara — each with its own regional sub-variants, but all sharing distinctive phonological and grammatical features that set them apart from all other Kannada varieties. The second is a bookish standard Kannada acquired as a second language through schooling, not a regional mother tongue. Only the first type counts as a genuine coastal dialect.
Two major structural features distinguish coastal from non-coastal Kannada:
-
Vowel-raising absence: Non-coastal dialects (and standard Kannada) underwent a sound change around the 8th century CE whereby e became i and o became u in certain syllable environments (e.g., kemi > kivi ‘ear’, beli > bili ‘white’, oli > uli ‘to remain’). Coastal dialects including Havyaka and Halakki preserve the older, unraised vowels.
-
Retention of archaic tense paradigms: Old Kannada had four verbal paradigms: present, past, future, and negative. Non-coastal dialects retain only the past and have developed new non-past, subjunctive, and perfect paradigms. Coastal dialects retain the future and negative paradigms (in addition to the past), losing only the present — which they replaced with a construction similar to the non-coastal non-past.
A third difference concerns personal suffixes: coastal dialects have a single set of allomorphic conditioning, purely phonological, whereas non-coastal dialects show two to three sets conditioned by the tense/mood suffix. Additionally, finite verbs in Havyaka show no gender distinction in the strict sex-based sense; the masculine/feminine distinction of other Kannada dialects corresponds instead to an honorific/non-honorific distinction in Havyaka.
Phonology
Vowel Inventory
Havyaka has five vowel qualities: a, e, i, o, u. All five contrast for length (short vs. long), and all five can be nasalized. Quantity and quality contrasts are both phonemic:
- Quantity: kadu ‘strong’ : ka:du ‘forest’; bidu ‘to leave’ : ni:du ‘to stretch’
- Quality: kattu ‘to tie’, kittu ‘to smear’, kuttu ‘to hit’, hettu ‘to hit’, kottu ‘spade’
- Nasalization: banda ‘he came’ : banda (nasalized) ‘she came’
Consonant Inventory
The consonant system has the following structure:
| Labial | Dental | Retroflex | Palatal | Velar | |
|---|---|---|---|---|---|
| Stops (voiceless) | p | t | t | c | k |
| Stops (voiced) | b | d | d | j | g |
| Nasals | m | n | n | n | ng |
| Lateral/Flap | l | ||||
| Other sonorants | v, y, r, l | ||||
| Fricatives | s, s, h |
Note the absence of aspirated stops — Havyaka, like all Kannada dialects, does not phonemically contrast aspirated and unaspirated consonants.
Allophonic Features
All initial vowels are preceded by a glottal catch. The vowel e is lower-mid word-finally; a is higher and slightly centralized when not followed by length. The retroflex voiced stop (d) is flapped intervocalically and after a non-identical stop or a non-homorganic nasal. Gemination is phonemically contrastive: hata ‘obstinacy’ vs. hatta ‘attic’.
Nouns: Derivation
Derivation from Verbal Bases
Havyaka has a productive system of nominal derivation from verbal roots. Several suffixes are still productive:
-
-ate / -yate: Creates nouns denoting a disliked or undesirable quality. Added directly to longer bases (kayrjate ‘bitter and hence not good’), or after a reinforcing consonant for (c)vcv-type bases — e.g., -k- (harkate ‘torn’), -kk- (odakkate ‘broken pot’), -pp- (kanappate ‘astringent’), or -ang- (ba:dange ‘withered’).
-
-ku / -kku: Added mostly after (c)vcv pattern verbs. kolakku ‘dirty (adj.)’, harku ‘torn piece of cloth’, murku ‘broken furniture’.
-
-ike / -ikke / -ige / -ge: A productive action-noun suffix. nambike ‘belief’, hedarike ‘fear’, meccige ‘praise’, heccige ‘increase’.
-
-a: Added after bases ending in -u; in (c)vcv-type bases the penultimate consonant is devoiced. a:ta ‘play’, o:ta ‘running’, tumba ‘full, much’, katta ‘tie, bund’.
-
-i / -e: After (c)vccv and (c)vcvcv type roots. kutti ‘peg’, tatte ‘plate’, kole ‘murder’, kette ‘chip’.
-
-vu: After (c)vcv bases ending in e (base-final e becomes a) and (c)v:cv bases ending in yi. kajavu ‘theft’, karavu ‘milking’, maravu ‘forgetfulness’, sa:vu ‘death’, ga:vu ‘heat’.
-
-a:na / -ya:na: A productive suffix denoting a tiresome, unending activity. sa:rjka:na, ke:ra:na, ogeya:na, holiya:na.
-
-e:lu / -e:la: Added after bases ending in -u; denotes an agent or habitual doer. bagge:lu ‘bent’, ta:de:lu ‘goring cow’, cu:nte:la ‘one who pinches’.
-
-tta: After bases ending in i or e. bigitta ‘tight’, haritta ‘flow’, holetta ‘shining’.
Derivation from Nominal Bases
The suffix -a derives nouns denoting a male person from nominal or adjectival bases: kallu ‘stone’ → kalla ‘mason’; kurudu ‘blindness’ → kuruda ‘blind man’. The corresponding female suffix is -i: kurudi ‘blind woman’.
A more formal male/female pair uses -a/-etti: gattiga/gattigetti ‘clever man/woman’, ga:niga/ga:nigetti ‘oil-man/woman’. When the penultimate consonant is l or l, the female suffix is -ti instead.
Nouns: Plural
Plural forms are derived by adding -ngo after bases ending in a or a, and -go after other endings. Word-final e becomes a before this suffix:
- mara ‘tree’ → maranrgo ‘trees’
- medi ‘tender berry’ → medigo ‘tender berries’
- kannu ‘eye’ → kannugo ‘eyes’
- ka:ke ‘crow’ → ka:kago ‘crows’
Kinship terms form plurals differently:
- Words ending in a: add -ndru (maga ‘son’ → magandru)
- Words ending in i or u: add -akko (ajji ‘grandmother’ → ajyakko)
- Words ending in e: add -kko (abbe ‘mother’ → abbekko)
Nouns denoting persons of respect (priest batta, physician da:ktra, astrologer jo:ys¸a, teacher mantra) take -akko in the honorific plural and -ngo in the plain plural. In the singular these can also take -ru (batru, da:ktru) in the honorific sense, triggering plural verb agreement.
Nouns: Case
Five case suffixes occur, called accusative (A), instrumental (I), dative (D), genitive (G), and locative (L). The case allomorphy is conditioned by a classification of noun bases into five main classes:
| Noun Class | A | I | D | G | L |
|---|---|---|---|---|---|
| Class 1 (ends -a, rational) | -na | -nda | -nge | -na | — |
| Class 2a (ends -a, irrational) | -va | -nda | -kke | -da | -lli |
| Class 2b,c,d (numerals, demonstratives, verbal nouns) | -a | -anda | -kke | -a | -alli |
| Class 3a,b (ends -e, -i) | -ya | -nda | -ge | -ya | -li |
| Class 3c (rational plural in -o) | -aja | -alanda | -ge | -aja | — |
| Class 3d (irrational plural in -o) | -aja | -alanda | -okke | -aja | -alli |
| Class 4a,b (ends -u) | -ina | -inda | -inge | -ina | -ili |
| Class 5 (direction/time nouns) | — | -nda | -nge | -a:na | — |
When followed by particles beginning with a vowel (emphatic -ude, interrogative -o:), accusative takes increment -nn- and instrumental takes -l-.
There is no vocative suffix; the optional particle e: precedes the noun. A locative-emphatic suffix -a:re occurs with a small set of body-part and relational nouns: kanna:re ‘by one’s own eyes’, benna:re ‘just behind him’.
Pronouns
Personal pronouns:
| Singular | Plural | |
|---|---|---|
| 1st | a:nu | engo (exclusive), na:vu (inclusive) |
| 2nd | ni:nu | ningo (also honorific singular) |
Before the dative suffix -ge, the 1st sg., 2nd sg., and 1st pl. inclusive show allomorphs end, nina, and nova respectively. Before other case suffixes these have allomorphs enn-, ninn-, and namm- respectively.
Third person is derived from demonstrative bases:
- Remote: a-, a:-, ha:-
- Proximate: i-, i:-, hi-
- Interrogative: e-, e:-, he:-, a:-
Derived 3rd person forms include: ava ‘that man’, iva ‘this man’, adu ‘that thing/woman’, idu ‘this thing/woman’, avu ‘those things’, a:ru ‘who’. Adverbial demonstratives: alii ‘there’, illi ‘here’, elli ‘where’; a:ca ‘that’, i:ca ‘this’, e:ca ‘which’; a:ga ‘then’, i:ga ‘now’, e:va:ga ‘when’; ha:rge ‘like that’, hi:rge ‘like this’, he:rge ‘how’.
An inclusive-exclusive contrast operates in first person plural: engo (exclusive, speaker and others excluding hearer) vs. na:vu (inclusive, speaker and hearer).
Verb: Overview and Root Types
Verbal roots are classified by syllable structure:
- (c)vcv: kadi ‘to cut’, sigi ‘to split’, kadu ‘to give’, ele ‘to pull’, kade ‘to grind’
- (c)vccv: attu ‘to drive’, narjgu ‘to dimple’, kayku ‘to be bitter’, danku ‘to stumble’
- (c)v:cv: a:du ‘to play’, ku:ru ‘to sit’, te:gu ‘to belch’
- (c)v:ccv: cu:ntu ‘to pinch’, nu:nku ‘to push’, sa:ncu ‘to lean’
- (c)vcvcv: manugu ‘to sleep’, samalu ‘to overflow’, aralu ‘to blossom’
- (c)vcvccv: kenakku ‘to tease’, kalanku ‘to be turbid’, mogaccu ‘to overturn’
Verbal Derivation
Three productive derivational suffixes modify verbal bases:
-
Causative -su / -su: Derives transitive/causative from intransitive roots. me: ‘graze (intr.)’ → me:su ‘graze (tr.)’; soli ‘peel’ → solsu ‘make to peel’; ma:du ‘do’ → ma:dsu ‘cause to do’. The allomorph su vs. su is conditioned by the final phoneme of the root.
-
Reflexive -go / -ko / -kko: Derives reflexive or middle voice forms. ere ‘to pour’ → erakko ‘pour for oneself’; sigi ‘to split’ → sigko ‘split oneself’; hogalu ‘to praise’ → hogaligo ‘praise oneself’. Reflexive bases add ll before vowel-initial suffixes or particles: tekko ‘take!’ → tekkojji ‘you (pl.) take!’.
-
Completive -kku: Added to past participle stems; implies finishing off an action and often carries a polite request nuance. eje ‘to pull’ → eladikku ‘to pull off’; bidu ‘to leave’ → bittikku ‘to leave off’.
The causative and reflexive suffixes can combine: ma:digo ‘do oneself’, ma:dsigo ‘cause to do oneself’, ma:dikku ‘finish off doing’.
Verb: Finite Forms — Sample Paradigm of ma:du ‘to do’
Present
| Person | Singular | Plural |
|---|---|---|
| I | ma:dutte | ma:duttu (incl.), ma:dutteyo (excl.) |
| II | ma:dutte | ma:dutti |
| III M(h) | ma:dutta | ma:duttavu |
| III F(h) | ma:dutta | |
| III non-h | ma:duttu |
Past
| Person | Singular | Plural |
|---|---|---|
| I | ma:dide | ma:dittu (incl.), ma:dideyo (excl.) |
| II | ma:dide | ma:didi |
| III M(h) | ma:dida | ma:didavu |
| III F(h) | ma:dida | |
| III non-h | ma:dittu |
Future
| Person | Singular | Plural |
|---|---|---|
| I | ma:duve | ma:dugu (incl.), ma:duveyo (excl.) |
| II | ma:duve | ma:duvi |
| III | ma:dugu | ma:dugu |
Negative
| Person | Singular | Plural |
|---|---|---|
| I | ma:de | ma:da (incl.), ma:deyo (excl.) |
| II | ma:de | (no II pl. negative form) |
| III M(h) | ma:da | ma:davu |
| III non-h | ma:da |
Other Modal Forms
- Imperative sg.: ma:du; pl.: ma:di
- Prohibitive 1 (sg.): ma:dada; Prohibitive 2 (3rd): ma:dadde ‘let him not do’
- Concessive 1: ma:duvo ‘let us do’; Concessive 2: ma:dali ‘let him do’
- Assertive 1: ma:dida ‘it could be done’; Assertive 2: ma:dekku ‘it should be done’
Non-Finite Forms
- Conditional (present): ma:duttare ‘if (you) do’; (past): maididare ‘if (you) did’; (negative): ma:daddare ‘if (you) do not do’
- Simultaneous conditional: ma:duvadde ‘as soon as (he) did’
- Purposive: ma:dle ‘in order to do’
- Participial (positive): ma:di ‘having done’; (negative): ma:dadde ‘without doing’
- Relative participles: ma:duva ‘one who does’, ma:dida ‘one who did’, ma:dadda ‘one who did not do’
Verb: Personal Suffixes
| Person | Singular | Plural |
|---|---|---|
| I | e | tu ~ attu ~ u (incl.), eyo (excl.) |
| II | e | i |
| III M-honorific | a | avu |
| III F-honorific | a | |
| III non-honorific | tu ~ attu ~ u |
Key observations:
- I person singular and II person singular are homophonous in non-extended forms but distinguish in extended forms (before emphatic -ude or interrogative -o:): I sg. extends to -an, II sg. to -ey.
- I person inclusive plural and III non-honorific singular are homophonous throughout.
- The honorific system encodes social-caste relationships: male-honorific is used only for males of the speaker’s own community; female-honorific is restricted to relatives holding mother status (mother, aunt, grandmother, mother-in-law). All other referents take non-honorific forms. Plural forms are used in extra-honorific singular contexts (e.g., a wife addressing her husband).
Verb: Tense Suffixes
Present Tense Suffix
The present suffix is -t- after unnu ‘to dine’ and tinnu ‘to eat’, and -tt- after all remaining bases. The bases a:gu and ho:gu have allomorphs a:vu and ho:vu before this suffix.
Past Tense Suffix
The past tense suffix has multiple allomorphs conditioned by verb class: t, t, k, d, d. The distribution is complex:
- t: after bases like kedu ‘to decay’, bidu ‘to leave’, sa:yu ‘to die’, horu ‘to carry’
- k: after ogu ‘to boil over’, migu ‘to remain over’, hogu ‘to enter’
- d: after unnu ‘to dine’, reflexive bases (which take n as increment)
- d: after all remaining bases (the most common class)
Transitive vs. intransitive distinction in the past: bases ending in i or e that belong to the intransitive set take the III non-honorific singular suffix -tu, while the same bases in the transitive set take -attu. This creates a morphological distinction: haruttu ‘it was torn (intr.)’ vs. harudattu ‘she tore it (tr.)’; odattu ‘it broke’ vs. odadattu ‘he broke it’.
Future Tense Suffix
The future suffix has two allomorphic sets: k/g before III person u and I inclusive plural u, and v/p/b before other personal suffixes. The specific allomorphs are conditioned by verbal base class, giving forms such as ma:duve ‘I will do’, ma:dugu ‘he will do’; appe ‘I will become’, akku ‘it will become’; ho:peyo ‘we will go’, ho:ku ‘it will go’.
Negative
There is no negative tense suffix as such. The negative paradigm is formed by adding personal suffixes directly to the verbal base, with vowel modifications: bases ending in i, i:, e, or e: take -y- as increment; bases ending in u drop the final vowel; reflexive bases take extended form with ll; ta: and ba: take -r- as increment.
The Root iru ‘to be’
The section on iru details the paradigms of this auxiliary root, which participates in compound constructions and expresses existence and progressive aspect. Its forms are distinct from the main verb paradigms in certain respects, particularly in the way compound tenses are formed.
Sentences and Narrative Text
The grammar includes a set of elicited sentences illustrating grammatical constructions, followed by a narrative text (story) with interlinear glossing. These provide natural-speech data beyond the elicited paradigms and illustrate clause combination, topic marking, and discourse structure in Havyaka.
Vocabulary
The grammar concludes with a classified vocabulary (glossary), providing Havyaka forms alongside English translations across semantic domains. This lexical resource is foundational for comparative Dravidian work on the coastal dialects.
File Index
| File | Contents |
|---|---|
20-havyaka-outline-grammar-djvu.md | DjVu OCR text (raw, with OCR errors) |
20-havyaka-outline-grammar-en.md | This file — English chapter summaries |
20-havyaka-outline-grammar-website.md | Website metadata |
20-havyaka-outline-grammar-claude-prompt.md | AI prompt |
20-havyaka-outline-grammar.md | Collection overview |
| Internet Archive | Original DjVu source |
Cross-References
| Related Book | Connection |
|---|---|
| 09 — Havyaka Kannada | Companion Kannada-language popular work on the same dialect by the same author |
| 19 — The Koraga Language | Companion 1971 monograph from the same fieldwork period and same Deccan College series |
| 03 — Kannada Padagala Olarachane | Bhat’s word-morphology framework, applicable to the nominal derivation described here |
| 25 — Kannada Vakyagala Olarachane | Bhat’s sentence-structure framework, relevant to the syntax section |
| 24 — Grammatical Relations | Theoretical framework Bhat later developed, rooted in fieldwork like this |
This document is an English-language summary of D. N. Shankara Bhat’s An Outline Grammar of Havyaka (1971), based on the DjVu OCR source. The full OCR text (with errors) is in 20-havyaka-outline-grammar-djvu.md.