Lund University


Back to 2004 Conference page

Panel No. 27

Panel Title: South Asian languages in linguistic focus

Convenor: Dr. Ruth Laila Schmidt, Dept of East-European and Oriental Studies, University of Oslo, Norway
Co-convenor: Dr. Vit Bubenik, Department of Linguistics, Memorial University of Newfoundland, Canada

    Venue: Tornrummet (Tower Room) in the Academic Society Building, Sandgatan 2
    Tuesday 6 July, 13–18

Panel Abstract: The focus of this panel is the analytical treatment of language. Papers may analyze features of any modern or ancient South Asian language (some examples: analysis of the phonology, morphology or syntax of an Indo-Aryan, Dravidian or Munda language), or analyze historical linguistic or sociolinguistic aspects of modern South Asian languages (some examples: historical development of a South Asian language; language use in instruction, language use in ethnic or regional politics, endangered languages and so on). Papers will be based on linguistic, sociolinguistic, historical linguistic, lexical or textual data, and employ an accredited theoretical framework from some field of linguistics or philology.

Papers accepted for presentation in the panel:

Paper Giver 1: Ondrej Sefcik, Department of Linguistics, Masaryk University, Brno, Czech Republic

Paper 1 Title: On OIA root patterns

Paper Abstract: My paper attempts to explain the structure of the OIA root (based on the Vedic material) and to postulate models of the OIA root.
The first step is division of the root to the onset (consonantal), the coda (consonantal) and the peak (vocalic). The main difference between the onset and the coda on the one side and the peak on the other is the fact that the first two do not participate in any paradigmatic alternation whereas the peak enters in a paradigmatic alternation (= apophony).
The second step is an analysis of the patterns both of the onset (STR-/TSR- etc.) and of the coda (-RTS/-RST etc.)
The final part of my paper presents an interpretation of the syntagmatic rules used for the formation of the root – the rule of the voice, the rule of the aspiration and the rule of the mirror-shaping of the onset and of the coda.

Paper Giver 2: Václav Blazek, Filozoficka fakulta, Masarykova univerzita, Brno, Czech Republic

Paper 2 Title: Dravidian numerals

Paper Abstract: The Dravidian cardinal numerals have been reconstructed as follows:
*oru _ (C) / *dr _ (V) (DEDR 990a; Zvelebil 1977, 34) = *or- (G. Starostin).
*oo=u (DEDR 990d) = *on-tu (Krishnamurti 2001, 255: *o + -r-, -n-, -k- where the root *o had to be attested in Old Tamil *o "to unite") = *ond- (G. Starostin).
*okk- "one, single, alone" (DEDR 990b) = *ok(k)- (G. Starostin).
*ozri "alone, single" (DEDR 990c).
*iru _ (C) / *ír_ (V) (DEDR 474; Zvelebil 1977, 34) = *ir- (G. Starostin).
*muv _ (C) / *m¨ _ (V) (DEDR 5052; Zvelebil 1977, 34-35) = *m¨- (G. Starostin) = *muH- (Krishnamurti 2001, 330: plus the neuter marker *-ntu).
*nWl (DEDR 3655; Zvelebil 1977, 34; G. Starostin).
*cayN _ (C) / *cay _ (V) (DEDR 2826; Zvelebil 1977, 34-35) = *_ai- (G. Starostin).
*c!ru _ (C) / *cW= _ (V) (DEDR 2485; Zvelebil 1977, 35) = *_Wd- (G. Starostin).
*e8u _ (C) / *¯8 _ (V) (DEDR 910; Zvelebil 1977, 35) = *e¬-u- / *¯¬- (Krishnamurti 2001, 63) = *jd8- (G. Starostin: vocalization after Gondwan *jd0- while *-¯- in other branches should have been influenced by the following numeral *ez- "8"; in his dissertation Starostin 2000, #350 reconstructs *¯, i.e. *j¯8-, in his transcription *j¯0-).
*erru / *ez (DEDR 784; Zvelebil 1977, 35) = *ez-(-nr-) (G. Starostin).
*on-/or-paCtu (DEDR 1025).
*toJ-(paC-)tu (DEDR 3532) = *toJ-pad- (G. Starostin).
*paCtu (DEDR 3918) = *paT- (G. Starostin) = *paH- & ntr. suf. *-tu (Krishnamurti 2001, 328).
*n¨r(-tu ) (DEDR 3729) = *n¨d- (G. Starostin).
Telugu v¯yi, veyi, veyyi, pl. v¯lu "1000", v¯na-v¯lu "thousands by thousands", cf. Tamil viyam "extensiveness, height", viyal "greatness, width, expansion", viyan "greatness, vastness, excellence", Malayalam viyam "extension", Gondi weeya "high" (DEDR 5404).

Paper Giver 3: Lars Martin Fosse, University of Oslo, Norway

Paper 3 Title: The Sanskrit absolutive as a fuzzy form

Paper Abstract: The Sanskrit absolutive has been discussed regularly for the last 150 years or so, the topics being the status of the form within the verbal paradigm, its voice, its agent, and its syntactic role.
a) This paper will discuss the voice of the absolutive in connection with its agent, and
b) Also take a closer look at the way the absolutive functions syntactically in the sentence.
a) It has been proposed that the absolutive is partly indifferent to voice, and that it may have a passive interpretation. I shall argue that the absolutive is inherently active, and that the combination of an active verbal form with a passive main verb creates a theoretical problem that is difficult to solve with standard syntactic theory.
b) It is clear from the way the absolutive is used that it functions as a converb. However, it can also be used in a clause-chaining construction which is neither argemental nor adnominal. This raises the question whether the Sanskrit absolutive can also be regarded as a medial verb in addition to functioning as a converb. In many languages, medial verbs are used as a means clause-chaining. However, the absolutive does not behave exactly like a medial verb in all respects. Rather, it would appear that the absolutive adopts the role of a medial verb on some occasions, while adopting the role of a converb on other occasions.
As a form, the absolutive in itself is fuzzy both in the semantic sense and in the syntactic sense. It is extremely flexible in use, and therefore rapidly grew popular, contributing to the erosion of the Vedic verbal system which lead to the simpler, more straightforward but less precise verbal usage of the post-Vedic period.

      Full paper to be downloaded (as a pdf-file)

Paper Giver 4: Vit Bubenik, Memorial University of Newfoundland, St. John’s, Canada

Paper 4 Title: On the Evolutionary Changes in the Middle and New Indo-Aryan Systems of Case and Adpositions

Paper Abstract: The Old Indo-Aryan case system, outlined as a starting point, will be contrasted with those of Old (Pali), Middle (Ardha-Magadhi) and Late Middle Indo-Aryan (Apabhramsa) case systems to demonstrate the gradual erosion of the synthetic morphology of case during the Middle Indo-Aryan period. The trade-off between the restructured nominal case system and the evolving system of adpositions during the MIA period will be discussed in a systemic fashion (along the lines of Bubenik (1996) and (1998)). It will be shown that Late MIA ended up with only one form for an earlier nominative vs. accusative, instrumental vs. locative, and genitive vs. ablative. Given this drastic reduction from the seven fusional cases of OIA to four by the end of the MIA period, the adpositions grew steadily in importance in denoting the relational aspects of their head nouns. (This will be demonstrated for the notions of source, location, and accompaniment/instrumentality where the postpositions supplied unambiguous markers compared with the indistinctive case morphology of ablative, locative and instrumental).
In Early NIA the four cases of Late MIA were further revamped (Nom/Acc > Direct, Gen/Dat/Abl > Oblique, Instr > ergative postposition, Loc > locative postpositions). The system of postpositions was further developed whereby the simple (primary) postpositions express the basic topological notions (‘at’ - ‘of’, ‘to’ - ‘from’, ‘up’) and the projective notions of interiority/exteriority, anteriority/posteriority, superiority/inferiority, accompaniment and circumference are expressed by secondary postpositions whose adverb is preceded by the oblique form of the genitive postposition ka). Some important typological similarities and differences between Hindi and European Romani will be outlined. Most remarkably Romani differs from all the other NIA languages in realizing the layer of secondary postpositions by prepositions. This typological change in the structure of the postpositional phrase was accompanied (conditioned ?) by the overall shift from SOV to SVO word order.

Paper Giver 5: Zakharyin, Boris A., Moscow State University, Moscow, Russian Federation
Paper 5 Title: Pa:Nini’s ‘Lopa’ and Zeroing in Modern Hindi
Paper Abstract:
In ‘AshTa:dhya:yi:’ Pa:Nini suggests the following general definition of zero replacements: A 1.1.60: adar_anaM lopaH ‘(The term) ‘lopa’ (signifies) ‘invisibility/disappearance’ (of a language unit)’. This means that the corresponding unit at first appears in full in its place in an utterance, and only afterwards it becomes zeroed. The process might be exemplified by su:tra A 1.3.9: tasya lopaH ‘Zero replacement of that (marker is to take place)’, which implies an unconditioned and complete ellipsis of a technical marker ‘IT’.
More specific, more determined by context and structurally more important are subdivisions of lopa (described as -lu- by Pa:Nini) that signify zero replacements of affixes alone: A 1.1.61: pratyayasya luK _lu luPaH ‘luK’, ‘_lu’ (and) ‘luP’ (denote different types of zero replacements of) affixes’.
The spheres of applying of each of these three subdivisions are not determined unequivocally by Pa:Nini, though it is clear that the use of luP is restricted to taddhita-affixes exclusively. - It should be reminded in this connection that, according to Pa:Nini, all affixes (pratyaya) belong to either of the two groups: to ‘kRt’ (see A 3.1.93), serving to derive from nominals or verbs more or less regular nominal bases (pra:tipadika), or to ‘taddhita’, deriving - generally with irregularity - secondary nominal bases from nominal or verbal elements, or even from pairs of words related semantically and syntactically (characterized by samartha). As has been rather convincingly demonstrated by S.D. Joshi and J.A.F. Roodbergen, the whole of taddhita-section in Pa:Nini’s grammar (including rules of luP-elision) may not belong to the original text of ‘AshTa:dhya:yi:’.
Contrary to luP, the other two affixal zero replacements may apply to either kRt or to taddhita. The same is true also for lopa, that occurs not only with affixes but with any language elements, including phonemes or phoneme-sequences. The difference in use between luK, on one side, and _lu and lopa ,on the other , has been clearly shown by G. Cardona: any operation that is allowed to take place after a deletion of a certain affix does not take effect in case the zero replacing this affix is denoted by luK. For example, in deriving the nominative singular masculine form go-ma:n from the base go-mat- ‘someone having many cows’ we apply lopa to the universal affix -sU (of Nom. sg. m. from stems in -a) at the second step of the derivation: gomat- + -sU > gomat-s; gomat-s + lopa (in accordance with A 6.1.68) > gomat-. And this lopa, after it happens, does not prevent any of the rest of operations necessary for the derivation process: gomat- > goma:t (by A 6.4.14) > goma:nt (by A 7.1.70) > goma:n (by A 8.2.23). Contrary to this, luK when applied to the same affix -sU at the same (second) step of the derivation, prohibits the last three operations that follow luK. Thus, we get only the stem gomat- (< gomat- + -sU) that may serve as a part of compounds - e.g., in gomat-priya- ‘one who likes someone with cows’. Otherwise (when using lopa instead of luK) we ‘ll get wrong compounds *goma:t-priya-, *goma:nt-priya-, etc.
As for _lu, it normally denotes zeroing (usually accompanied by vowel alternations) of the final constituent in the first syllable of the reduplicated stem. The most typical examples imply zeroing (accompanied by stem-reduplication in accordance with A 2.4.75) of the thematic universal suffix -a- (PaNini’s ‘_aP’, first introduced and later deleted) after verbal roots of the 3-d ("reduplicated") class beginning with hu- ‘to offer oblations’: hu- + _aP + tiP > hu- + _lu + ti P > hu-hu- + # + tiP >... > ju-ho-ti ‘offers oblation’, - or after nij- ‘to wash, bathe’ beginning another set of the "reduplicated" class of verbs: nij- + _aP + tiP > nij- + _lu + tiP > ni-nij + # + tiP > ... > ne-nek-ti ‘washes’.
Zeroing of an affix is conditioned by two metarules: A 1.1.62 pratyayalope pratyayalakshaNam ‘After (the whole) affix has been replaced by zero the operation conditioned (by it still) takes effect (as if that affix is manifested)’. But the next su:tra A 1.1.63 provides important restrictions for cases when zeroing is denoted by any term with -lu- (lumata:): na lumata:Mgasya ‘(In case zeroing of an affix is denoted by a term) with -lu- (and the affix has been deleted, an operation, formerly applicable to its) stem (in the presence of that affix, now) does not (take place)’ - see examples with gomat-s + lopa > ... goma:n, but gomat-s + luK > ... gomat- above.
Pa:Nini’s refined system of zero replacements may also be applied to Hindi data. For example, while combining with the root laR- ‘to be militant; to make war’, the first vowel of the ‘taddhita’-suffix -ak- (signifying a person characterised by activity ‘laR-’) is obligatorily deleted before regular inflections in word-changing paradigms (e.g., laR-#k-a: ‘a boy’, laR-#k-e ‘boys’, etc.) but remains intact before the other ‘taddhita’-suffix of abstract nouns -pan ( laR-ak-pan ‘boyhood; youth’). The zero replacements of the first type might be denoted by the term ‘lopa’, while the second type occurences might be described as, first, introducing the same regular inflections (-a:, -e, etc.) and then dropping them through ‘luK’ replacement that prohibits elision of the first vowel of suffix -ak: laR- + -ak + -a:/ -e ... + pan > laR- + -ak + luK + -pan > laR-ak-# -pan > laR-ak-pan. Such, imitating Pa:Nini’s, approach to derivation (when at first step inflections -a:, -e, etc. are inserted after -ak) demonstrates irrefutable preference over other models as it also incorporates the other variant of the same lexeme: the form laR-#k-e-pan ‘boyhood; youth’ where delition of -a in -ak might be described through either lopa or _lu.
The zeroing procedures, suggested by Pa:Nini, also bring to life many a question important for Hindi morphology, for example: (1) As there exist a:-y-a: ‘came’ from a:- ‘to come’ or pa:-y-a: ‘acquired’ from pa:- ‘to acquire’, shouldn’t we posit the same affix -y- also in verb-forms like pahuMc-y-a: ‘arrived’ from pahuMc- ‘to arrive’ implying zero replacement: pahuMc- + -y- + -a: / -e ... > pahuMc- + lopa + -a: /-e ... > pahuMc- # -a: /-e ...> pahuMc-a: / e ... ? (2) How many ‘null-elements’ are we to posit in declensional patterns of the type ghar- ‘a house’ ? - As there exists an opposition between ghar (a) ‘a house’; (b) ‘houses’; © ‘inside the house’ ( in ghar + meM) and ghar-oM + meM ‘inside the houses’ should we postulate three different ‘nulls’ or only one as opposed to -oM (Obl. pl.) ? (3) Are we allowed to posit derivational zero in bakr-#-a: ‘ he-goat’ as there exists bakr-i:-# ‘ she-goat’, and - again - how many of those zeroes the declensional pattern of the latter word implies: two or only one ? (4) Can we apply rules like Pa:Nini’s su:tras A 2.4.75; 6.1.10 when deriving forms of the type ‘calte calte’ ‘while moving continuously’ (in sentences like ra:m calte calte gir paRa: ‘In the process of continuous movement Ram (suddenly) fell down’) ? - That is, calta: (hua:) Ra:m + REDUPLICATION + ADVERBALIZATION + MOVEMENT > Ra:m calte hue calte hue + _lau (in accordance with A 6.1.10) > Ra:m calte calte ... ? - These and similar problems are to be discussed and - probably - solved in the forthcoming paper.

Paper Giver 6: Annie Montaut, INALCO, Paris, & CEIAS/EHESS, Paris, France

Paper 6 Title: Intransitivity in Hindi: verb morphonology, semantic roles and argumental structure

Paper Abstract: Hindi, as well as other Indo-Aryan languages is characterized by a verbal lexicon where basic forms are intransitive, the transitive being derived from it, whereas an important majority of the world languages derive, the other way round, intransitive from basic transitives or causatives: middle in ancient languages is derived from active which is considered as the basic form, and the so-called modern middle forms of modern Roman languages (Fr "se casser, s’ouvrir", Sp "rumpir se, abrirse", are also derived from the corresponding transitive. Among the some thirty verbal bases listed by Haspelmath (1993) as representative of the most elementary predicates, Hindi has a large majority of basic intransitives of the middle or medio-passive type (TuuTnaa "break-int", toRnaa "break-tr", khulnaa "to open-int", kholnaa "to open-tr"), whereas French has a large majority of basic transitives, with little relevance of semantic roles, and English a majority of equipollent verbs ("to break, to open").
I propose to explore the class of Hindi intransitives in their relation to the morphologically related transitives/causatives in order to inquire in the relation between argumental structure, semantic roles and morpho-phonological shape of verbal bases. I will study in particular the ability of a sub-class of intransitives to passivize, an ability linked to their argumental structure (+agent) and the ability of a distinct sub-class of intransitives (- agent) to allow an optional extra argument with very similar modal meanings as the passivable intransitive:
mujhse calaa nahiin gaya
I-instr walk neg passive
I could not (bring myself to) walk/go
mujhse darvaazaa khul nahiin rahaa thaa
I-instr door open-intr neg progr impft
I could not (manage to) open the door
The semantic distinction of both structures, as the third semantically "special" meaning of another sub-class of intransitives (inadvertent process in non negative contexts), will be accounted for in terms of the semantic roles involved in the argumental structure of such predicates.
mujhse gilaas TuuTaa
I-instr glass broke-intr
I broke the glass by mistake (inadvertently, unconsciously)
I will then try to revisit some of the main semantics roles as exhibited by Hindi morpho-syntax, specially the role agent, present in some intransitives and all transitives providing we refine the features associated to the role in order to account for such alternances and contrasts as (cf. Montaut 2004, in Oblique Subjects in SA languages, Benjamins):
mai mahsuus kar rahaa thaa vs mujhe mahsuus ho rahaa thaa
I feeling do progr impft I-dat feeling be progr impft
I was feeling
mujhe tumse irSyaa thii par us samay uskaa bodh nahiin thaa
I-dat you-from jealousy was but that time of-it consciousness neg was
I was jealous from you but at that time I was not aware of it
*main tumse irSyaa kartaa thaa par uskaa bodh nahiin thaa
I you-from jealousy do impft but of-it consciousness neg was
maine Siitaa ko rote hue paayaa vs maine apne ko khoye hue paya
I-erg Sita acc crying being found, I-erg refl acc lost being found
I found Sita crying, I found myself lost
mujhe Siitaa rotii huii milii * mujhe main/apna khoyaa huaa milaa
I-dat Sita crying being be-found * I-dat I/refl lost being be-found

Paper Giver 7: L.V. Khokhlova, Moscow State University, Moscow, Russian Federation

Paper 7 Title: Resultative Constructions with overt Agent in Western NIA Languages

Paper Abstract: There are three types of resultative constructions with overt agent in Western NIA languages:
1) Objective resultative: stative participle refers to Object in Nominative case, controlling verbal agreement. Agent is in the Genitive case.
H. yah ciTThii merii likhii huii hai «This letter is written by me».
2) Possessive resultative: stative participle refers to Agent in Nominative (Marwari, standard Hindi, restricted to several verbs in Gujarati), Ergative (colloquial Hindi, Punjabi), Instrumental (Gujarati) cases. Verbal agreement is controlled by Agent (Marwari) or Object (Punjabi, Gujarati, colloquial Hindi). The Object NP usually denotes a part of Agent, something belonging to Agent or something being in direct contact with Agent. The result of the performed action affects the subject (possessor) more than the object.
G. gopale kaalii Topii paherelii che "Gopal is wearing a black cap".
3) Resultative with the Promoted Possessor (Possessor-to-Proleptic-Element-Promotion): Agent is marked by the Genitive postposition in the fixed masculine oblique case form, the verb agrees with Object. Similar to Object in Possessive resultative, Object NP in Resultative with the Promoted Posessor usually denotes something being in direct contact with Agent. Unlike the first two resultative constructions, resultative with the promoted possessor is possible only in case of the visuality of the stationary condition. Construction of this type is used only in Punjabi:
P. muNDe de niilii pagRii vajjhii hoii hai "As for the boy, he is wearing a blue turban". Unlike the first two resultative constructions, it is possible both with transitive and intransitive verbs.
Raising transformation results in Agent promotion in case of the possessive resultative and in Object promotion in case of Resultative with the Promoted Posessor, compare:
us nUU banduuk phaRiAA vekhke asII Dar gae "Having seen him with the rifle we got frightened" and
us de banduuk phaRii vekhke asII Dar gae "Having seen a rifle kept by him we got frightened".
Resultative constructions in Western NIA languages may be compared with similar constructions in different languages of the world. Possessor-to-Proleptic-Element-Promotion exists in Japanese. Russian dialects are rich in objective and possessive resultative constructions of various formal structures. They will be discussed in my paper when the comparisons with the resultative constructions in Western NIA languages are made.

Paper Giver 8: Indira Y. Junghare, Institute of Linguistics, ESL and Slavic Lgs., Minneapolis, USA

Paper 8 Title: Borrowing, Code-Mixing, Code-Switching and Code-Shift in Diasporic Marathi

Paper Abstract: In this paper, the term code-switching is employed in the sense of the alternative uses of two languages, Marathi and English, either within a sentence or between sentences. Code-switching within a sentence includes borrowing and code-mixing, while intersentential code-switching seems to include all the three, borrowing, code-mixing and code-switching. The inter-sentential code-switching often trigger code-shift. Our study of the Marathi-English language alternation is based primarily on spontaneous speeches used by adult members of the first generation Marathi-speaking community to each other in casual or semi-formal conversations and intergenerational conversations, i.e. between bilingual parents and their children in an intimate family situation.
The study includes an examination of extra-sentential and intra-sentential code switching, and their relation to various variables, i.e. extra-linguistic factors: sex, age, education, social network membership, ethnic identity, and occupation. The purpose behind this is to see what variables are capable of accounting more generally for patterns of language choice.
The code-switching can be studied from a number of perspectives: the grammatical, the socio-linguistic and the conversation analytic. From the grammatical perspective, a number of restrictions on code-switching within the sentence have been formulated ( Clyne, 2000; Gumperz, 1982; Poplack, 1978/81; and others). However, grammatical restrictions do not tell us anything about the interactional value or meaning of intra-sentential or extra- sentential code-switching. The same thing holds true for socio-linguistic perspective. General statements are made about the distribution of code switching in certain situations or among participants holding certain 'roles' and 'statuses' in a given society (Auer, 2000). It says little or nothing about the contribution of code-switching to the ongoing interaction, that is, about its local functioning. Although the value of both the approaches cannot be denied, they need to be incorporated into a third perspective which is to investigate the contribution of code-switching to community members' sense making activities.
Our analyses of social and family conversations, taking into account the grammatical restrictions when necessary, will relate to larger sociolinguistic statements. Some fundamental distinctions that are relevant for the production and interpretation of intra-sentential and extra-sentential code-switching in conversation will be presented in this paper. The social content behind the linguistic forms of borrowing, code-mixing, code-switching and code-shift will also be discussed. It will be shown that these forms form a continuum, borrowing being on the far left and code-shift on the far right of the continuum.

Paper Giver 9: John Peterson, Universität Osnabrück, Osnabrück, Germany

Paper 9 Title: Parts of speech in Kharia (Munda, Austro-Asiatic)

Paper Abstract: The Munda languages are well-known for the fact that the noun-verb distinction, if one exists in these languages, is very weak.
Typical of Munda languages is the so-called "precategoriality", i.e., any lexeme or pro-form can function as either a predicate or as the complement of a predicate. Consider the following examples which at least for many speakers of Kharia are perfectly grammatical utterances and in which the "nouns" and "pronouns" function both as complements and predicates:
(1) bhagwan lebu-ki-Ø ro Del-ki-Ø. lebu Del-ki-_.
God man-M.PT-S and come-M.PT-S man come-M.PT-S
'God became man (= Jesus) and came [to earth].' 'The / a man came.'
(in a play about me and you, in which both of us will be taking part):
(2) naTak-te iny-ga ho-kaR-na-iny ro am-ga iny-na-m.
play-OBL 1S-FOC 3-S.HUM-M.IRR-1S and 2S-FOC 1S-M.IRR-2S
'In the play I will be him and you will be me.'
In addition, "nouns" marked for genitive case can function as predicates. This is a fully productive process: ("?" = glottal stop)
(3) iny ho-kaR-te iny-a?-y-oj. am ho-kaR-te am-a?-y-ob.
'I adopted him/her (i.e., I made him/her mine).' 'You adopted him/her.'
But there is more to this issue in Kharia than merely "precategoriality". For example, entire "NPs" can function as predicates:
(4) ho rochob-ki-ny. ho rochob-te col-ki-ny.
that side-M.PT-1S that side-OBL go-M.PT-1S
'I moved to that side.' 'I moved to that side.'
(5) bharat-ya? lebu-ki bides-a? lebu-ki-y-a? ruprang-ki-may.
India-GEN person-P abroad-GEN person-P-y-GEN appearance-A.PT-3P
'The Indians took on the appearance of foreigners (e.g. by living abroad so long).'
The only restriction on this system is "world knowledge". For example, Tebul 'table' is virtually restricted to its function as the complement of a predicate, since things seldom 'become a table', which this would mean in the middle voice, nor are things commonly 'turned into a table' (active voice). Nevertheless, when a proper context is found (e.g. fairy tales), this predicative use is possible for any lexeme.
In my proposed talk I will address the issue of parts-of-speech in general in Kharia. I will show that from a purely language-internal perspective, lexical classes such as adjective, noun and verb have no place in Kharia grammar.
Instead, Kharia grammar makes direct use of functional / syntactic concepts such as "predicates" and their "complements", both of which are marked morphologically by what may be termed functional heads. These consist of purely grammatical markers: in the case of complements, these consist of markers for inalienable possession, case and number markers, in the case of complements, TAM, voice and person markers. It is these functional heads which signal the function of a phrase in a clause as being either a complement or a predicate. What is important is that all types of lexical heads, whether they refer to a physical entity or to an action, may occur with both types of functional heads.

Paper giver 10: Estella Del Bon, Ecole Pratique des Hautes Etudes, Paris, France

Paper 10 Title: Reference and Non-Reference in Impersonal Clauses of Kashmiri

Paper Abstract: For situations that cover the expression of natural phenomena and mental/physical affections (and that as such share the characteristic of lacking a participant that can potentially assume agentive properties) Kashmiri uses two different grammatical constructions. Either the clause is intransitive (ex.1) either it is impersonal (ex.2). Intransitive clauses are dominant in speech and are the unmarked form, whereas impersonal clauses occur rarely and have a marked aspectual value : they indicate a change of situation (with implicit reference to the prior situation) .
ru:d pyov
rain.MS-Š fall.PTMS
It rained.
ru:d koru-n
rain.MS-Š make.PTMS-III3S
It rained.(lit. It made rain.)
(before it was not raining, now it is raining)
Structurally impersonal clauses are transitive and show on the finite verb a third person singular marker (ex.2 : personal suffix of set III -n) that cannot be made explicit by any (pro)-nominal form and that has no referent, though with transitives it usually has the form of an agent-agreement marker. Furthermore, it turns out that part of these impersonal clauses can, as an alternative, have an additional non-referential marker on their finite verb (ex.3 : personal suffix of set IV -s). This marker is a third person singular personal suffix that is usually used for dative case marked arguments.
ru:d kor-n-as
rain.MS-Š make.PTMS-III3S-IV3S
It rained.(lit. It made rain to it.)
The present paper aims at answering the following questions :
- Why can’t this additional non-referential marker -s occur with all impersonal clauses ?
- Can this additional non-referential marker be assigned any particular semantic value or semantic role ?
- And finally, for clauses that do accept this marker, to what opposition corresponds the alternation presence vs absence of the marker (for instance, example 3 vs example 2)?
The study of this particular marker is of interest to the typology of personal suffixes and impersonals and shows how reference and non-reference interact with semantic roles and semantic properties of situations in a language like kashmiri. Furthermore, since impersonal constructions and complex personal suffixes systems are not characteristic of modern Indo-Aryan, and since impersonal constructions similar to that of Kashmiri have been reported in languages such as Shina, Kalasha, Burushaski and Limbu, the present study is of particular interest to areal typology.

Paper giver 11: Ruth Schmidt, Department of East European and Oriental Studies, University of Oslo, Norway

Paper 11 Title: Compound verbs and modality in the Shina of Pakistan

Paper Abstract: Verb sequences consisting of conjunctive participle + inflected verb, usually called "compound verbs" are one of the true innovations of New Indo Aryan (Masica 1991: 326). These verb sequences behave as a unit, in which the first member of the sequence contributes the basic lexical meaning, and the second (partially emptied lexically) specifies the manner of an action or event. The second verb has many designations; I refer to it as a "vector verb".
Compound verbs occur in Shina, although they are a less conspicuous feature of the language than in Hindi or Panjabi. Five vector verbs have been found, which meet the usual tests for compound verb status: "fall", "go", "sit", "leave" and "give". This paper provides examples of all five vectors and describes the type of manner specification which each contributes. Where relevant, Kashmiri compound verb constructions will be compared.
Modality is a more multiform category than compounding. Only one free modal verb has been found ("to finish"), which patterns like a vector verb but which can be differentiated by semantic tests. Other types of modality are expressed variously by a bound modal verb ("be able"), or an impersonal construction ("should"). Examples of modal constructions will also be given.

Back to SASNET

Search the SASNET Web Index

SASNET - Swedish South Asian Studies Network/Lund University
Address: Scheelevägen 15 D, SE-223 70 Lund, Sweden
Phone: +46 46 222 73 40
Webmaster: Lars Eklund
Last updated 2008-01-24