The case of addressees in Dravido-Portuguese

PAPIA, São Paulo, 24(2), p. 307-342, Jul/Dez 2014. The case of addressees in Dravido-Portuguese O caso dos interlocutores em Drávido-Português Hugo Cardoso1 Universidade de Lisboa, Portugal [email protected] Abstract: This article is a study of the Dravido-Portuguese creoles of the Malabar (modern-day Kerala, India) and Sri Lanka in terms of the formal means they use to mark the addressee arguments of various predicates of verbal interaction (e.g. talk, say, ask), and compares the functional range of such markers in the creoles with their counterparts in the lexifier language (Portuguese) and in the Dravidian adstrates (Malayalam and Tamil). This comparative study shows the different contributions of the lexifier and of the adstrates to the case-marking system of the present-day Dravido-Portuguese creoles, as well as the ways in which the functional range of the creole case-markers reveal diachronic processes of functional reanalysis. Keywords: Dravido-Portuguese creoles, addressees, casemarking. Resumo: Este artigo consiste num estudo dos crioulos drávidoportugueses do Malabar (actualmente Kerala, na Índia) e do Sri Lanka em termos dos meios usados para marcar formalmente os interlocutores de diversos predicados de interação verbal (ex. conversar, dizer, pedir), no qual se compara a amplitude funcional dos marcadores crioulos com a dos seus equivalentes na língua lexificadora (o português) e nos adstratos dravídicos 1 I would like to thank Ian Smith for kindly sharing with me unpublished Sri Lanka Indo-Portuguese data, and also an anonymous reviewer for his/her comments and suggestions. ISSN 0103-9415, e-ISSN 2316-2767 308 Hugo Cardoso (o malaiala e o tâmul). Este estudo comparativo demonstra os diferentes contributos do lexificador e dos adstratos para o sistema de marcação casual dos actuais crioulos drávidoportugueses, bem como o modo como a amplitude funcional dos marcadores casuais crioulos revela processos diacrónicos de reanálise funcional. Palavras-chave: Crioulos drávido-portugueses, interlocutores, marcação casual. 1 Introduction In Schuchardt’s (1889a) classification, Dravido-Portuguese (Dravidoportugiesisch) referred to the subset of Portuguese-lexified creoles formed in the Dravidian-majority areas of Southern India and Northeastern Sri Lanka. Although Indo-Portuguese creoles were once numerous across the Dravidosphere (Schuchardt 1889a; Tomás 1992; Smith 1995; Cardoso 2006), to the best of my knowledge they subsist only in Cannanore [Kannur] in the Indian state of Kerala, and in the Sri Lankan cities of Batticaloa and Trincomalee. Smith’s (1979a,b) research on the Indo-Portuguese of Batticaloa has shown that, in many domains, this language has come to resemble significantly the major language of the region, Tamil, with which it has been in close contact throughout its history (see also Bakker 2006) - and recent descriptive work on the Indo-Portuguese of India’s Southwestern coast reveals a comparable (though not entirely equivalent) process of metatypy under the influence of the dominant local language, Malayalam (Cardoso 2012). Among various examples of isomorphism, Smith (1979a) highlights the similarity in the functional range of case-markers in Sri Lanka Indo-Portuguese and in Tamil. This article focuses on a particular aspect of case-marking in the Dravido-Portuguese creoles, namely the assignment of case to Addressee arguments, arguments which are associated with verbs of verbal interaction such as talk,say, or ask (see 1.2). Addressee arguments can be treated in different ways in different languages and, as a matter of fact, constitute one domain in which the languages involved in the contact equation which gave rise to the Dravido-Portuguese Creoles (primarily Portuguese, Malayalam and Tamil) diverge quite markedly. I have therefore selected this particular domain of case-marking in order to explore the extent to which the DravidoPortuguese Creoles reveal the impact of the Dravidian languages among which they developed, innovate, or retain case-marking patterns modelled on their main lexifier (Portuguese). PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 The case of addressees in Dravido-Portuguese O caso dos interlocutores em Drávido-Português. . . 309 1.1 Dravido-Portuguese sources While some early data is available for Dravido-Portuguese creoles elsewhere (e.g. Teza 1872; Schuchardt 1882, 1883, 1889b; Dalgado 1900, 1917; Tavares de Mello 1908; Jackson 1990; see also Tomás 1992; Ladhams 2009), the only ones which have been documented in recent times are those of Batticaloa (studied in the 1970s by Ian Smith), Cannanore [Kannur] and the recently-disappeared creole of Vypeen, off Cochin [Kochi] (documented by myself since 2006). It is on the basis of these recent studies and corpora that I set out to explore the case-marking patterns of addressees in the Dravido-Portuguese creoles: • The Batticaloa variety described by Smith (1977, 1979a, 1979b, 2013), the only one to have been researched in recent times, is taken as representative of Sri Lanka Indo-Portuguese [slip]. As a matter of fact, plenty of the earlier records of Sri Lanka Indo-Portuguese refer to other geographical variants and therefore, in theory, one could ascertain to what extent patterns of case-assignment varied across the island in the past; however, since there are reasons to question the authenticity and reliability of a good deal of the early Sri Lankan corpus (see Smith 1977; also Cardoso, Baxter & Nunes 2012), the focus here will be mostly synchronic; • The corpora recently collected in Cochin and in Cannanore are largely equivalent, therefore we will refer to both varieties generically as Malabar Indo-Portuguese [mip], after the old name for the coast of modern-day Kerala. Here too, despite the availability of short 19 th -c. collections of texts, we will restrict our analysis to the modern data. Given that the Dravido-Portuguese creoles have no conventionalized orthography, the reader will encounter substantial variation. In the case of Sri Lanka Indo-Portuguese, I have retained the spelling developed by Smith for his published descriptions. In the case of Malabar Indo-Portuguese, given that linguistic documentation is an ongoing process, no final phonologicallybased orthography has yet been developed; therefore, I resort to phonetic transcription here. With respect to example sentences in other languages, I use either the official orthography (in the case of Portuguese) or the romanisation adopted by the authors of the source grammars (for Malayalam and Tamil). PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 310 Hugo Cardoso 1.2 Object of inquiry and method Predicates of verbal interaction or ‘linguistic action verbs’ (Verschueren 1981), such as Eng. talk,tell or ask, may be monotransitive or ditransitive. Their valency requires an argument which I will term the ‘Speaker’ [abbreviated as ‘S’] and one corresponding to the ‘Addressee’ [abbreviated as ‘A’], and often also an indication of the content of the verbal interaction (in whatever form) which I will call the ‘Message’ [abbreviated as ‘M’]. The following English sentences illustrate this classification: (1) a. Louise was talking to/with her mother. S A b. Louise told her mother the news. S A M c. Louise asked her mother a question. S A M Even though the argument her mother is identified as the addressee in all of the sentences above, it can be conceptualised differently in terms of its semantic role: notice, for instance, how in (a), the Addressee can be treated either as a Companion (marked by the preposition with), thereby expressing reciprocity, or as a Recipient/Goal (marked by to). In fact, it is one of my aims to ascertain the default semantic role expressed by the case-marker(s) assigned to Addressee arguments in the various languages under analysis. Therefore, to try and capture a good deal of the potential variation within a particular language, I have collected data for three pragmatically-defined categories of predicates of verbal interaction: • The talk-type: types of verbal interaction in which both the Addressee and the Speaker are agentive interlocutors, i.e. predicates of reciprocal verbal engagement (includes e.g. talk,chat,discuss,argue,...); • The tell-type: types of verbal interaction in which the Addressee is conceptualised as a passive recipient of the Message and not necessarily expected to react (includes e.g. tell,say,inform,convey,. . .); • The ask-type: types of verbal interaction which are aimed at eliciting a reaction from the Addressee, whether it be in the form of verbal act (includes e.g. ask,...) or otherwise (includes e.g. demand,request,. .. ). PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 The case of addressees in Dravido-Portuguese O caso dos interlocutores em Drávido-Português. . . 311 In order to conduct this study, relevant data was collected not only for Malabar Indo-Portuguese and Sri Lanka Indo-Portuguese, but also for their respective adstrates (Malayalam and Tamil) and for their main lexifier (Portuguese). As the intention was to be able to compare the functional range of case-markers involved in Addressee-marking in each language, the research proceeded in three steps: Step 1: I collected example sentences for all five languages containing predicates of verbal interaction of the three types described above; Step 2: For each of the five languages, I identified the default semantic roles expressed by all of the case-markers compiled in Step 1; for the purposes of this particular study, I will label these the associate semantic roles (as they are associated with the case markers attested for Addressees). Overall, seven associate semantic roles were identified in step 2: Patient, Recipient, Companion, Instrument, Temporary Possessor, Location, Goal; Step 3: I collected example sentences for all five languages containing arguments with all seven associate semantic roles. The results are given in section 2 below, which includes examples of all three types of predicates of verbal interaction and of all seven associate semantic roles for each of the five languages. The data are then compared and discussed in section 3, revealing a complex interplay of influences. This study provides ample evidence that, in high-contact languages, the phonetic representation of a particular (grammatical) element and its function need not be sourced from the same donor language nor match any entirely. It therefore confirms the insights of language contact scholars who have argued for the need to distinguish between the transfer of linguistic forms and the transfer of semantic or functional range – what Matras (2009) terms matter (also fabric, in Grant 2012) and pattern respectively. 2 Case-marking of addressees and associate semantic roles 2.1 Malabar Indo-Portuguese The corpus of Malabar Indo-Portuguese which I have been collecting in Cannanore and Vypeen in the past years is composed of elicited material and recordings of spontaneous discourse. As mentioned earlier, the recordings made in Vypeen and in Cannanore are equivalent to a large extent, and therefore they will be treated here as complementary; nonetheless, for clarity, I indicate where and how each example sentence was collected. Although the treatment of this corpus is still ongoing, it was possible to extract several instances of both Addressee arguments and participants with the associate semantic roles. PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 312 Hugo Cardoso 2.1.1 Addressees in Malabar Indo-Portuguese In Malabar Indo-Portuguese, Addressee arguments of talk-type predicates (Addressee TALK ) most often consist of an np followed by the postposition pært @ ‘near’ (2), which requires a linking morpheme -s @ that also functions as a default genitive marker. However, elicitation reveals that -j˜utad @ , typically a comitative/instrumental marker, is a possible alternative (3b): (2) Cannanore Indo-Portuguese (Cardoso: Elicited corpus) pærmi 1s.obl bOs-s@ you-gen mãy-s@ mother-gen pært@ near kæra want kõb5rsa, talk bOs-s@ you-gen pært@ near nO not ‘I want to talk to your mother, not to you.’ (3) a. Cannanore Indo-Portuguese (Cardoso: Elicited corpus) Stanley Stanley portuguese portuguese l˜ ıgwaji language Yvonne-s@ Yvonne-gen pært@ near t5 be.prs kõb5rsa. talk b. Cannanore Indo-Portuguese (Cardoso: Elicited corpus) Stanley Stanley portuguese portuguese l˜ ıgwaji language Yvonne Yvonne j˜ utad@ com t5 be.prs kõb5rsa. talk ‘Stanley is talking in Portuguese with Yvonne.’ All other addressees in the corpus occur with the postposition pært @ . This is exemplified in (4) for Addressee arguments of tell-type predicates (Addressee TELL ), and in (5) for those of ask-type predicates (Addressee ASK ); notice that, in (4) and (5b), the fact that the 1spronoun has a dedicated Genitive form does away with the need for the linking morpheme -s@: (4) Cochin Indo-Portuguese (Cardoso: Oral corpus) minh@ 1s.gen pært@ near fala say ki what keræ. want ‘Tell me what you want.’ PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 The case of addressees in Dravido-Portuguese O caso dos interlocutores em Drávido-Português. . . 313 (5) a. Cochin Indo-Portuguese (Cardoso: Oral corpus) kawa then el 3sm j@ pst fala say pay-s@ father-gen pært@ near pis˜ ın little fana money tama take da. give ‘Then he told father to give a little money.’ b. Cannanore Indo-Portuguese (Cardoso: Elicited corpus) Olivia-s@ Olivia-gen filh@ daughter minh@ 1s.gen pært@ near j5 pst f˜ uta ask præla 3sm.obl pu to asisti. help ‘Olivia’s child asked me to help her.’ 2.1.2 Associate semantic roles in Malabar Indo-Portuguese In Malabar Indo-Portuguese, the case-marking of patients responds to animacy. As shown below, animate participants are marked with postposed -p @ (6a), while inanimate participants are assigned no case-marker at all (6b): (6) a. Cochin Indo-Portuguese (Cardoso: Oral corpus) ælfõti-p@ elephant-obl j@ pst faze make tir@ shot ‘[We] shot elephants.’ b. Cochin Indo-Portuguese (Cardoso: Oral corpus) ag@ water friw cool ker@ want faze. make ‘[You] must cool the water.’ In (6a), -p @ is glossed obl[ique] because, in addition to marking animate patients, it also marks recipients (7): (7) a. Cochin Indo-Portuguese (Cardoso: Oral corpus) ak@l dem ja pst da give s@, cond paymi 1s.obl ˜ uga one mil thousand rupi rupee lo irr da. give ‘If [I] delivered it [i.e., a skin], [they] would give me one thousand rupees.’ b. Cochin Indo-Portuguese (Cardoso: Oral corpus) paymi 1s.obl t@ be.prs kãd@ when acha, get olot@-p@ 3p-obl t@m˜ e also lo irr da give kov@ri. money ‘When I got [it], I would also give them money.’ PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 314 Hugo Cardoso Both companions (8a) and instruments (8b) occur with a case-marker which, in the corpus, assumes two slightly different realizations as -s˜utad@or -j˜utad@: (8) a. Cannanore Indo-Portuguese (Cardoso: Elicited corpus) Olivia-s@ Olivia-gen fili child minh@ 1s.gen kr@skrãs@-s˜ utad@ children-com t@ be.prs br˜ ıka. play ‘Olivia’s son is playing with my children.’ b. Cannanore Indo-Portuguese (Cardoso: Elicited corpus) nOz 1p fak@-j˜ utad@ knife-inst kok coconut lo irr kebræ break ‘One breaks coconuts with a knife.’ In predicative possessive constructions, generic possessors are typically dative-marked in Malabar Indo-Portuguese. However, when it comes to temporary possession, the possessor is embedded in a particular locative phrase meaning ‘inside X’s hand’: (9) Cannanore Indo-Portuguese (Cardoso: Elicitated corpus) ælA 1sf pærmi 1s.obl prasãti present nuk@ neg da give ælA-s@ 3sf-gen mõ-d˜ etr@ hand-inside fAn@ money nutæ neg.be suydA. because ‘She didn’t give me a present because she didn’t have money.’ The morphological markers of locations, as expected, include an array of alternatives to convey semantic distinctions; for our purposes, in additional to generic location, those used to indicate proximate location and adjacency are particularly relevant. The Malabar Indo-Portuguese corpus contains the generic locative marker -d˜etr @ (10a,b) and the marker of proximate location pært @ (10c), with the same collocational properties described for examples (2) and (3) above: (10) a. Cochin Indo-Portuguese (Cardoso: Oral corpus) ak@ dem sister sister paymi 1s.obl hotel-d˜ etr@ hotel-loc ˜ uga one sirvis job ja pst oLa look da. give ‘That sister found me a job at a hotel.’ PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 The case of addressees in Dravido-Portuguese O caso dos interlocutores em Drávido-Português. . . 315 b. Cannanore Indo-Portuguese (Cardoso: Oral corpus) market-d˜ etr@ market-loc tud@ all t˜ e. be.prs ‘There’s everything in the market.’ c. Cannanore Indo-Portuguese (Cardoso: Elicited corpus) hotel-s@ hotel-gen pært@ near ˜ unga one igrædji church tæ. be.prs ‘There’s a church near the hotel.’ Goals are often unmarked when the predicates disambiguate the intended reading, as in (11a), or indicated by the oblique [dative-accusative] (11b) or locative marker (11c): (11) a. Cochin Indo-Portuguese (Cardoso: Oral corpus) olotr@ 3p ak@l dem Mysore Mysore lo irr mãda. send ‘They would send it to Mysore.’ b. Cannanore Indo-Portuguese (Cardoso: Elicited corpus) yo 1s minh@ 1s.gen irm˜ e-p@ sister-obl lo irr mãda. send ‘I will send [it] to my sister.’ c. Cannanore Indo-Portuguese (Cardoso: Elicited corpus)2 yo 1s minh@ 1s.gen irm˜ e sister pært@ near lo irr mãda. send ‘I will send [it] to my sister.’ 2.2 Sri Lanka Indo-Portuguese For this exploration of the case of Sri Lanka Indo-Portuguese addressees, as discussed in 1.1., I rely mostly on the descriptions published by Ian Smith (1977, 1979a, 1979b, 2013), which are based on his work with the Indo-Portuguese community of Batticaloa, as well as additional data Smith kindly provided for this study. While, admittedly, Batticaloa Indo-Portuguese may diverge somewhat from other past and present variants of Sri Lanka, at this point no significant and reliable sources are available for other variants. 2 The reader will notice that this particular example contradicts what has been said earlier about the requirement that pært @ occurs after a linking morpheme -s @ . It is indeed one of the few examples of this, which I interpret as a result of the conversational setting (elicitation) and/or perhaps the over-productivity of the blocking effect of minh@. PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 322 Hugo Cardoso (25) Portuguese (constructed sentence) Vou go.prs.1s dar give.inf um ind.m presente gift a dat um ind.m amigo. friend ‘[I] will give a present to a friend.’ Both companions (26a) and instruments (26b) take the preposition com ‘with’, although there are several other choices of prepositions to introduce companions: (26) a. Portuguese (constructed sentence) A def.f Luísa Luísa vai go.prs.3s à to.def.f escola school (junto) together com with / mais in.addition a def.f amiga. friend ‘Luísa goes to school with [her] friend.’ b. Portuguese (constructed sentence) Abre open.imp a def.f porta door com with esta this.f chave. key ‘Open the door with this key.’ In Portuguese, possessors which function as subjects of predicative possessive constructions with the verb ter ‘to have’ are unmarked. However, in order to convey a notion of temporary possession, an adjunct pp may be added in which the possessor is recovered (often anaphorically) and marked by the comitative/instrumental preposition com (27). It is worth noting that, syntactically, this construction is not comparable to the argument case-marking observed for Sri Lanka Indo-Portuguese: (27) Portuguese (constructed sentence) A def.f Luísa Luísa não neg tinha have.pst.ipfv.3s dinheiro money com com ela 3sf / consigo. 3.com ‘Luísa didn’t have money on her.’ As shown in (26), one option to introduce companions is to use the word junto ‘together, next to’ before the comitative preposition com. This construction is nowadays particularly common in Brazilian Portuguese, and builds on the inherent semantics of the word: as an adverb, junto also introduces the notion of shared agency (e.g. fomos juntos ‘[we] went together’) and, in combination of other prepositions (de or a), constructs proximate or adjacent location (28): PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 The case of addressees in Dravido-Portuguese O caso dos interlocutores em Drávido-Português. . . 323 (28) Portuguese (constructed sentence) A def.f chave key está be.prs.3s junto next a to esse that vaso. pot ‘The key is near/next to that pot.’ In addition, em ‘in, at’ marks generic location in Portuguese (29a), whereas the complex perto de ‘near’ is a dedicated marker of proximate location (29b), and the language furthermore has a host of other complex prepositions to establish finer semantic distinctions: (29) a. Portuguese (constructed sentence) A def.f chave key ficou stay.pst.3s em in casa. home ‘The key stayed at home.’ b. Portuguese (constructed sentence) A def.f chave key está be.prs.3s perto near de to um ind.m vaso. pot ‘The key is near a pot.’ The preposition awe have seen earlier as a marker of addressees, patients and recipients recurs with goals, in which function it competes with para ‘to’; the distinction between the two lies in the time-stability of the predicate, in that goals introduced by para are seen as more permanent than those introduced by a: (30) Portuguese (constructed sentence) Vamos go.prs.1p a to / para to casa. home ‘[We]’re going home.’ 2.4 Malayalam Malayalam, the major language of the modern Indian state of Kerala (which includes the coastal stretch formerly known as the Malabar), is especially important as the major sub-/adstrate of Malabar Indo-Portuguese. Yet, given that the earliest Portuguese settlements in Asia were precisely located in the Malayalam-speaking region, scholars have hypothesised that the contact between Malayalam and Portuguese was foundational in the formation of Portuguese-based creoles elsewhere in Asia (see Smith 1977: 143ff; Clements 2000, 2009), which would grant Malayalam a seminal role over a wider subset of Asian-Portuguese than just that of the Malabar. PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 324 Hugo Cardoso 2.4.1 Addressees in Malayalam Malayalam has a nominal suffix -oo ú@ expressing a typically Dravidian case often called sociative or associative in the literature (see Krishnamurti 2003: 235-237; Asher & Kumari 1997: 191ff), reserved for the stimuli of psych verbs (e.g. love, pity, be indebted) and the arguments of various predicates indicating social interaction (e.g. fight, borrow, excuse). This case is the one assigned to addressees of all types. Sentence (31a) exemplifies this for Addressee TALK arguments, (31b) for Addressee TELL arguments (and 35c,d) for Addressee ASK arguments: (31) a. Malayalam (Asher & Kumari 1997: 61) ñaan 1s ãookúar-ooú@ doctor-soc samsaariccu speak.pst ‘I spoke to the doctor.’ b. Malayalam (Asher & Kumari 1997: 109) raaman Raman raadhay-ooú@ Radha-soc oru a kaaryam thing par ¯aññu tell.pst ‘Raman told Radha something.’ c. Malayalam (Asher & Kumari 1997: 48) ñaan 1s ampalatt-il temple-loc pookukayaaïoo go.impfv.prs.int enn@ quot kuúúi child en ¯n ¯-ooú@ 1s-soc coodiccu ask.pst ‘The child asked me if I was going to the temple.’ d. Malayalam (Nair 2012: 35) keeral .am Kerala san ¯dar .śikkan .am visit.oblg enn˘ u quot aŋŋay-oot .˘ u you-soc avar they apeeks .iccirunnu request.pst.cont ‘They had requested you to visit Kerala.’ 2.4.2 Associate semantic roles in Malayalam Typical patients take the accusative marker -e, as shown in (32): (32) Malayalam (Asher & Kumari 1997: 60) pooliiskaaran policeman vaúi stick koïú@ inst kuúúiy-e child-acc aúiccu beat-pst ‘The policeman beat the child with a stick.’ PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 The case of addressees in Dravido-Portuguese O caso dos interlocutores em Drávido-Português. . . 325 Recipients, on the other hand, take a dative suffix which is sensitive to the phonological context of insertion and can therefore be realised as -kk @ , -n @ or -@: (33) a. Malayalam (Asher & Kumari 1997: 181) raaman Raman innale yesterday kôùïan ¯-n ¯@ Krishna-dat raïú@ two pustakamm book koúuttu give.pst ‘Krishna gave Krishnan two books yesterday.’ b. Malayalam (Asher & Kumari 1997: 193) kutti-kk@ child-dat amma mother paal milk koúuttu give.pst ‘The mother gave the child milk.’ In Malayalam, there are several ways of signalling companions (Asher & Kumari 1997: 196): either with the noun in the sociative case followed by the postpositions kuu ú i(34a), kuu ú e,oppam or ott @ , or with the noun in the genitive case followed by kuuúe(34b). (34) a. Malayalam (Asher & Kumari 1997: 62) saar ¯a Sarah kuuúúukaariy-ooú@ friend.f-soc kuuúi along.with vannu come.pst ‘Sarah came with a friend.’ b. Malayalam (Asher & Kumari 1997: 196) avaí 3 kuuúúukaariy-uúe friend.f-gen kuuúe along.with vannu come.pst ‘She came with a friend.’ Instruments can be indicated by a specific instrumental postposition koïú@ – see (32) above – or the instrumental case suffix -aal: (35) Malayalam (Asher & Kumari 1997: 210) tiiyin-aal fire-inst kuúil hut naSippikkapeúúu destroy.pass.pst ‘The hut was destroyed with(/by) fire.’ While Malayalam possessors are normally dative-marked (Asher & Kumari 1997: 174-175) in predicative possessive constructions, temporary possessors can be expressed in various ways: either by embedding the possessor in a locative phrase meaning ‘in X’s hand’ (36a) or by adding the locative postposition aúutt@‘near’ to a genitive-marked noun (36b): PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 326 Hugo Cardoso (36) a. Malayalam (Asher & Kumari 1997: 175) enre 1s.gen kayy-il hand-loc pa6am money u6ú@ be.prs ‘I have money (on me).’ b. Malayalam (Asher & Kumari 1997: 175) ayaaí-uúe 3s-gen aúutt@ near nalla good saarikaí sari.p u6ú@ be.prs ‘He has good saris (with him).’ Malayalam uses the case affix -il – see (36a) – to indicate generic location (37a), while proximate location uses the postposition a ú utt @ ‘near’ seen in the previous example sentence (37b) and, interestingly, the selection of the sociative case-marker -oo ú@ conveys an idea of adjacent location (37c). In addition, an array of other postpositions is used to make finer locative distinctions (see Asher & Kumari 1997: 64-65). (37) a. Malayalam (Asher & Kumari 1997: 63) ii prox kaaú-úil forest-loc vaíare many aanakaí elephant uïú@ be.prs ‘There are many elephants in this forest.’ b. Malayalam (Asher & Kumari 1997: 196) ki6ar ¯r ¯in-r ¯e well-gen aúutt@ next.to ninnu stand.pst ‘[He] stood near the well.’ c. Malayalam (Nair 2012: 23) amma mother kuññin ¯-e child-acc neñc-oot .u bosom-soc ceer .ttirunnu hold.close.pst.cont ‘The mother had held the baby close to her bosom.’ Finally, goals take the same case-marker as recipients: (38) Malayalam (Asher & Kumari 1997: 62) raajamma Rajamma kaúa-kk@ shop-dat pooyi go.pst ‘Rajamma went to the shop.’ PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 The case of addressees in Dravido-Portuguese O caso dos interlocutores em Drávido-Português. . . 327 2.5 Tamil At one point, Indo-Portuguese creoles were used in a wide area of Sri Lanka, comprising both the Tamil-majority and Sinhala-majority parts of the island. The data on which our discussion of Sri Lanka Indo-Portuguese, however, comes from the Tamil-speaking region (see 1.1), and therefore we take Tamil as the most relevant adstrate in this study – which is not to say Sinhala linguistic material and/or models could not have made their way into Batticaloa IndoPortuguese. The primacy of Tamil in this study is supported by previous ones: Smith (1977, 1979a) identified early on significant similarities between Batticaloa Indo-Portuguese and Tamil, and later studies (Smith 2012; Cardoso 2012) show this link to be stronger than that with Sinhala. 2.5.1 Addressees in Tamil In Tamil, Addressee TALK arguments can be treated in a number of ways. They may take the sociative affix -oo d . e 7 (39a), the animate locative affix -ki t .t . e (39b), or the dative affix -ukku (39c): (39) a. Tamil (Asher 1985: 30) kaïïan Kannan daakúar-ooúe doctor-soc peecinaan speak.pst.3sm ‘Kannan spoke to the doctor.’ b. Tamil (Asher 1985: 64) neettu yesterday vantu come.ptcp en-kiúúe 1s-loc peecana speak.pst.rel payyan boy ‘The boy who came yesterday and spoke to me’ c. Tamil (Asher 1985: 86) avan 3sm tan-akku self-dat taanee emph peecikiúúaan speak.refl.pst.3sm ‘He spoke to himself.’ Addressee TELL arguments can also take a variety of case-markers, but the array is a little different. Instead of the sociative, the literature reports the accusative marker -e (40a), but also the animate locative (40b) as a formal/deferential alternative, and the dative (40c) in informal speech: 7 As examples in this section come from different sources, and these use slightly different conventions for romanisation, some orthographical inconsistency is unavoidable. PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 328 Hugo Cardoso (40) a. Tamil (Asher 1985: 23) . . . kaïïan-e Kannan-acc vara come.inf connaaru say.pst.3s ‘. .. told Kannan to come’ b. Tamil (Schiffman 1999: 32) d .airekt .ar director kit .t .e loc en 1s kaariyatte concern sonneen lay.pst.defer ‘I laid my concern before the Director.’ c. Tamil (Steever 2005: 119) un ¯-akku you-dat laks .am 100,000 taram time colli tell irukkir ¯¯ en ¯, be.prs.1s ippat .i this.way vet .t .i cut poz .utu p¯ okk¯ at¯ e pass.neg.imp en ¯r ¯u say ‘I’ve told you 100,000 times not to fritter away your time wastefully like this.’ When it comes to Addressee ASK arguments, the postposition paattu (41a) and the animate locative marker (41b) are attested: (41) a. Tamil (Schiffman 1999: 43) avan-e he-acc paattu at keet .t .aal .a? ask.pst ‘Did she ask him?’ b. Tamil (Schiffman 1999: 189) ayyaa-kit .t .ee master-loc oru a koraccalaana cheap vaad .aheyle rent-loc oru a viid .u house paakka to.find sollungal .-een say.imp-defer ‘Would you be so kind as to ask the master to find me a cheap place to live?’ 2.5.2 Associate semantic roles in Tamil Typical patients take the accusative case affix-e; accusative-marking carries the notion of definiteness, so that indefinite patients may be unmarked if they are inanimate (Schiffman 1999: 36-37): (42) a. Tamil (Schiffman 1999: 90) tirud .anga thieves en my nay-e dog-acc kon .n .upt .t .aanga kill.malice.pst.3p ‘Thieves (in cold blood) killed my dog.’ PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 The case of addressees in Dravido-Portuguese O caso dos interlocutores em Drávido-Português. . . 329 b. Tamil (Schiffman 1999: 94) kaaru car tondaravu trouble kod .uttu-kit .t .irundadu; give-dur ad-e it-acc vittu sell toleccut .t .een impat.pst.1s ‘My car was giving me trouble; I sold it off (and was finished with it).’ Recipients typically take the suffix -ukku/akku/(y)kki, as demonstrated in (43a); in case they are assigned the animate locative marker ki t .t . e, the implication is that the predicate restores previous ownership (43b): (43) a. Tamil (Schiffman 1999: 32) avar-ukku 3s-dat muttu pearl maaleye necklace kud .utteen give.pst.1s ‘I gave him the pearl necklace.’ b. Tamil (Schiffman 1999: 32) avar-kit .t .e 3s-loc muttu pearl maaleye necklace kud .utteen give.pst.1s ‘I returned the pearl necklace to him.’ The sociative case-marker -oo d . eintroduced in 2.5.1 is typically assigned to companions (44a). There are, however, alternative strategies involving the postpositions kuud .e(44b) and ton .eyle (44c): (44) a. Tamil (Schiffman 1999: 35) sneydar-ood .e friend-soc ‘with a friend’ b. Tamil (Schiffman 1999: 39) attaan brother-in-law.nom kuud .e with ‘with brother-in-law’ c. Tamil (Schiffman 1999: 40) enga we.nom ton .eyle with ‘with us’ PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 330 Hugo Cardoso Instruments have a dedicated case affix -aal(e), demonstrated in (45); however, when applied to inanimates, sociative -oo d . emay also have an instrumental reading (Schiffman 1999: 34ff). (45) Tamil (Krishnamurti 2003: 236) villin-¯ al bow-inst ‘with the bow’ Just as in Malayalam, typical possessors are dative-marked (Schiffman 1999: 29-30) but, in order to convey the notion of temporary possession, the animate locative -kit .t .eis selected: (46) a. Tamil (Schiffman 1999: 32) en-kit .t .e 1s-loc pan .am money irukku be.prs ‘I have money on me.’ The marking of generic location is sensitive to animacy, in that inanimate locations receive the marker –le (47a) – reported in Smith (1977) as -(i)la(y) for Sri Lanka Tamil (47b) – and animate locations take the marker -(ki) t .t . e we have encountered earlier in a variety of functions (47c) – see also (40b): (47) a. Tamil (Schiffman 1999: 31) kooyil-le temple-loc ‘in the temple’ b. Tamil (Smith 1977: 146) enka-Ta we-gen vi:TT-ila house-loc kataykk-a:ma speak-without iŕukk-ra:nka be-prs ta:y-takappan!? mother-father ‘in our house are [her] parents not speaking it?’ c. Tamil (Schiffman 1999: 31) avar-t .t .e he-loc ‘on him’ Finer semantic locative distinctions make use of several different postpositions. Notions of proximate location, for instance, may be indicated by pakkattle (48a), pakkam (48b) or -(a)n .d .e(48c): PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 The case of addressees in Dravido-Portuguese O caso dos interlocutores em Drávido-Português. . . 331 (48) a. Tamil (Schiffman 1999: 37) kooyil-ukku temple.dat pakkattle near ‘near the temple’ b. Tamil (Schiffman 1999: 39) med .raas Madras.nom pakkam near ‘near Madras’ c. Tamil (Schiffman 1999: 40) avar-an .d .e he.nom-near ‘near him’ Inanimate goals are dative-marked in Tamil (49a), while an animate goal takes the animate locative marker (49b): (49) a. Tamil (Steever 2005: 119) avan ¯ 3sm v¯ ıt .t .u-ku house-dat ¯ ot .i run p¯ on ¯¯ an ¯ go.pst.3sm ‘He went running to the house.’ b. Tamil (Smith 1979a: 198) appa:-TTa father-loc po: go ‘Go to father.’ 3 Analysis To aid in the discussion, the results of the survey above are given in Table 1, which indicates what markers can be associated with each semantic role in all of the five languages under analysis. In this table, the first column refers to Portuguese and, to make the comparison easier, each of the DravidoPortuguese creoles are shown next to their respective Dravidian adstrate. In each column, those markers that occur with addressees are given in bold type, so that it becomes clearer what functions they perform in other domains of the language. PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 338 Hugo Cardoso coincidental, their association is perhaps based on the notion that, logically, adjacency implies proximate location (but not the other way around). Once this connection was established, the process of functional reanalysis (in this case, functional expansion) of ptg perto (de) on the basis of mal -oo ú@ could proceed, resulting in the modern functional range of mip -(s@) pært@. As for mip -j˜utad @ , we have established that its form is derived from ptg juntado and much of its functional distribution appears to be modelled on ptg com. While ptg juntado (a participle form of the verb juntar ‘to join, connect’) does not occur as such in Table 1, the related adverb junto ‘together, next to’ (see 2.3.2) does, involved in the expression of proximate location and comitative. The shared comitative semantics of Portuguese junto (com)/junto(s) and com is perhaps the key to why, in Malabar Indo-Portuguese, a derivative of the former could take over a portion of the functional range of the latter to encompass instrument-marking and AddresseeTALK-marking. The same Portuguese cluster (junto and juntado) is also involved in the development of three Sri Lanka Indo-Portuguese markers, viz. juuntu,-(u)ntu and -(u)ntaa. However, as noted in 3.1, their function range reveals the influence of two different Tamil case-markers, viz. -oo d . eand -(ki) t .t . e. With respect to slip juuntu, the process of functional reanalysis was probably motivated by the coincidence of ptg junto and tam -oo d . eas markers of companions, while in the case of-(u)ntaa the relevant functional match is that between ptg junto(/juntado) and tam -(ki) t .t . ein the locative domain. In the case of -(u)ntu, the explanation is similar to that for -(u)ntaa, but there is the additional question of whether there was ever a stage in the creole in which a single form (perhaps juuntu or similar) covered the combined functional range of modern-day juuntu +-(u)ntu, before a process of speciation and phonological reduction resulted in the modern-day distribution. An answer to this question is likely to lie in a critical analysis of older texts written in Sri Lanka Indo-Portuguese, which lies beyond the scope of this study. Finally, the case of slip -p @ is a little more complex. If we were to limit ourselves to Table 1, we might propose that the fact that both ptg para (the etymon, also typically a marker of beneficiaries) and ptg a(one of the relevant functional models) can be used to mark goals provides the grounds for the association between matter and pattern which resulted in modern slip -p @ . This may be so, but the fact that para or derivatives of it have additional functions (e.g. as addresseeor recipient-markers) not only in various other Portuguese-based creoles but also in some varieties of Portuguese (such as Brazilian Portuguese) raises the possibility that, in spoken Classical Portuguese, the functional similarity between para and awas closer than the available sources and descriptions indicate. This issue, once again, goes beyond the scope of the current comparative work, and stands as a final suggestion of further research which highlights the potential of the study of creole languages to challenge and motivate other domains of linguistic inquiry. PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 The case of addressees in Dravido-Portuguese O caso dos interlocutores em Drávido-Português. . . 339 Abbreviations acc accusative, com comitative, cond conditional, conj conjunction, cont continuous, dat dative, def definite, defer deferential, dem demonstrative, dur durative, emph emphasis, fut future, gen genitive, hab habitual, hon honorific, imp imperative, impfv imperfective, ind indefinite, inf infinitive, inst instrumental, int interrogative, irr irrealis, loc locative, mal Malayalam, mip Malabar Indo-Portuguese, nnoun, neg negative, nom nominative, np noun phrase, obl oblique, oblig obligation, pass passive, pfv perfective, pot potential, prox proximate, prs present, pst past, ptcp participle, ptg Portuguese, quot quotative, refl reflexive, rel relative, slip Sri Lanka Indo-Portuguese, soc sociative, tag tag, tam Tamil. References Asher, R.E. 1985. Tamil. London: Croom Helm. Asher, R.E. & T.C. Kumari. 1997. Malayalam [Descriptive Grammars Series]. London/New York: Routledge. Bakker, Peter. 2006. The Sri Lanka Sprachbund: The newcomers Portuguese and Malay. In Matras, Yaron, April McMahon & Nigel Vincent (eds.) Linguistic Areas: Convergence in Historical and Typological Perspective, 135–159. New York: Palgrave. Baxter, Alan N. 1988. A Grammar of Kristang (Malacca Creole Portuguese). Canberra: Pacific Linguistics. Cardoso, Hugo C. 2006. Challenges to Indo-Portuguese across India. In Elangayan, R., R. McKenna Brown, N. Ostler & M.K. Verma (eds.) Proceedings of the FEL X, 23-30. Mysore: Central Institute of Indian Languages. Cardoso, Hugo C. 2009. The Indo-Portuguese Language of Diu. Doctoral dissertation, University of Amsterdam. Cardoso, Hugo C. 2012. Luso-Asian comparatives in comparison. In Cardoso, Hugo C., Alan N. Baxter & Mário Pinharanda Nunes (eds.) Ibero-Asian Creoles: Comparative perspectives, 81-123. Amesterdam: John Benjamins. Cardoso, Hugo C., Alan N. Baxter & Mário Pinharanda Nunes. 2012. Introduction. In Cardoso, Hugo C., Alan N. Baxter & Mário Pinharanda Nunes (eds.) Ibero-Asian Creoles: Comparative perspectives, 1-14. Amesterdam: John Benjamins. PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 340 Hugo Cardoso Clements, J. Clancy. 1996. The Genesis of a language: The formation and development of Korlai Portuguese. Amsterdam: John Benjamins. Clements, J. Clancy. 2000. Evidência para a existência de um pidgin português asiático. In d ' Andrade, E., Dulce Pereira & M.A. Mota, (eds.) Crioulos de base lexical portuguesa. Braga: Associação Portuguesa de Linguística. Clements, J. Clancy. 2009. The linguistic legacy of Spanish and Portuguese; Colonial expansion and language change. Cambridge: Cambridge University Press. Dalgado, Sebastião Rodolfo. 1900. Dialecto Indo-Portugues de Ceylão. Lisbon: Imprensa Nacional. Dalgado, Sebastião Rodolfo. 1917. Dialecto Indo-Português de Negapatão. Revista Lusitana 20: 40-53. Grant, Anthony P. 2012. Mindanao Chabacano and other ‘mixed creoles’; Sourcing the morphemic components. In Cardoso, Hugo C., Alan N. Baxter & Mário Pinharanda Nunes (eds.) Ibero-Asian Creoles: Comparative perspectives, 327-364. Amesterdam: John Benjamins. Jackson, K. David. 1990. Sing Without Shame. Oral Traditions in IndoPortuguese Creole Verse. Amsterdam: John Benjamins. Krishnamurti, Bhadriraju. 2003. The Dravidian languages. Cambridge: Cambridge University Press. Ladhams, John. 2009. The formation of the Portuguese-based Creoles: Gradual or abrupt? In Selbach, Rachel, Hugo C. Cardoso & Margot van den Berg (eds.) Gradual creolization: Studies Celebrating Jacques Arends, 279-303. Amsterdam: John Benjamins. Matras, Yaron. 2009. Language contact. Cambridge: Cambridge University Press. Nair, R.S.S. 2012. A Grammar of Malayalam. Language in India <http://www.languageinindia.com/nov2012/ravisankarmalayalamgrammar.pdf> Schäfer-Prieß, Barbara. 2002. O acusativo preposicional na história da língua portuguesa. In Head, Brian F., J. Teixeira, A.S. Lemos, A.L. de Barros & A. Pereira (eds.) História da língua e história da gramática, 405-419. Braga: Universidade do Minho/ilch. Schiffman, Harold F. 1999. A Reference Grammar of Spoken Tamil. Cambridge: Cambridge University Press. Schuchardt, Hugo. 1882. Kreolische Studien II. Über das Indoportugiesische von Cochim. Sitzungsberichte der Kaiserlichen Akademie der Wissenschaften zu Wien (Philosophisch-historische Klasse) 102: 799-816. PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 The case of addressees in Dravido-Portuguese O caso dos interlocutores em Drávido-Português. . . 341 Schuchardt, Hugo. 1883. Kreolische Studien VI. Über das Indoportugiesische von Mangalore. Sitzungsberichte der Kaiserlichen Akademie der Wissenschaften zu Wien (philosophisch-historische Klasse) 105(III): 882-904. Schuchardt, Hugo. 1889a. Beiträge zur Kenntnis des creolischen Romanisich. V. Allgemeineres über das Indoportugiesische (Asioportugiesische). Zeitschrift für Romanische Philologie 13: 476-516. Schuchardt, Hugo. 1889b. Beiträge zur Kenntnis des kreolischen Romanisch: VI. Zum Indoportugiesischen von Mahé und Cannanore. Zeitschrift für Romanische Philologie 13: 516-524. Silva Neto, Serafim da. 1970. História da Língua Portuguesa (2nd edition). Rio de Janeiro: Livros de Portugal. Smith, Ian R. 1977. Sri Lanka Creole Portuguese Phonology. Doctoral dissertation, Cornell University. Smith, Ian R. 1979a. Convergence in South Asia: a creole example. Lingua 48: 193-222. Smith, Ian R. 1979b. Substrata vs.universals in the formation of Sri Lanka Portuguese. Papers in Pidgin and Creole Linguistics 2: 183-200. Smith, Ian R. 2012. Measuring substrate influence: Word order features in Ibero-Asian creoles. In Cardoso, Hugo C., Alan N. Baxter & Mário Pinharanda Nunes (eds.) Ibero-Asian Creoles: Comparative perspectives, 125-148. Amesterdam: John Benjamins. Smith, Ian R. 2013. Sri Lanka Portuguese structure dataset. In Michaelis, Susanne M., Philippe Maurer, Martin Haspelmath & Magnus Huber (eds.) Atlas of Pidgin and Creole Language Structures Online. Leipzig: Max Planck Institute for Evolutionary Anthropology. (Available online at http://apicsonline.info/contributions/41, accessed 18.12.2013) Smith, Norval. 1995. An annotated list of creoles, pidgins, and mixed languages. In Arends, Jacques, Pieter Muysken & Norval Smith (eds.) Pidgins and Creoles - An introduction, 331-374. Amsterdam/Philadelphia: John Benjamins. Steever, S.B. 2005. The Tamil Auxiliary Verb System. Oxon: Routledge. Tavares de Mello, B.C. 1908. Dialecto Indo-Portuguez de Ceylão; Te Deum Laudamus. O Oriente Portuguez V: 373-390. [Reprinted in Tavares de Mello, B.C. 1998. Folclores Ceilonenses: Colectânea de textos do crioulo português do Ceilão, 115-130. Lisbon: Comissão Nacional para as Comemorações dos Descobrimentos Portugueses.] Teyssier, Paul. 2005. A língua de Gil Vicente. Lisbon: Imprensa Nacional – Casa da Moeda. PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767 342 Hugo Cardoso Teza. Emilio. 1872. Indo-portoghese. Il Propugnatore 5: 1-10. Tomás, Maria Isabel. 1992. Os Crioulos Portugueses do Oriente; Uma Bibliografia. Macau: Instituto Cultural de Macau. Verschueren, Jef. 1981. Basic linguistic action verbs. Cahiers de Linguistique Française 2: 71-88. Received: 06/30/2013 Accepted: 12/01/2013 PAPIA, 24(2), ISSN 0103-9415, e-ISSN 2316-2767