The synchrony and diachrony of an Asian-Portuguese causal morpheme1 Hugo C. Cardoso Universidade de Lisboa This study explores the use of a particular causal morpheme, derived from a protoform that may be reconstructed as *[V/B]IDA, in the Portuguese-lexified creoles of Asia. A survey of the various formal means employed by the AsianPortuguese creoles to establish relationships of cause, reason, and purpose demonstrates that only three may be said with certainty to use or have used a *[V/B]IDA-related morpheme, viz. those of the Malabar (South India), Sri Lanka, and Batavia/Tugu (Java, Indonesia) – a geographical distribution which, it is argued, calls for an assessment of the exact role of South Asian populations in the formation of Batavia/Tugu creole, in addition to other pieces of linguistic and ethnographic evidence. In order to determine the etymology and synchronic transformations of these causal morphemes, this study also explores several diachronic and dialectal corpora of Portuguese, which reveals that the Portuguese expression por via de ‘by way of’ is a more likely source than another proposed alternative, por vida de ‘by the life of’. Keywords: Asia, Asian-Portuguese creoles, cause, reason, purpose, etymology. 1. Introduction The Asian-Portuguese creoles were once spoken across coastal Asia, from India to Timor, although they are now much reduced in number. A consequence of such a wide geographical dispersion is the fact that there is significant diversity in terms of the substrate and adstrate languages which have contributed to their formation and development, and the sociohistorical conditions which have 1 I am indebted to several colleagues and consultants who have taken the trouble to provide me with relevant data and/or access to particular corpora: Alan Baxter, Catarina Magro, Chiara Truppi, Clancy Clements, Clifford Netto, Mahesh Radhakrishnan, Patrícia Costa, Tjerk Hagemeijer, and Vanessa Köbke. I am also thankful to two reviewers and to the participants in the 2017 Annual Meeting of the ACBLPE (Stockholm University) for insightful and helpful suggestions. Any shortcomings remain my own responsibility. The research necessary for this study was funded by Fundação para a Ciência e a Tecnologia (FCT) under the Investigador FCT programme (IF/01009/2012). Journal of Ibero-Romance Creoles 9.1 (2019), 27-54 ISSN 2184-5360 © Journal of Ibero-Romance Creoles Hugo C. Cardoso presided over these processes. As a result, the Asian-Portuguese creoles are characterised by considerable variation, but also, interestingly, significant continuities (Ferraz 1987). Since, after a period of some scholarly neglect, there are now descriptions and edited sources for all of the creoles currently spoken (and a few of the extinct varieties), comparative studies aimed at determining and explaining the degree of divergence or similarity between them have become common (see e.g. Clements 2000; also the contributions in Cardoso & Ansaldo 2009, and in Cardoso, Baxter & Nunes 2012). This article is very much part of that tradition and operates on the premise that establishing synchronic (or recent) continuities between discrete Asian-Portuguese creoles can shed light on their historical relationships and, as a result, on the history of the populations that spoke – and, in some cases, continue to speak – them. In this case, I will explore the distribution and use of a range of causal morphemes which appear to stem from a protoform tentatively reconstructed as *[V/B]IDA. As we will see, among the Asian-Portuguese creoles, such causal morphemes are geographically widespread but by no means universal, as only a few of them use it (either at present or historically), and this lends support to the necessity of fully establishing the historical and linguistic links between creole-speaking communities of two distinct Asian regions: Southern South Asia (i.e. South India and Sri Lanka) and Java (in Indonesia). The demonstration of the geographical and diachronic distribution of these causal morphemes among the Asian-Portuguese creoles will be the object of section 2 below. More than just establish the current and/or past use of a *[V/B]IDArelated morpheme, however, I am also interested in advancing an etymological scenario to account for these forms. To that end, I will explore diachronic and dialectal corpora of European Portuguese in search of two distinct proposed etymological sources – the expressions por via de ‘by way of’ and por vida de ‘by [the] life of’ – and survey other Portuguese-based creoles for cognate forms that may shed light on the issue. The results of this research are reported and discussed in section 3. Finally, in section 4, I will consider the diachronic implications of the data presented here for our understanding of the development and historical relationships between the Asian-Portuguese creoles, both living and extinct. 28 The synchrony and diachrony of an Asian-Portuguese causal morpheme 2. Distribution in Asia This section will survey the various Portuguese-lexified creoles of Asia to ascertain the distribution of *[V/B]IDA-related causal morphemes, in modern as well as historical sources, adding, for clarity, the alternative forms used with comparable values in each one. 2.1. South Asia Our knowledge of the Malabar Indo-Portuguese [henceforth MIP] creoles is based on two rather different sets of data: a) a short 19th-century corpus composed of language samples collected by several different collaborators for Hugo Schuchardt (in Cannanore, Mahé, and Cochin) and published in his articles on the subject (Schuchardt 1882; 1889); and b) a modern oral corpus collected by myself (in Cannanore, Vypeen, and Calicut) since 2006 (Cardoso 2006-2015; 2007-2010; 2015; 2018). The modern MIP corpora provide plenty of instances of a causal morpheme suydɐ, occurring at the right edge of a subordinate clause, very frequently in adverbial clauses of cause (1a), reason (1b)2 and, more rarely, in purpose clauses (1c): (1) a. [Cannanore Creole, elicited data] [chuva jə-vi-suydɐ], rain tudu-kə molæsti jə-vi PST-come-CAUS all-OBL sickness PST-come ‘Because it rained, everyone got sick.’ b. [Cannanore Creole, elicited data; Cardoso 2006-2015] bombə chuyvə tini-suydɐ, iskɔla ja-ficha much rain EXS.PST-REAS school PST-close ‘Because there was so much rain, the school was closed.’ 2 In this article, following Croft (1991) and Luraghi (2005), the distinction between Cause and Reason is based on intentionality: whereas the states of affairs given in Reason clauses motivate an action which implies the intention of an agent (as is also the case with Purpose clauses), those expressed in Cause clauses have effects that require no intentionality on the part of an agent. 29 Hugo C. Cardoso c. [Cannanore Creole, elicited data; Cardoso 2006-2015] Michael [pə-ræza-suyda] igreja tə-vay Michael INF-pray-PURP church IPFV-go ‘Michael goes to church to pray.’ The use of -suydɐ with NPs is unattested in the unconstrained speech sections of the corpus, and even elicitation does not clarify this possibility, as the request to translate phrases of the ‘because of NP’ type usually produces full clauses. However, there is a very frequent combination with the demonstrative akə which, as a whole, establishes a result nexus between two clauses, as in (2): (2) [Cannanore Creole, elicited data; Cardoso 2006-2015] æli-pərɐ unga bicycle kerɐ, [akə-suydɐ yo lo-pidi da] 3s.m-OBL one bicycle want DEM-CAUS 1s IRR-buy give ‘He wants/needs a bicycle, therefore [i.e., because of that] I’ll buy him one.’ Although there is currently no evidence that this form is multimorphemic, it is very likely that, at least in diachronic terms, it was so. In fact, -suydɐ can be interpreted as being composed of the genitive case-marker –sə plus a causal morpheme *–(w)ydɐ, on the basis that, in MIP, several (but not all) postpositions select a genitive NP. As a result, –sə functions as a linking morpheme between the NP and the postpositions, optionally in the case of dæ̃trə ‘in’, but obligatorily with many others (e.g. bayxi ‘under’, dyãti ‘in front’, lɔ̃ji ‘far’, rivɐ ‘above’, etc.). In addition to -suydɐ, more focused elicitation conducted for this particular study has revealed another equivalent postposition in modern MIP. In (3), kawsu (from Ptg. causa ‘cause’) is seen to select a genitive-marked nominal, in this case, in a reason context:3 3 Determining the distributional properties of kawsu beyond the specific syntactic context provided in (3) requires further investigation. 30 The synchrony and diachrony of an Asian-Portuguese causal morpheme (3) [Cannanore Creole, elicited data] olotə Kanur 3p jə-vi Kannur PST-come [minha kawsu]. 1s.GEN REAS ‘They have come to Kannur because of me.’ In contrast with the abundance of tokens of causal -suydɐ in the modern MIP corpora, the 19th-century data – which are relatively scarce and, in part, of questionable descriptive reliability (see Cardoso 2014: 94-100) – provide no instances of a *[V/B]IDA-type causal morpheme, resorting instead to other forms. The very few cause/reason clauses which do occur in the 19th-century corpus use a subordinator derived from standard Portuguese por causa ‘because’ (related to the modern postposition kawsu mentioned above) + complementiser que ‘that’, or por razão ‘by reason’ + complementiser que.4 In comparison with MIP, we have an abundance of sources for Sri Lanka Portuguese [henceforth SLP]. Not only was there a substantial written production in the 19th and early 20th century (see Cardoso, Hagemeijer & Alexandre 2015; Smith 2016), the language has also been the object of documentation and grammatical description in recent times, especially in the work of Ian Smith (e.g. Smith 1977, 1979, 1984, 2013a) and, even more recently, as part of the project Documentation of Sri Lanka Portuguese [henceforth DSLP] developed at the University of Lisbon (data available in Cardoso 2017). The use of a cognate causal morpheme in modern SLP is, in some ways, similar to what has been described for MIP. In this creole, the corresponding form is wi:də/viida,5 (i.e., excluding the genitive linker which can be recognised in the MIP form) and also occurs at the right edge of the clause or phrase. It occurs mostly in adverbial clauses of cause (4a), reason (4b), and purpose (4c): 4 Some examples from the Cochin corpus: não por causa que sou pela natureza melhor que outros, mas somente por rezão que eu tamou lugar de um culpado peccador diante de Deos... ‘not because I am by nature better than others but only because I have taken the place of a guilty sinner before God’ (Schuchardt 1882: 806-807). 5 The orthographical variation observed in the modern Sri Lankan data reflects the source of the materials transcribed, since Ian Smith and the DSLP team use slightly different orthographies. In the case of this pair of optional spellings, wi:də is Smith’s variant, and viida is the one used in the DSLP transcriptions. 31 Hugo C. Cardoso (4) a. [Batticaloa Creole; unconstrained; Cardoso 2017: slp046_1] [áklóóra málváár papiyáá viida], etus málváár then Tamil 3p.HON Tamil mee papiyáá papiyáá ya-ka-kustumáá. EMPH speak PST-PFV-practice speak CAUS speak ‘Because they spoke Tamil then [i.e., growing up], they [i.e., youngsters] have gotten used to speaking in Tamil.’ b. [Batticaloa Creole; Smith 1977: 163] [ja:-fəla: wi:də me:] ja:-wi: PST-tell REAS EMPH PST-come istɔ:ndə this.way ‘It is precisely because he told me that [I] came over here.’ c. [Batticaloa Creole; unconstrained; Cardoso 2017: slp038_1] [elispa reskatáá viida mee] 3p-OBL rescue nóóspa luumi triiya… PURP EMPH 1p-OBL light bring ‘In order to rescue them, we [would have to] bring a light…’ In addition to clauses, wi:də / viida can also be used with NPs (5a), and, as in MIP, there is a very frequent combination with the demonstrative áka introducing a result (5b): (5) a. [Trincomalee Creole; unconstrained; Cardoso 2017: slp014_1 [áka istóóri viida] saylááv, isi DEM history REAS Ceilão, DEM place-OBL fala-ntu mee garáá speak-PFV FOC keep luváára-pa saylááv Ceilão voo. Q ‘Because of that history, [we should] keep the name Ceilão for this land, right?’ 32 The synchrony and diachrony of an Asian-Portuguese causal morpheme b. [Batticaloa Creole; Smith 1977: 150] [əkə wi:də etus DEM CAUS 3p.HON shyness-LOC wærñə-ntu na:-pəpiya:] NEG.FUT-talk ‘Therefore, in shyness, they won't talk.’ Modern SLP admits alternatives to wi:də / viida. The DSLP corpus contains many instances of a causal/reason marker dáála (also, occasionally, daláá), which has not been recorded in earlier descriptions of the language. Interestingly, dáála occurs overwhelmingly in speech collected in Trincomalee, with a single speaker in Batticaloa producing it. Therefore, it appears to be a regional variant,6 even though Trincomalee speakers also use wi:də / viida abundantly. The distribution of the two causal/reason markers seems to be equivalent, as dáála occurs in adverbial clauses of cause (6a) and reason (6b) – but is unattested in clauses of purpose –, with pronouns (6c), and with demonstratives in result contexts (6d): (6) a. [Trincomalee Creole; unconstrained; Cardoso 2017: slp063_1] […] [kilááy áka how nósa DEM tudu ya-ka-fikáá dáála], all CAUS PST-PFV-become “music” pooku tráás ya-ka-fikáá 1p.GEN music little behind PST-PFV-become avóóra. now ‘[…] because of how all that has become, our music has now fallen behind a little.’ b. [Trincomalee Creole; unconstrained; Cardoso 2017: slp056_3] mesta-juvisáá, [óóras nunteem OBLIG-evaluate time NEG.have dáála], seem. REAS yes ‘I need to check, because I have no time, yes.’ This fact may explain why dáála went unrecorded in Ian Smith’s corpus, since it focused heavily on data collected in Batticaloa. The etymology of dáála is unclear. 6 33 Hugo C. Cardoso c. [Trincomalee Creole; unconstrained; Cardoso 2017: slp063_5] [osiir daláá], osiir tinha 3s.HON CAUS 3s.HON viida mee, áka EXS.PST CAUS EMPH DEM isi “recording” loov DEM recording immediately can pooy faya=ley tinha. do=like be.PST ‘Because of him, because he was [here], [we] could do that recording straight away.’ d. [Trincomalee Creole; unconstrained; Cardoso 2017: slp056_3] […] senáá [áka so dáála kááza-ntu etrus DEM CAUS house-LOC 3p prenda ta-faya] […] learn PRS-do málváár mee Tamil EMPH ‘So, because of that, at home they learn Tamil.’ In addition to constructions with postposed wi:də / viida and dáála, clauses of cause and reason may also be introduced by a complex connector composed of kiipa ‘why/what for’7 and a conditional subordinator (kiipa see or kiipa kamfaláá; see also Smith 2016: 240). Purpose clauses, on the other hand, are most often flagged simply with a verbal affix pa ‘INF/PURP’ (Smith 2013a: 116, 118). When it comes to early sources, once again, SLP is much more richly recorded than MIP. As a result, the written corpora of the 19th and early 20th century contain many instances of causal morphemes cognate with modern wi:də / viida. It occurs in those texts with the forms vide/vidè or videque/vide que (the latter integrating what appears to be the Ptg. complementiser que ‘that’) but, instead of the right edge of the clause/phrase, it precedes it (which is consistent with the well-established opposition between these early SLP sources and the modern oral creole in terms of word order, as explored and discussed e.g. by Smith 1984, 2016; Iken 2000; Bakker 2006). The nexus it establishes is 7 Smith (e.g. 2013b: 41-28) interprets this form as being transparently composed of Q-word kii ‘what’ + oblique suffix –pa; in the DSLP corpus, the form is treated as a single Q-word meaning ‘why’. 34 The synchrony and diachrony of an Asian-Portuguese causal morpheme always that of cause or reason. With respect to syntactic distribution, it is a strong generalisation that vide/vidè precedes NPs, pronouns or demonstratives (7a), whereas videque/vide que introduces adverbial clauses (7b) (see also Smith 2016: 240); occasionally, however, the corpus contradicts this (7c,d): (7) a. [Tavares de Mello 1998[1908]: 106; ‘Oração’] nós pôde contá nós mesmo digno per suffrí [vidè tua CAUS your santa nome]. holy name ‘We can count ourselves worthy of suffering because of your holy name.’ b. [Newstead 1852: Matthew 3: 2] E fallando, Arrepende vossotros, [videque REAS ceos tem chegado perto]. heaven.p be arrive.PTCP close o reyno de ART kingdom of ‘And saying: “Repent, for the kingdom of heaven is drawing near”.’ c. [Tavares de Mello 1998[1910-1913]: 152] O prendido femes de este tempo não confiá ne amor, [vide ellotros CAUS 3p tem assi prendido que ellotros te be thus learn.PTCP COMP 3p grande dodice per obedecê per big folly INF obey lembrá tem hum PRS think sua OBL GEN be one maridos]. husband.p ‘Learned women of this time do not trust love, because they have learnt that they find it a great folly to obey their husbands.’ 35 Hugo C. Cardoso d. [Tavares de Mello 1998[1908]: 108; ‘Oitavel Estação’] […] Jesus sabendo o cousas que tinha per susdê sobre Jerusalem, [vidèque de rejectação de elle] […] CAUS of rejection of 3sm ‘Jesus, knowing the things that would happen to Jerusalem because of [its] rejection of him’ In addition, a form videaquel (rarely also aquelvide or vide este), consisting of the causal morpheme and a demonstrative, is very frequent in result contexts: (8) [Newstead 1852; Acts 16: 36] […] O capitaõs ja manda orde per larga per vossotros, [videaquel parte agora, e CAUS.DEM leave now anda ne and go paz]. LOC peace ‘The captains have sent orders to let you go, therefore, leave now, and go in peace.’ All these forms were recorded in early descriptive works of SLP (then known as Ceylon Portuguese), including Berrenger’s 1811 grammar of the language and William Buckley Fox’s 1819 trilingual Ceylon Portuguese-Sinhala-English dictionary. In one of his dictionaries, John Callaway (1820: 31) even makes an explicit distinction between videaquel (translated as ‘therefore’), videque (translated as ‘because’), and vide X (translated as ‘on account of X’). The prevalence of *[V/B]IDA-type causal morphemes in MIP and SLP has no parallel in the other extant Indo-Portuguese creoles, which have derived etymologically distinct forms for the same functions. In the cases of Diu Creole and Daman Creole, which are linguistically (as well as geographically and historically) close, the terms used for establishing a causal nexus are quite similar to the most common terms in modern Portuguese. In the creole of Diu, recent documentation has recorded pərki/purki (from Ptg. porque ‘because’) with adverbial clauses of reason or cause (Cardoso 2009: 135ff), and pə/pu (from Ptg. por ‘by’ or para ‘for’) with purpose clauses (Cardoso 2009: 155); kawz or pukawz (from Ptg. (por) causa ‘because [Lit. ‘by cause]’) is a causal morpheme which may precede a nominal, if mediated by a genitive preposition 36 The synchrony and diachrony of an Asian-Portuguese causal morpheme də, or a clause of cause/reason, if mediated by the complementiser ki (Cardoso 2009: 185ff); as for result clauses, they typically use the connector pures (from Ptg. por isso ‘therefore [Lit. ‘by that’]’) (Cardoso 2009: 135ff, 185ff, 240). Similarly, for Daman, Clements records the causal subordinator purkə/purki with clauses of cause and reason (2012: 37; 2014: 135), pə with purpose clauses (2014: 137), and puris to indicate a result nexus (2014: 154); Clements and Koontz-Garboden (2002: 224) further note a complex causal preposition pur kawz də, similar to the one observed in Diu. Korlai creole is, in this domain, rather different from either Daman and Diu, to the north, or MIP and SLP, to the south. Clements notes that, in Korlai, NPs interpreted as expressing reason/cause or purpose are treated as a genitive phrase, with the use of the postposition -su (from Ptg. sua ‘his/her [f.]’) (Clements 1996: 140-141), and there is also a cause/reason preposition rhəpəd/rhəspəd (from Ptg. (a) respeito de ‘with respect to’) – also used, in combination with a demonstrative (resulting in rhəpədəkə; Clements p.c.) to introduce result clauses (Clements 1996: 162). Elsewhere, in a multi-language wordlist focusing on the Indo-Portuguese creoles, Clements also includes the form karən as a translation of ‘because’ in Korlai (2012: 37). This form corresponds to one of the conjunctive particles which introduce adverbial clauses of cause in Marathi (see Dhongde & Wali 2009: 228-229) and occurs at te left edge of a cause/reason clause (Clements p.c.). Finally, as for purpose clauses, examples provided by Clements (2013: exs. 40-16, 40-124) indicate that they are treated very much like complement clauses, with the use of the complementiser (also relativiser) ki. 2.2. (South)east Asia Among the Portuguese-lexified creoles of Southeast and East Asia, the only one which can be said to have made use of a *[V/B]IDA-type causal morpheme is the extinct creole of Batavia and Tugu, known to us through the corpus published by Schuchardt in 1890 and a few other sources, which have been collated and studied by Maurer (2011). According to Maurer’s study, the formal means employed in these sources to express causality include the prepositions di (from Ptg. de ‘of, from’, a multifunctional preposition in the Tugu and Batavia corpora, meaning ‘of, from, out of, by’; see Maurer 2011: 6, 224) and perbida (Maurer 2011: 87), and conjunctions dibida di, and perbida, in addition to the Indonesian subordinator sebab ‘because’ (Maurer 2011: 109). Of these, dibida and perbida appear to contain a *[V/B]IDA-type root, added either to the already mentioned preposition di, or to per (also par/por) ‘for, in order to’ (see 37 Hugo C. Cardoso Maurer 2011: 278). In the sources, perbida introduces both clauses of cause8 (9a) – very often with an explicative function (9b) – and NPs (9c), whereas dibida di only occurs once, introducing a causal explicative clause (9d): (9) a. [Tugu Creole; Schuchardt 1890: 47; see Maurer 2011: 152] Noos fadji sirbis grósoe, noengsabe fadji sirbis finjoe, [perbida noos djèntih serah]. CAUS people countryside 1p ‘We do rough work, [we] don’t know how to do delicate work, because we [are] countryside people.’ b. [Tugu Creole; Schuchardt 1890: 44; see Maurer 2011: 147-148] Ingoewal toedoe nasèdoe nasèdoe na těra Djawa, thing alber neli ki koeměsoe [perbida toedoe pobes CAUS all kiä-tja diake people těra Djawa koemi aros land Java eat rice froeta alber neli]. REL-get from.DEM fruit plant rice ‘Among all that grows on the land of Java, rice is the earliest, because all the Java people eat the rice they get from the rice plant.’ c. [Tugu Creole; Schuchardt 1890: 43; Maurer 2011: 146] ki thing noentè inggoeäl koe oter REL COP NEG.COP same as other djenti djenti kampong naäli [perbida soeä person~person village LOC-there CAUS POSS modoe ò redjang]. manner and language ‘those who are there are not the same as the other people of the village because of their customs and their language.’ 8 Only one sentence in these sources contains what may be unequivocally classified as a clause of reason, and it is introduced by sebab (Schuchardt 1890: 64). 38 The synchrony and diachrony of an Asian-Portuguese causal morpheme d. [Tugu Creole; Schuchardt 1890: 43; Maurer 2011: 146] Djenti Toegoe atja tantoe bondadi disïoe Gobernemen [dibida CAUS disïoe Gobernemen dja da koe nos oenga of.man government PST give OBJ 1p one gredja per conserta doti prindèsoe (…)] church for exert ? religion ‘The people of Tugu have received much favour from the Government people, because the Government people have given us a church to practice our religion (…)’ Furthermore, an anonymous 1937 manuscript containing words and a few sentences in Dutch, Indonesian, and Tugu Portuguese, attributed to a Dutch priest (see Maurer 2011: 2), translates Dutch omdat ‘because’ as aka bida in Tugu Portuguese, though no example sentence is provided. However, considering that, in Dutch, omdat only introduces clauses and that aka bida consists of a demonstrative plus the causal morpheme, it is likely that it would also be used with clauses, rather than nominals. This is the extent of the information found in sources that can unequivocally be seen to record the Portuguese-based creole spoken on the island of Java. However, a manuscript collection of pantoons in Malay and Creole Portuguese,9 which belonged to Hugo Schuchardt and was passed on to Leite de Vasconcelos, contains a relevant causal form, in the verses transcribed in (10): (10) Soo per wia wos: / qiú penna agora / mostra rijaldadie / only by cause 2s suffer now COMPL.1s Ô dossie senhora. Oh sweet lady show toughness ‘It is because of you / that I suffer now. / (You) demonstrate toughness, / oh sweet lady.’ 9 Panton Malaijoe dan Portugees, currently preserved in the library of the Museu Nacional de Arqueologia, Lisbon. The manuscript was located and brought to my attention by Prof. Ivo Castro. 39 Hugo C. Cardoso The exact geographical origin of the manuscript is not known, but the Indonesian archipelago is a likely candidate; nor is the date of the text clear, although it was definitely composed no later (and probably much earlier) than the 1880s. The *[V/B]IDA-related form recorded there is highly significant, not only given the dearth of attestations from the Southeast Asian region but also because of its particular form: wia (see section 3 below). What makes these data from Batavia and Tugu (plus the anonymous manuscript) especially interesting is that the other (South)east Asian regions that host or have hosted a Portuguese-based creole provide no evidence of similar *[V/B]IDA-type causal morphemes. In the case of the modern creole of Malacca, which has been documented and studied in depth, the form of the most frequent causal morpheme is kauzu (from Ptg. causa ‘cause’). Kauzu, or a reduced form kau, can introduce causal nominals – either mediated by a relator di (postposed; see Baxter 1988: 162) or genitive sa (preposed; see Baxter 1988: 207; Baxter & de Silva 2004: 31) – and clauses – either bare or followed by (di) ki (see Baxter 1988: 13, 203-204; 2013: 127-128). Akeh kauzu or isti kauzu, the Malacca variant of the familiar combination of a demonstrative and a causal operator, is used to establish a result nexus between two assertions (Baxter & de Silva 2004: 13, 31, 98). In addition, kifoi (from Ptg. que ‘what’ + foi ‘be/go.PST’), elsewhere used as a Q-word ‘why’, may also introduce clauses of cause (Baxter 1988: 203-204). Baxter & de Silva (2004: 50) further record a causal conjunction porkih (from Ptg. porque ‘because’), which, however, they classify as archaic.10 Still in Southeast Asia, the few records of Timorese creole contain no instances relevant for our study (Baxter 1990; Baxter & Cardoso 2017). In the case of the creole of Macau, recent (20th-century) investigations reveal a range of causal operators which are not dissimilar form those found in Malacca. Ferreira (1990[1978]: 249) records the causal conjunctions qui and po qui (presumably from Ptg. porque ‘because’). Arana-Ward’s research of the Macanese community in Hong Kong also registers the form porke but indicates an interesting alternative bikozo, which is clearly an English loan (Arana-Ward 1977: 150). In his corpus of Macau creole as spoken among the Macanese diaspora in Canada and the United States of America, Pinharanda Nunes (2010) encounters not only the causal subordinator porqu, but also pocósa (from Ptg. In fact, in a partial 19th-c. translation of the Gospels identified as being “escripto no crioulo de Malacca” (‘written in the creole of Malacca’), the causal subordinator parqui can be found. However, as Baxter (2018: 264) points out, it is important to take this evidence with a grain of salt, since it has been established that missions in Malacca made use of Indo-Portuguese translations produced in South Asia. 10 40 The synchrony and diachrony of an Asian-Portuguese causal morpheme por causa ‘because [Lit. ‘by cause’]’) and promódi (from Ptg. por modo de or, possibly, por meio de ‘by way of’), and records both po isso (from Ptg. por isso ‘for that/therefore’) and so (an English loan) introducing result clauses. Earlier written records of Macau creole also contain several of these forms and add a few more. 19th-century texts (collated in Barreiros 1943-4) provide plentiful instances of porque, but also use qui fui/foi (from Ptg. que ‘what’ + foi ‘be/go.PST’; cognate with Malacca kifoi; see Barreiros 1943-4: 31) to introduce causal clauses, and porisso (que/qui) (from Ptg. por isso + que ‘that’; see Barreiros 1943-4: 252-254, 350, 353, 360) with result clauses. With causal nominals, the early texts in Barreiros (1943-4: 254) record pro cosu di (from Ptg. por causa de ‘because of [Lit. ‘by cause of’]’). 3. Etymology Having established the geographical distribution of *[V/B]IDA-type causal morphemes in Asia, we will now explore their etymology. Commenting on the Tugu Creole forms perbida and dibida di, Maurer proposes an etymological derivation from Ptg. para ‘for’ + vida ‘life’ (2011: 282), in the first case, and de ‘of’ + vida + de (Maurer 2011: 225), in the second. If correct, Maurer’s etymological proposal could, of course, be extended to MIP and SLP, though with the suppression of the Ptg. preposition por or para which accounts for part of the causal form from Java. The identification of ‘life’ as one of the components of these causal operators finds some support in the fact that, in the creoles seen to have used a *[V/B]IDA-type morpheme, the corresponding word for ‘life’, derived from Ptg. vida ‘life’, is very similar (bida/vida in Tugu, viida in SLP, unattested in MIP). However, I suggest a different etymology here, viz. a derivation from Ptg. por via (de) ‘by way of’. One reason for this is that this expression, unlike por vida de, features more or less robustly in diachronic and dialectal corpora of European Portuguese with causal readings (among others). To explore the history of both expressions in Portuguese, I surveyed several diachronic corpora of Portuguese, with the following three yielding relevant results: - Corpus Eletrónico do CELGA – Português do Período Clássico (CECPPC) – 16th-17th centuries; - Corpus P.S. Post Scriptum (CLUL 2014), an archive of private letters – 16th-20th centuries; - Corpus Tycho Brahe (Galves et al. 2017) – 15th-20th centuries. 41 Hugo C. Cardoso Even though written corpora of Portuguese are somewhat limited for the period preceding the 16th century, they do attest a wide range of values for por via (de) in different periods, including cause. Since we are interested in ascertaining whether it occurs at a time that is consistent with the formation of the AsianPortuguese creoles, I first restricted the search for attestations to the period between the early 15th century (when Portuguese expansion took off) and the mid-17th century (when Portuguese decline in Asia intensified). For that period, we find this expression, with the forms por via + genitive nominal (NP/pronoun), or por + pronoun + via, introducing constituents which carry the values of Path (11a), Means (11b), Manner (11c), or – significantly for our purposes – Cause (11d) and Reason (11e): (11) a. [1542; Diogo do Couto, Décadas; Tycho Brahe] Ruy Lopes de Villa-Lobos vendo que arribara o São Joanilho, determinou de mandar recado a Hespanha [por via da Índia], e pera isto se falou com um Gaspar Melio e lhe deo instruções. ‘Seeing as the São Joanilho had arrived, Ruy Lopes de Villa-Lobos decided to send a message to Spain by way of India, and to do so spoke to a certain Gaspar Melio with instructions.’ b. [1502; Pêro Magalhães de Gandavo, História da Província de Santa Cruz; Tycho Brahe] E a maneira de como isto se veio a denunciar e ter por coisa averiguada, foi [por via dos Índios da terra]. ‘And the manner by which this came to be known and verified was through the Indians of the land.’ c. [1502; D. João III, Cartas de D. João III; Tycho Brahe] Agora que o concerto é feito, hei por meu serviço que não faleis nisto [por via de descontentamento], senão mui amigavelmente […] ‘No that the agreement is made, I consider it to be in my interest that you talk about this not in displeasure, but rather in a very friendly way’ 42 The synchrony and diachrony of an Asian-Portuguese causal morpheme d. [1556; Frei Luís de Sousa, A Vida de Frei Bertolameu dos Mártires; Tycho Brahe] […] por isso muito torno a pedir a Vossa Reverência que se esforce muito e creça seu zelo em apagar o fogo da luxúria por esse arcebispado e de nenhuma cousa tenha tanto escrúpulo como de ser remisso nisso e por nenhuma outra cousa tema tanto o juízo de Deus, e preze-se muito de ganhar enemigos [por esta via] e excitar línguas contra si, porque padecer isto é sinal da eterna predestinação. ‘therefore I once again ask Your Reverence to make great efforts and that your zeal in extinguishing the fire of lust in that archdiocese may grow, and that you do not concern yourself as much with anything else as with being remiss in this, and that you do not fear the judgement of God as much for anything else, and that you take pride in making enemies because of this and exciting tongues against you, because suffering this is a sign of eternal predestination.’ e. [1502; D. João III, Cartas de D. João III; Tycho Brahe] E porque me parece, segundo o que dizeis em vossa carta, que os dez mil cruzados que prometeste ao Almirante não são [por via do concerto], mas que vós lhe os prometestes por vos parecer que era assim muito meu serviço […] ‘And since it seems to me, according to what you say in your letter, that the ten thousand cruzados you have promised the Admiral were not due to the agreement but because you have promised them to him as you thought this was in my interest’ If we expand the temporal width of the search, we notice that, for later periods, the corpora – especially P.S. Post Scriptum, the one that collects non-official documents, including much more informal and colloquial written productions, and mostly in the 19th century – yield the same type of data but also feature the expression (por via de or por via que) introducing clauses with the values of Reason (12a) and Purpose (12b): (12) a. [1827; ‘Carta de Pedro Fernandes, boleeiro, para José Matias Monteiro, soldado’; P.S. Post Scriptum] Agora o eu Retirarme da sua Caza foi [por via de eu Saber o seu genio e Vmce não atender a Rezão] ‘Now, the reason for me to retire from your home was because I am aware of your temper and [because] you do not listen to reason’ 43 Hugo C. Cardoso b. [1822; ‘Carta assinada sob nome falso, Pedro Leal, para destinatário não identificado’; P.S. Post Scriptum] entregarlhe este dinheiro embrolhado em huma coisa q paressa huma emcomenda e q seja em Ouro ou em Papel, [por via q os mais prezos não persebam o que he] ‘To deliver this money to him wrapped in something that looks like a parcel and that it be in gold or paper, so that the other prisoners do not realise what it is’ In both periods, the combination with a demonstrative (por esta via or por via disso), which anaphorically recuperates a previous predication, also occurs in what appear to be result contexts, as in (13); however, the occurrences are few and the contexts somewhat unclear: (13) [1502; D. João III, Cartas de D. João III; Tycho Brahe] Gaspar Vaz me escreveu, que Afonso de Sevilha lhe dissera, que a minha armada metera no fundo uma nao em que ele tinha parte, e mandara secretamente um correio a essa cidade, e que lhe o negara, e que [por esta via se tem lá todos os avisos das minhas armadas]. ‘Gaspar Vaz wrote to me that Afonso de Sevilha told him that my armada had sunk a vessel of which he was a partner, and that [he] had secretly sent a message to that city, and that [he] had denied it, and that, therefore, everyone in those parts is wary of my armadas.’ To complement the diachronic study, I also surveyed the use of por via (de) in a synchronic corpus of dialectal European Portuguese, CORDIAL-SIN (Martins 2000-), which contains interviews collected in several regions of the country. While the use of the expression is not particularly frequent, it occurs, in this corpus, exclusively to introduce clauses of Cause (14a) and Purpose (14b), in productions from the districts of Horta (Azores), Bragança, Vila Real, Viana do Castelo (Northern Portugal), and Setúbal (Central Portugal): (14) a. [Larinho (Bragança); CORDIAL SIN, LAR07-C] Os estrumos perto dos animais não estão bem. Estão melhor no campo. Porque aparecem muitas doenças é [por via de os estrumos estarem perto dos palheiros]. ‘Manure placed near the animals is not right. It is better in the fields. The reason why many diseases appear is because manure is close to the barns.’ 44 The synchrony and diachrony of an Asian-Portuguese causal morpheme b. [Fajãzinha (Horta); CORDIAL-SIN, FLF50-C] E do 'rabisco' pega duas tiras que vai pregada à sela, que é [por via de a sela não caminhar para a frente]. ‘And from the rear-piece depart two straps that attach to the saddle, which is so that the saddle doesn’t move forward.’ Por vida de, the alternative etymology, is much less frequent in the corpora with values that are consistent with those of *[V/B]IDA in the Asian-Portuguese creoles. The diachronic corpora only reveal one instance of por vida de in which it can be said to introduce a cause NP: (15) [1653; ‘Carta não autógrafa de Francisco Gomes Henriques para Jerónimo Nunes’; P.S. Post Scriptum] Ao amigo, e amigos minhas lembranças e que mto deveras me encomenden a ds e a nossa andrada, do que fico certo [por vida de nossa amizade] […] ‘My regards to my friend [i.e. the addressee] and my friends, and may you commend me much to God and to our Andrada, of which I am certain on account of our friendship’ Other than this, por vida de is frequent in the corpus, but only as a rhetorical formula when entreating someone (16a) or as contingent in a promise (16b), and only with nominals: (16) a. [1672; ‘Carta de Abigail de Brito, [dona de roça], para Francisco de Medina’; P.S. Post Scriptum] 6 culheres de prata que devo a 2 annos a hũa ingreca tenho mandado em mtas pedir a Vm agora o torno a fazer [por vida de Vm] sendo posivel me faca esta mce […] ‘6 silver spoons which I have owed an Englishwoman for 2 years, I have asked you for it many [times, and] now I do it again, that, upon your life, if it is possible, you do me this favour’ b. [1672; ‘Carta de Abigail de Brito, [dona de roça], para Francisco de Medina’; P.S. Post Scriptum] […] que lhe juro a Vm [por vida de todos os meos filhos] avera sinco meces perecemos e não somos nos sos se não em geral […] ‘that I swear to you on all my children’s lives that we have been perishing for some five months now, and not just us but [everyone] in general’ 45 Hugo C. Cardoso It should be noted that, in any of its functions, por vida de never introduces clauses in the corpora, unlike por via de (though only attested for a later period). If we combine this fact with the demonstrated functional range of both expressions in the corpora, it is safe to say that the diachronic and dialectal research supports por via de as a more likely ancestor of the *[V/B]IDA-type Asian-Portuguese morphemes than por vida de, despite the fact that, formally, vida appears to be closer to the Asian-Portuguese morphemes (with the exception of per wia, given in example (10) above). However, it is a significant fact that most West African Portuguese-based creoles, as well as Papiamentu, have causal conjunctions/adpositions whose forms are much more clearly reflective of por via (de) (see Jacobs 2012: 111-113 for an overview): a) in Forro/Santome: plôvya (di) (Araújo & Hagemeijer 2013: 76) is overwhelmingly a preposition occurring with NPs but there are also rare attestations introducing a clause (Hagemeijer, p.c.), even though, as a causal operator, it is secondary to the alternative form punda; b) in Principense: pôvya is defined as a conjunction ‘because’, although there are other causal subordinators – modi, pidi¸ pôkê (Maurer 2009: 168, 233); c) in Cape Verdean Creole: pabia / pabiâ / pa biâ (di) is a causal operator, though seen as dialectal and archaic (Brüser et al. 2002: 529; Jacobs 2012: 111), therefore not the dominant form at present; d) in Guinea-Bissau Creole, or Kriyol: pabia (di) is a regular causal conjunction, alongside other forms such as purke (Doneux & Rougé 1988: 56; Scantamburlo 1999: 184); e) in Casamancese Creole: pabiya or parbiya is used as a causal connector with clauses and NPs, and also as a Q-word ‘why’ (Biagui 2012: 272ff; Biagui & Quint 2013: exs. 34-20, 34-68); f) in Papiamentu: Jacobs (2012: 111) identifies causal pa via / pa bia (plus di with nominals, plus ku with clauses) but considers it archaic, having been replaced with alternative expressions in this language. Since the formation of the West African creoles predates that of the Asian creoles, these African causal morphemes provide evidence that causal por via (de) was available in oral Early Modern Portuguese even before it reached Asian shores, and sufficiently salient to be retained in the developing creoles as a functional element, lending support to the proposed derivation of *[V/B]IDA. The only remaining issue to resolve in this scenario, therefore, is that of the particular transformation of the Portuguese expression into *[V/B]IDA in Asia. One possibility is that, at some point of the creoles’ history (either at the 46 The synchrony and diachrony of an Asian-Portuguese causal morpheme formative stage or later), the morphological boundaries of the Portuguese complex via de were reanalysed to incorporate both via and the preposition de into a single form. The absence of a morphological boundary is clear in the current stage of MIP and SLP, in which suydɐ and viida are postposed to the NP/clause and de/da/dɐ no longer correspond to any adposition – which, in the case of MIP, has resulted in the necessity to add another linker morpheme -sə, alien to the etymological construction but, as mentioned earlier, required by several of this language’s postpositions. As for the earlier corpus of SLP (Ceylon Portuguese), the issue of the morphological unity of vide/vidè/videque is a little less clear. There, the causal operator is preposed to the NP or clause, and de does occur in other contexts as a genitive/ablative preposition. This means that, in theory, it is possible that de was (still) a linker preposition and interpreted as such. One argument against this is that, even though de was available as an independent preposition to the writers of these texts, they never apply an orthographic boundary between *vi and de, nor is *vi ever recorded as an entry in the coeval dictionaries or given as such in the grammars of Ceylon Portuguese. In both MIP and SLP, the preposition por, which is an integral part of the Portuguese expression, has left no trace, unlike in the case of the West African creoles. In the case of Tugu, it can be recognised in the first section of perbida. However, it is not clear whether this form was interpreted as monomorphemic, not only because per is a productive preposition elsewhere in these texts, but also because it is replaced with a different Tugu Creole preposition in the attested form dibida.11 4. Discussion This survey of causal morphemes in the Asian-Portuguese creoles has revealed a range of alternative forms, most of which are derived from Portuguese. One of these consists of or integrates a sequence cognate with *[V/B]IDA, which has been the main object of this study, with the available linguistic sources providing evidence of cognate operators in the Portuguese-lexified creoles of the Malabar (currently), Sri Lanka (currently and in the past), and Tugu (in the past). An alternative etymological derivation for dibida could be Ptg. devido (a) ‘due to’. This does not seem likely, however, because a Batavia/Tugu-internal development is plausible from the attested morphological elements – and that too with resort to a linker (di) which is indeed associated with causal relations in this creole. 11 47 Hugo C. Cardoso Of two proposed etymologies – por vida de and por via de –, research in diachronic and dialectal corpora of European Portuguese revealed that por via de is the most likely candidate, as it is shown to have (had) functions and a syntactic distribution (with both NPs and clauses) that are consistent with those of the cognate forms in the Asian creoles. In addition, the West African Portuguese-lexified creoles have also retained forms derived from por via de to express a nexus of causality. In formal terms, however, this scenario requires a morphological reinterpretation to conflate the via and the de elements of the Portuguese expression, an instance of change which is not unlikely but also not evident (as demonstrated by the fact that it was not effected in the case of the West African Portuguese-lexified creoles). This has a few interesting implications for our understanding of the diachrony and relationship between the various Asian-Portuguese creoles. On the one hand, the coincidence of lexical resources between MIP and SLP is not unexpected, and has been highlighted before in support of a certain historical relatedness, resulting from either retention of an initial form of contact Portuguese, or diffusion dynamics linking Southern India and Sri Lanka (Cardoso 2013). The fact that the Portuguese presence in both these regions decreased significantly after the mid-17th century, as the result of a Dutch colonial takeover, indicates that the cooption of Ptg. por via de for causal functions and even the morphological reinterpretation which has resulted in a *[V/B]IDA-type root must have taken place before that, although, in truth, the scope for cultural and linguistic circulation between the Luso-Asian communities of both regions in later periods (under Dutch and British rule) has not yet been properly assessed. The presence of a cognate morpheme on Java, on the other hand, requires a channel for this form to diffuse no earlier than the 17th-century, when Portuguese(-creole)-speaking populations began to converge on the region of Batavia from various parts of Asia under Dutch rule – including the Malabar and Ceylon (Maurer 2011: 3-5). In order to approach the issue of where this particular form may have travelled to Java from, it is important to realise that the non-attestation of similar forms in the corpora of other Asian-Portuguese creoles (Diu, Daman, Korlai, Malacca, Macau, and Bidau) does not necessarily mean they were never used in those varieties. As such, we cannot be certain that it was the incoming populations from either the Malabar or Ceylon who brought along this form, even though we have evidence of that particular migration. Nonetheless, the distribution we can observe does lend credence to the hypothesis of a South Asian origin. As a matter of fact, there are a few other lexemes in the Batavia/Tugu corpus (see the wordlist in Maurer 2011: 199ff) 48 The synchrony and diachrony of an Asian-Portuguese causal morpheme which are reminiscent of the southern Indo-Portuguese varieties (especially SLP) but appear have no discernible equivalent in the Southeast Asian creoles, including: - Batavia/Tugu alung, alunga ‘some’ [from Ptg. algum(a) ‘some’]; in MIP aluŋɐ ‘some’, in SLP álung ‘some’; - Batavia ascroea/askoera ‘hulled rice’ [from Ptg. arroz ‘rice’ + cru(a) ‘raw']; in SLP áskruuva ‘raw rice’;12 - Tugu beenvetoe ‘well’ [from Ptg. bem-feito ‘well-done’]; in MIP befet, in SLP befeentu;13 - Batavia kadora ‘always’ [from Ptg. cada ‘each’ + hora ‘time’]; in SLP kadóóra ‘always’; - Batavia/Tugu kaaudoe/kawdu ‘tortoise’ [from Ptg. cágado ‘tortoise’]; in SLP kávdu ‘tortoise’; - Batavia werdoera ‘vegetables’ [from Ptg. verdura ‘greenness; vegetables’]; in SLP verduura ‘vegetables’, in MIP værdurɐ ‘vegetables’. Other lexemes in the Batavia/Tugu corpus may have close equivalent forms in Southeast Asian creoles (especially Malacca) but nonetheless display certain formal characteristics which establish a closer connection with cognates found in the southern Indo-Portuguese varieties; e.g.: - Tugu matjika ‘to tread’ [from Ptg. machucar? ‘to crush, to pound, to hurt’], with a high front vowel /i/ in the first syllable; in SLP machikáá ‘to press, to crush’; in Malacca Creole machukah ‘to crush’ (Baxter & de Silva 2004: 38); - Tugu metay/météj ‘half’ [from Ptg. metade ‘half’], having dropped the intervocalic /d/ of the Portuguese etymon; in SLP metááy ‘half’; in Malacca Creole metadi (Silva & de Silva 2004: 42); - Batavia/Tugu palmiang ‘morning’ [from Ptg. pela manhã ‘in the morning’], with a lateral consonant /l/ in the coda of the first syllable; in SLP pálmiyaam ‘morning’; in Malacca Creole pamiang ‘morning’ (Baxter & de Silva 2004: 46); - Batavia/Tugu saaudoe/saoedoe ‘saturday’ [from Ptg. sábado ‘saturday’], with the transformation of the plosive /b/ of the Portuguese etymon; 12 In Malacca Creole, the expression that contains both etymological components of these forms, aros krua, means ‘acne’ (Baxter & de Silva 2004: 5). 13 Bemfetu in Malacca means ‘beautiful, handsome’ (Baxter & de Silva 2004: 63), therefore an adverbial use similar to that found in Tugu and SLP would only be possible as a metaphor. 49 Hugo C. Cardoso in SLP sáávdu ‘saturday’, in MIP sawdi ‘Saturday’ [from Ptg. sábado + dia ‘day’]; in Malacca Creole sabdu ‘Saturday’ (Baxter & de Silva 2004: 53); - Batavia soesti ‘to happen’ [from Ptg. suceder ‘to happen’], with a voiceless plosive /t/ in the final syllable; in SLP susta ‘to happen’; in Malacca Creole susudeh or sudeh ‘to happen’. However, the lexicon is not the only domain to reveal potential links between South and Southeast Asia, when it comes to characteristics associated with Asian-Portuguese communities. In the domain of oral and musical traditions too, there is also a significant parallel between the two regions, namely in the preservation of musical and dance genres known by terms cognate with cafrinha. To this day, the Portuguese Burghers of Sri Lanka cultivate a highly distinctive music-dance tradition known as Káfriinha, while Tugu musicians perform the Kafrinyu/Cafrinho as part of the Keroncong Tugu musical heritage – both traditionally sung in Creole Portuguese; according to Tan (2016: 146), however, a musical connection between the two traditions has not been clearly shown. Interestingly, this term recurs elsewhere in insular Southeast Asia, including in Sumatra – where the Sikambang kapri genre and the Kaparinyo song can be found (Kartomi 2012: 221ff, 406) – as well as the Moluccas and Timor – which saw the establishment of the Kafrínu or Kafrinia genres, respectively (Dalgado 1919: 474). In conclusion, while these regional links may be tentative at this stage, they do conspire with the evidence unearthed with respect to the use of *[V/B]IDA-derived causal morphemes to call for a systematic assessment of the South Asian – especially Sri Lankan – contribution to the extinct creole of Batavia and Tugu. Abbreviations ART=Article; CAUS=Cause marker; COMP=Comparative marker; COMPL=Complementiser; COP=Copula; DEM=Demonstrative; DSLP=Documentation of Sri Lanka Portuguese; EMPH=Emphatic marker; EXS=Existential; FOC=Focus marker; FUT=Future; GEN=Genitive; HON=Honorific; INF=Infinitive; IPFV=Imperfective; IRR=Irrealis; LOC=Locative; M=Masculine; MIP=Malabar Indo-Portuguese; NEG=Negator; NP=Noun phrase; OBJ=Object marker; OBL=Oblique; OBLIG=Obligative; p=Plural; PFV=Perfective; POSS=Possessive; PRS=Present; PST=Past; PTCP=Participle; Ptg.=Portuguese; PURP=Purpose marker; Q=Question particle; REAS=Reason marker; REL=Relative particle; s=Singular; SLP=Sri Lanka Portuguese 50 The synchrony and diachrony of an Asian-Portuguese causal morpheme References Araújo, Gabriel & Tjerk Hagemeijer. 2013. Dicionário Livre Santome/Português. São Paulo: Hedra. Arana-Ward, Marie. 1977. A synchronic and diachronic investigation of Macanese: The Portuguese-based creole of Macao. Hong Kong: The University of Hong Kong, M.A. dissertation. Bakker, Peter. 2006. The Sri Lanka Sprachbund: The newcomers Portuguese and Malay. In Yaron Matras, April McMahon & Nigel Vincent (eds.), Linguistic Areas: Convergence in Historical and Typological Perspective, 135-159. New York: Palgrave. Barreiros, Danilo L. 1943-4. O Dialecto Português de Macau: Antologia. Macau: Renascimento. Baxter, Alan N. 1988. A Grammar of Kristang (Malacca Creole Portuguese) (Pacific Linguistics. Series B; no. 95). Canberra: Pacific Linguistics. Baxter, Alan N. 1990. Notes on the Creole Portuguese of Bidau, East Timor. Journal of Pidgin and Creole Languages 5(1). 1-38. Baxter, Alan N. 2013. Papiá Kristang. In Susanne Michaelis, Philippe Maurer, Martin Haspelmath & Magnus Huber (eds.), The Survey of Pidgin and Creole Languages, vol. 2, 122-130. Oxford: Oxford University Press. Baxter, Alan N. 2018. Malacca Creole Portuguese in the 19th century; Evidence of a wider lectal range? Journal of Pidgin and Creole Languages 33(2). 247-279. Baxter, Alan N. & Hugo Cardoso. 2017. Early notices regarding Creole Portuguese in former Portuguese Timor. Journal of Language Contact 10. 264-317. Baxter, Alan N. & Patrick de Silva. 2004. A Dictionary of Kristang (Malacca Creole Portuguese), with an English-Kristang Finderlist. Canberra: Pacific Linguistics/Australian National University. Berrenger. 1811. A Grammatical Arrangement on the Method of Learning the Corrupted Portuguese as Spoken in India, 2nd ed. Colombo: Frans de Bruin. Biagui, Noël Bernard. 2012. Description générale du Créole Afro-Portugais parlé à Ziguinchor (Sénégal). Université Cheikh Anta Diop (UCAD)/Institut National des Langues et Civilisations Orientales (INaLCO), PhD dissertation. Biagui, Noël Bernard & Nicolas Quint. 2013. Casamancese Creole structure dataset. In Susanne Michaelis, Philippe Maurer, Martin Haspelmath & Magnus Huber (eds.), Atlas of Pidgin and Creole Language Structures Online. Leipzig: Max Planck Institute for Evolutionary Anthropology, http://apics-online.info/contributions/34. (28 November, 2018.) Brüser, Martina & André dos Reis Santos. 2002. Dicionário do Crioulo da Ilha de Santiago (Cabo Verde), Jürgen Lang (ed.). Tübingen: Gunter Narr. Callaway, John. 1820. A Vocabulary, in the Ceylon Portuguese, and English Languages, with a Series of Familiar Phrases. Colombo: Wesleyan Mission Press. Cardoso, Hugo C. 2006-2015. Corpus of Cannanore Indo-Portuguese speech. Unpublished, Centro de Linguística da Universidade de Lisboa. Cardoso, Hugo C. 2007-2010. Corpus of Vypeen [Cochin] Indo-Portuguese speech. Unpublished, Centro de Linguística da Universidade de Lisboa. Cardoso, Hugo C. 2009. The Indo-Portuguese Language of Diu. Utrecht: Landelijke Onderzoekschool Taalwetenschap. 51 Hugo C. Cardoso Cardoso, Hugo C. 2013. The southern cluster: Indo-Portuguese in the Malabar and in Sri Lanka. Paper presented at the joint conference of the Society for Pidgin and Creole Linguistics and Associação Crioulos de Base Lexical Portuguesa e Espanhola, University of Lisbon, June 22nd. Cardoso, Hugo C. 2014. Factoring sociolinguistic variation into the history of Indo-Portuguese. Revista de Crioulos de Base Lexical Portuguesa e Espanhola 5. 87-114. Cardoso, Hugo C. 2015. Corpus of Calicut Indo-Portuguese speech. Unpublished, Centro de Linguística da Universidade de Lisboa. Cardoso, Hugo C. 2017. Documentation of Sri Lanka Portuguese. London: SOAS, Endangered Languages Archive, https://wurin.lis.soas.ac.uk/Collection/MPI1035102. (14 November, 2018.) Cardoso, Hugo C. 2018. Oral Corpus of Malabar Indo-Portuguese Creole. London: SOAS: Endangered Languages Archive, https://elar.soas.ac.uk/Collection/MPI1225883. (14 November, 2018.) Cardoso, Hugo C. & Umberto Ansaldo (guest eds.). 2009. Accounting for similarities among the Asian-Portuguese varieties. Journal of Portuguese Linguistics 8(2) (Special issue). Cardoso, Hugo C., Alan N. Baxter & Mário Pinharanda Nunes (eds.). 2012. Ibero-Asian Creoles: Comparative Perspectives. Amsterdam: John Benjamins. Cardoso, Hugo C., Tjerk Hagemeijer & Nélia Alexandre. 2015. Crioulos de base lexical portuguesa. In Maria Iliescu & Eugeen Roegiest (eds.), Manuel des Anthologies, Corpus et Textes Romans, 670-692. Berlin: Mouton de Gruyter. Clements, J. Clancy. 1996. The Genesis of a Language; The Formation and Development of Korlai Portuguese. Amsterdam/ Philadelphia: John Benjamins. Clements, J. Clancy. 2000. Evidência para a existência de um pidgin português asiático. In Ernesto d'Andrade, Dulce Pereira & Maria Antónia Mota (eds.), Crioulos de Base Lexical Portuguesa. Braga: Associação Portuguesa de Linguística. Clements, J. Clancy. 2012. Luso-Asian comparatives in comparison. In Hugo C. Cardoso, Alan N. Baxter & Mário Pinharanda Nunes (eds.), Ibero-Asian Creoles: Comparative Perspectives, 15-46. Amsterdam: John Benjamins. Clements, J. Clancy. 2013. Korlai structure dataset. In Susanne Michaelis, Philippe Maurer, Martin Haspelmath & Magnus Huber (eds.), Atlas of Pidgin and Creole Language Structures Online. Leipzig: Max Planck Institute for Evolutionary Anthropology, http://apics-online.info/contributions/41. (15 November, 2018.) Clements, J. Clancy. 2014. Lectal differences in Daman Indo-Portuguese. Revista de Crioulos de Base Lexical Portuguesa e Espanhola 5: 115-156. Clements, J. Clancy & Andrew J. Koontz-Garboden. 2002. Two Indo-Portuguese creoles in contrast. Journal of Pidgin and Creole Languages 17(2). 191-236. CLUL (ed.). 2014. P.S. Post Scriptum. Arquivo Digital de Escrita Quotidiana em Portugal e Espanha na Época Moderna, http://ps.clul.ul.pt. (3 June, 2017.) Croft, William. 1991. Syntactic Categories and Grammatical Relations. Chicago: University of Chicago Press. Dalgado, Sebastião R. 1919. Glossário Luso-Asiático. Coimbra: Imprensa da Universidade. Dhongde, Ramesh V & Kashi Wali. 2009. Marathi (London Oriental and African Language Library 13). Amsterdam/Philadelphia: John Benjamins. Doneux, Jean Léonce & Jean-Louis Rougé. 1988. En Apprenant le Créole à Bissau ou Ziguinchor. Paris: L’Harmattan. 52 The synchrony and diachrony of an Asian-Portuguese causal morpheme Ferraz, Luís I. 1987. Portuguese creoles of West Africa and Asia. In Glenn Gilbert (ed.), Pidgin and Creole Languages, 337-359. Honolulu: University of Hawai’i Press. Ferreira, José dos Santos. 1990. Doci Papiaçám di Macau. Macau: Instituto Cultural [Includes 1978. Papiá Cristâm di Macau: Epitome de Gramática Comparada e Vocabulário Dialecto Macaense. Macau: Tipografia da Missão]. Fox, William Buckley. 1819. A Dictionary in the Ceylon-Portuguese, Singhalese and English languages... to which is Prefixed a Compendium of the Ceylon-Portuguese Language. Colombo: Wesleyan Mission Press. Galves, Charlotte, Aroldo Leal de Andrade & Pablo Faria. 2017. Tycho Brahe Parsed Corpus of Historical Portuguese, http://www.tycho.iel.unicamp.br/~tycho/corpus/texts/ psd.zip. (3 June, 2017.) Iken, Sebastião. 2000. Morphosyntactic realignment in Dravido-Portuguese Creoles under a diachronic perspective. In Charles J. Borges, Óscar G. Pereira & Hannes Stubbe (eds.), Goa and Portugal: History and Development, 348-365. New Delhi: Ashok Kumar Mithal. Jacobs, Bart. 2012. Origins of a Creole: The History of Papiamentu and its African Ties. Berlin: De Gruyter Mouton. Kartomi, Margaret. 2012. Musical Journeys in Sumatra. Urbana/Chicago/Springfield: University of Illinois Press. Luraghi, Silvia. 2005. Paths of semantic extension. From cause to beneficiary and purpose. In Michael Fortescue, Eva Jensen, Jens Mogensen & Lene Schøsler (eds.), Historical Linguistics 2003. Selected Papers from the 16th International Conference on Historical Linguistics, 141-157. Amsterdam: John Benjamins. Martins, Ana Maria (coord.). 2000-. CORDIAL-SIN: Corpus Dialectal para o Estudo da Sintaxe / Syntax-oriented Corpus of Portuguese Dialects. Lisbon: Centro de Linguística da Universidade de Lisboa, http://www.clul.ulisboa.pt/en/10-research/314-cordial-sincorpus. (3 June 6, 2017.) Maurer, Philippe. 2009. Principense: Grammar, Texts, and Vocabulary of the Afro-Portuguese Creole of the Island of Príncipe, Gulf of Guinea. London/Colombo: Battlebridge. Maurer, Philippe. 2011. The former Portuguese Creole of Batavia and Tugu (Indonesia). London: Battlebridge. Newstead, Robert (transl.). 1852. O Novo Testamento de Nossa Senhor e Salvador Jesus Christo. Traduzido ne Indo-Portugueza. Colombo: Wesleyan Misison. Pinharanda Nunes, Mário. 2010. Estudo da expressão morfo-sintáctica das categorias de Tempo, Modo e Aspecto em Maquista. Macau: Universidade de Macau, PhD dissertation. Scantamburlo, Luigi. 1999. Dicionário do Guineense, vol. 1 [Introdução e notas gramaticais]. Lisbon: Edições Colibri/FASPEBI. 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. Schuchardt, Hugo 1889. Beiträge zur Kenntnis des kreolischen Romanisch: VI. Zum Indoportugiesischen von Mahé und Cannanore. Zeitschrift für Romanische Philologie 13. 516-524. Schuchardt, Hugo. 1890. Kreolische Studien IX. Über das Malaioportugiesische von Batavia und Tugu. Sitzungsberichte der Kaiserlichen Akademie der Wissenschaften zu Wien (philosophisch-historische Klasse) 122(9). 1-256. 53 Hugo C. Cardoso Smith, Ian R. 1977. Sri Lanka Creole Portuguese phonology. Ithaca, NY: Cornell University, PhD dissertation. Smith, Ian R. 1979. Convergence in South Asia: A creole example. Lingua 48(2-3). 193-222. Smith, Ian R. 1984. The development of morphosyntax in Sri Lanka Portuguese. In Mark Sebba & Loreto Todd (eds.), York Papers in Linguistics. York: University of York. Smith, Ian R. 2013a. Sri Lanka Portuguese. In Susanne Michaelis, Philippe Maurer, Martin Haspelmath & Magnus Huber (eds.), The Survey of Pidgin and Creole Languages, vol. 2, 111-121. Oxford: Oxford University Press. Smith, Ian R. 2013b. Sri Lanka Portuguese structure dataset. In Susanne Michaelis, Philippe Maurer, Martin Haspelmath & Magnus Huber (eds.), Atlas of Pidgin and Creole Language Structures Online. Leipzig: Max Planck Institute for Evolutionary Anthropology, http://apics-online.info/contributions/41. (15 November, 2018.) Smith, Ian R. 2016. The earliest grammars of Sri Lanka Portuguese. Papia 26(2). 237-281. Tan, Raan-Hann. 2016. Por-Tugu-Ese? The protestant Tugu community of Jakarta, Indonesia. Lisbon: ISCTE-IUL, PhD dissertation. Tavares de Mello. 1998. Folclores Ceilonenses: Colectânea de Textos do Crioulo Português de Ceilão. Lisbon: Comissão Nacional para as Comemorações dos Descobrimentos Portugueses [reprint of: Tavares de Mello. 1908. Dialecto Indo-Portuguez de Ceylão: Santa Via Cruz. O Oriente Portuguez V. 155-170; Tavares de Mello. 1910-1913. Dialecto Indo-Portuguez de Ceylão: Diálogos. O Oriente Portuguez VII. 363-374; VIII. 218-230; IX. 20-29; X. 16-25, 275-282]. 54