Two Indo-Portuguese Creoles in Contrast

J. Clancy Clements and Andrew J. Koontz-Garboden

(To appear in Journal of Pidgin and Creole Languages)

Abstract: This paper presents a comparative study of two Indo-Portuguese creoles, Korlai Creole Portuguese (KP) and Daman Creole Portuguese (DP). Using recently collected data, the phonology, pronominal systems, TMA markers, syntactic properties, and lexical items of KP and DP are compared and contrasted. The question of the common vs. independent origin of KP and DP is also discussed.

Keywords: creole, creolization, pidgin, Indo-Portuguese, Korlai, Daman

Correspondence Addresses:

J. Clancy Clements

Dept. of Spanish and Portuguese/ BH 844

Indiana University

1020 E. Kirkwood Ave.

Bloomington, IN 47405

email: clements@indiana.edu

Andrew J. Koontz-Garboden

Department of Linguistics

Margaret Jacks Hall, Building 460

Stanford University

Stanford, CA 94305-2150

email: andrewkg@stanford.edu

Introduction

A significant collection of work is beginning to develop on the Portuguese-based creoles in West Africa (d’Andrade and Kihm 1992, Kihm 1994, Honório do Couto 1994, Maurer 1995, Lorenzino 1998, to mention some of the most recent publications). For the Asian-Portuguese creoles, we find an increasing body of work as well, as evidenced by Theban (1975, 1977, 1985), Smith (1977, 1979, 1984), Baxter (1988, 1990a, 1990b, 1994, 1996), Charpentier (1992), Clements (1990, 1991, 1992a, 1992b, 1993a, 1993b, 1996, in press-a, in press-b), Thurgood and Thurgood (1996), and Tomás (1992). With the exception of the work just mentioned, and a dissertation on Daman Portuguese in 1960 (Santos Lopez 1959-60), the studies of Indo-Portuguese creoles date back to the turn of the century (e.g. Coelho 1880-1886, Dalgado 1998a, 1998b and Schuchardt 1889). Comparative studies, such as Hancock (1975) and Ferraz (1987), are based on the data from these scholars. To date, no comprehensive comparisons have been attempted using recently collected data from both newly discovered (e.g. Korlai) and already known (e.g. Daman) Indo-Portuguese creoles.¹ The present study seeks to bring new data to bear on the comparison between the creoles of Korlai (KP) and Daman (DP). After contrasting and comparing the grammars of these two creoles, we discuss to what extent the results of the comparison offer any insight into the independent formation of KP and DP or whether there may have been any influence of KP in the formation of DP.

Historical Background

Before beginning the comparison between KP and DP, we review in broad strokes some history of the areas in which these creoles emerged. ²

The initial policy of the Portuguese in India was not to establish themselves there permanently, but rather to send annual expeditions to India to bring back Indian commodities (Maharashtra State Gazetteer, p. 75). They reached the southwest coast of the Indian subcontinent, at Calicut, in 1498 (cf. map 1), and by the time they became aware of the Chaul area in 1505 (Fernandes 1926, p. 66), they had decided to make their presence permanent. By 1516, the Portuguese had been given permission by their crown to establish a trading post in Chaul, which was fortified in 1524. Chaul, an important trading center since the 6^th century c.e., became a key control point for Portuguese commerce during the 16^th century.

Across from the Chaul fort, a promontory overlooking the Chaul harbor was at the heart of three major confrontations during the 16^th century (see map 3). After the third dispute in 1594, the Portuguese finally took and fortified the promontory. A village formed at its base, taking on the name of the surrounding valley, Korlai, as its own. The inhabitants of this village were speakers of the Portuguese creole of the area.

While the 17^th century witnessed the decline of Chaul as a major trading center for the Portuguese, the 18^th century would see the Portuguese abandon the Chaul area altogether. In 1740, the Hindu Marathas took the Mumbai (formerly Bombay) area, Chaul, and the Korlai fort on the promontory (see map 3). All the Christians who were able to leave abandoned Chaul for Goa; those that did not have the means to do so remained behind. The inhabitants of Korlai, lower-caste Indian Christian tenant farmers, were the ones who stayed. From 1740 to the present, Chaul and Korlai have been largely isolated from Portuguese influence. The Portuguese presence in Korlai was maintained only by one or two Portuguese-speaking parish priests in Chaul who most likely communicated with the Korlai people in simple Portuguese, as was the case in the creole-speaking communities in the Mumbai area in the early 20^th century (Dalgado 1906, p. 143). The number of Christians in Chaul gradually decreased, and at one point the parish priest shifted to Korlai (Meersman 1972). The last Portuguese-speaking priest abandoned Korlai around in 1964, after it was decided in Vatican II that the Catholic faithful should worship in the vernacular of the place in which they live, which in Korlai’s case was and is Marathi, an Indo-Aryan SOV language. Korlai currently has 775-800 native speakers of KP.

Daman was first reached by the Portuguese in 1523, but it was not until 1529 that they launched an offensive in the area. In Moniz (1923, pp. 14-15), we read that the Portuguese were brutal in their onslaught of Daman, attacking it from land as well as from the sea, burning all ships in the harbor. They burned and destroyed the fort there as well, devastating the surrounding villages and assaulting their inhabitants. After the onslaught, Moniz (ibid.) writes that the Portuguese soldiers pulled back, victorious, and returned to Chaul, which reveals the importance of Chaul as a military center during this time of Portuguese colonial expansion.

Daman, however, was not conquered in a day, nor only by soldiers based in Chaul. Although the Portuguese attacked Daman in 1529, and then again in 1534 with Portuguese forces from Chaul and Ormuz, it was not until 1559 that Daman was brought under the control of the Portuguese, who then built a fort and organized a government with a military presence of 1200 men taken from Goa, Chaul, and Mumbai as well (Moniz 1923, pp 19-24; see map 1). In 1581, Daman was conceded the rights and privileges of a Portuguese city (Moniz 1923, pp. 25). When in 1740 the Marathas took Chaul and Bassein in the Portuguese Northern Province, in exchange they allowed the Portuguese to keep Goa, Diu, and Daman (see maps 1 and 2). This last city was then reduced to a minor administrative presence in the Portuguese empire. After 1740, although the Portuguese political influence in India was minimal, Daman was still a cultural and commercial center during the 18^th and 19^th centuries. In 1961, Daman, along with Diu and Goa, was taken over by the Indian government, and given territorial status. That is, Daman belongs to no state proper, but answers directly to the federal government.

Despite the fact that the Portuguese presence existed officially in Daman only up until 1961, there remains even today an unmistakable Portuguese presence there, apparent in the Catholic religion of the people, many of whom still worship in the Portuguese language, and in Catholic schools that taught Portuguese as a subject until 1993. Portuguese was again reinstated as a subject matter in 2000, and some children currently study it in private classes as well. In Daman, there are currently around 4000 native speakers of DP.

With this overview of the histories of Korlai and Daman in mind, let us return to the early days of the Portuguese colonization of India in the 16^th century in order to consider the question of how the respective creoles probably emerged. When Portugal began its colonization of India, 2,000 to 4,000 Portuguese men sailed to Asia annually, the majority ‘being able-bodied and unmarried young men, bound for Golden Goa and further east, relatively few of whom ever returned to Europe’ (Boxer 1975, p. 67). For the most part, these men were from the lower classes of Portuguese society (Boxer 1963, p. 62). The Portuguese practiced miscegenation from the beginning of the colonization. Indeed, the reason there were few if any Portuguese women involved in the colonization was ‘the fact that so many Portuguese men, including the soldados ‘soldiers’ preferred to live with a harem of slave girls rather than to marry, at any rate in their younger and more virile days’ (Boxer 1975, p. 68). The Portuguese casados (i.e. those who married) also had concubines and often owned slaves as well, both men and women, to carry out farm and other types of work (cf. Boxer 1963, p. 61-62).

From the beginning of the colonization of India, we find evidence of Indo-Portuguese offspring. Fernno da Veiga, the judge in charge of the orphans in Goa, reported in 1516 via letter to the King of Portugal that 70 orphans, fathered by the Portuguese, were living from alms in Goa, and requested money in order to provide for them. This type of situation must have been prevalent at that time, not only in Goa, but also in Chaul, Daman, and other settlements as well, given that an overseer of orphans became an official posting in these Portuguese settlements. As an example, among the privileges given to the city of Daman in 1581 by D. Henrique, King of Portugal, was the appointment of a ‘clerk of the orphans’ (Moniz 1923, pp. 25).

As was their practice elsewhere, the Portuguese in the settlements in India took it upon themselves to convert their slaves and their offspring to Christianity (Boxer 1963, pp. 59-60). The incorporation of Indo-Portuguese offspring as well as native Indians into the Catholic church constituted a new element in the Indian society, a society reasonably characterized as one in which religious barriers were quite rigid and in which the social structure, i.e. the caste system, was strict and hierarchical. By becoming Christians, the lower-caste Indians became doubly isolated: first, they were isolated from their erstwhile same-caste peers by their new religion, and from their Portuguese ‘landlords’ by their lower caste. Boxer (1963, p. 75) notes: ‘the Portuguese at first tried to abolish caste distinctions among their Indian converts, but they soon found that this was impossible and they were forced, however reluctantly, to compromise with this immensely powerful and deep-rooted social and religious system.’

Boxer (1963, p. 62-83) also tells us that the general caste system that developed among the Christians in Portuguese India was as follows:

(1) Castes among Indian Christians

Reinol European-born Portuguese

(no term) Portuguese born in India of pure European parentage (very few)

Castiços people born of a European father and a Eurasian mother

Mestiços people born of Eurasian-Indian or Eurasian-Eurasian parents

Brahmins Corresponding to the Brahmin caste in Hinduism. In Hinduism only Brahmins were allowed into the priesthood, and the Portuguese followed this practice with their converts up until the 19th century.

Char(o)dos Corresponding to the Kshatriyan and Vaishyan castes in Hinduism; these were the warriors and merchants. Few of these were ever ordained priests.

Shudras Corresponding to subgroups of the Vaishyan and Shudra castes in Hinduism; this caste included some types of menial laborers, peasants and artisans.

Corumbins Corresponding to part of the Shudra caste in Hinduism; members of this caste were chiefly landless workers and peasants.

Farazes Corresponding to the pariahs or untouchables in Hinduism; the professions of the Farazes were the most menial jobs such as sweeping, grave-digging, etc.

Following a decree from Alfonso de Albuquerque, the second governor of Portuguese India, the pure-blooded Portuguese attempted to marry higher-caste Hindu women, but were unsuccessful. The Mestiços, and most likely the Castiços as well, tried to marry pure-blooded Portuguese, also with meager success as the Reinols and Brahmins alike tended to despise the Mestiços/Castiços. These, in turn, were not friendly with purely Indian Christians of any caste (cf. Boxer 1963, pp. 62-83). Given the above, all Christians, including the lower-caste Indian Christians, formed an extension of the all-pervasive caste system. In fact, it was largely due to the barriers of religion and caste that Korlai, Daman and their respective creole languages themselves have been maintained up until today.

Regarding a possible scenario for the formation of the Portuguese creoles of Korlai and Daman, we assume that when the Portuguese arrived in Chaul and then later in Daman, they found that the Chaul and Daman populations were mostly Marathi- and Gujarati-speaking Hindus respectively, although both areas were under the rule of Muslim lords. It was the lower-caste Hindus that the Portuguese enslaved and with whom they mixed (cf. Boxer 1963, pp. 59-61). As we saw above, the first Portuguese to arrive in Chaul and Daman were soldiers, most of whom were of socially lower classes (Boxer 1963, p. 62). It is reasonable to assume that these men, having traveled several months in Africa before arriving to India, were at least familiar with some form of pidgin Portuguese and could communicate in at least some mixture of Portuguese foreigner talk and Portuguese pidgin. Moreover, they spoke their own dialect of Portuguese, and possibly some of the lingua franca used around the Mediterranean during medieval times (cf. Clements 1992b, 1993a, 1993b). When communicating with the indigenous people, the Portuguese most likely used any and all means available to them to make themselves understood, as did their interlocutors.

What we have, then, are two cases of a two-language contact situation, Portuguese-Marathi for KP and Portuguese-Gujarati for DP, with Portuguese as the target language. More precisely, the target for the newly enslaved and converted Indians was the language the soldiers spoke—most likely a mixture of pidgin Portuguese, Portuguese foreigner talk, and any other linguistic resource the soldiers might have developed. It is reasonable to assume that because the Portuguese used pidgin and/or foreigner talk, the Indians’ access to the standard colloquial Portuguese was partial at best, but that their access to Portuguese pidgin etc. was good. Thus, faced with a new living situation and presented with pidginized and/or simplified input of a new language, which they were obliged to somehow learn —recall that they did convert to Christianity, which entailed adopting a new language and culture— the Indians would have tried their best to make themselves understood, making guesses about what their Portuguese interlocutors would understand. Guesses promoting intelligibility were arguably considered the ‘right’ guesses. Thus, the grammar of the incipient pidgin in Korlai, as well as in Daman, was initially a direct reflection of the shared ‘right’ guesses made by the shifting recent converts (cf. Thomason and Kaufman 1988, pp. 152-53). Some of these guesses reflect the imprint of the native grammar onto the variety of Portuguese being learned. For example, in 16^th century Portuguese, the verb ter ‘have’ expressed possession, as well as being used as an auxiliary for the perfect tenses. Given that in Marathi and Gujarati the existential verb was also the copula as well as the auxiliary for compound tenses, and also was used to express possession, ter ended up performing all of these functions in KP and all but the copulative functions in DP. Moreover, lexical items like KP and DP pe `foot and leg' (< Ptg. pL ‘foot’) were reanalyzed on the Indo-Aryan model in which there is one word that corresponds to ‘foot’ and ‘leg’. (Cf. Thomason and Kaufman 1988, p. 153 for a comparable situation in Snotomense).

In both Korlai and Daman, creolization occurred when the offspring of the Indian slaves of the Portuguese began learning as a first language the variety of Portuguese spoken by the soldiers and/or the slaves’ parents. It is important to keep in mind that, from very early on, evidence such as the fact that Portuguese soldiers kept harems of slave women suggests that there existed a significant number of Indo-Portuguese (IP) offspring in Korlai and Daman. Assuming this scenario, it seems to us justified to characterize the type of creole formation in Korlai/Chaul as well as in Daman as abrupt creolization (cf. Thomason and Kaufman 1988, ch. 6). In other words, before or during the time pidgin Portuguese established itself among the newly converted Christians, Indo-Portuguese offspring were exposed to and picked up the pidgin, nativizing it, and thereby converted it into a creole. With the formation of a new primary language and the adoption of new religious customs, these new Christians forged a new identity. In this process, it is important to keep in mind that after the lower-caste Hindus converted to Catholicism, not only was their access to the higher caste Portuguese barred, but social interactions with Hindus of their same caste decreased significantly because of religion. It was not that they did not interact with the same-caste Hindus or with Christians of other castes; rather, the range of interaction narrowed. For example, where before they could not marry outside of their caste, after they became Christian, they no longer married Hindus of comparable caste. This is largely the case still in Korlai, though possibly somewhat less so in Daman.

There is, however, the issue of two lectal varieties in Daman. As we will see, the differences, found most notably in the respective pronominal systems, can be accounted for by appealing to the relatively recent influence of the lexifier on the variety spoken in the urban center in Small Daman, and relatively less influence on the variety spoken in Badrapur, a geographically separated community in Big Daman (see map 4).

We have presented a possible scenario for the origin of KP and DP. To recap: We assume that KP and DP emerged abruptly, as described above. In Chaul/Korlai, the creole was probably spoken as early on as 1520. In Daman, by the time Daman was given official status as a Portuguese city in 1581, we assume DP was already being spoken, given that Daman was occupied by the Portuguese around 1559.

It is an open question as to what input KP had in the formation of DP. That is, DP could have formed under substantial KP influence or influence from one or more IP creoles, or it may well have formed largely independently of this input. After a comparative view of key aspects of KP and DP grammar, we will return to this question to see whether there are any unequivocal signs that point one way or another.

Comparison of KP and DP

Although the reasons for the differences between KP and DP grammars may be fairly complex, there is one underlying general observation that can account for some of them: from 1500 up to the present, the Portuguese military, administrative, and cultural presence (including religion), was felt in the Korlai area for around 235 years (between 1505 and 1740), in the last 100 of which this presence decreased markedly. We saw that the first Portuguese arrived in Chaul in 1505. KP formed around 1520, spoken by lower caste Indian Christians in an area that waned in importance as a Portuguese commercial center at the end of the 16^th century (Clements 1996, pp. 7-12). In 1740, the vast majority of Portuguese left the Korlai area. Since that time until today, the Portuguese influence has been minimal (i.e. Portuguese-speaking priests until 1964). In contrast, Daman was a Portuguese colony until 1961, and the Portuguese presence is still apparent today in Daman churches and parochial schools. The creole in Daman arguably formed by 1581 and its speakers were initially lower-caste Indian Christians, as in Korlai. Given that the community subsequently developed more diversity, the Indo-Portuguese offspring became soldiers or part of lower middle and middle class who were traditionally, and are even today, employed princIPally in schools and the government offices. Portuguese continues to be used in church services, church choirs sing in Portuguese, and Portuguese is taught as a second language in Catholic schools in the area (English is the schools’ primary medium). Thus, although few people speak standard Portuguese, its presence, as well as the presence of English which has largely replaced Portuguese, blocks large-scale influence from Gujarati, the regional language. It is not surprising, then, that the general make-up of DP contains relatively more Portuguese lexicon and structure, whereas more Marathi influence is apparent in KP.

In examining the linguistic data, we will compare various areas of the KP and DP grammar to determine to what extent the two grammars are similar or different. Based upon this, we will then take up the question regarding the possible influence of KP on the formation of DP.

Phonology

KP and DP have the same oral vowel inventory, shown in (2a), but differ in their nasal vowel inventory in that in KP the mid nasal vowels are lax, whereas they are tense in DP (cf. 2b).

(2) a. KP and DP : /i, e, g, a, c, ], o, u/³

b. KP: /«, , n, , é/; DP: /«, , n, , é/⁴

The consonantal phonemes of KP and DP are given in (3).

(3) Consonants⁵

KP DP

/p, t, (T), k/ /p, t, k/

/b, d, (D), g/ /b, d, g/

/ts, tš/ /tš/

/dz, dž/ /dž/

/th, (Th), kh/ —

/bh, (dh), (gh)/ —

/f, s, š/ /f, s, š/

— /v, z, ž /

/l/ (lack of velar [l]) /l/ (presence of velar [l])

---- /λ/

/m, n/ /m, n, Z/

/r/ /r, /

/w, y/ /w, y/

/h/ —

There are several points to note here. Firstly, two facts reveal that the variety to which the eventual first speakers of KP and DP had exposure was quite accessible: the vowel inventory of KP and DP are virtually identical, and, with the already noted exception of the two nasal vowels /, / the KP and DP inventories correspond to those of Middle Portuguese (cf. Clements 1996, p. 60). Given that the KP and DP grammars are, as we shall see, far from standard Portuguese, the deduction to be made here is that the initial speakers of KP and DP had ample access to the Portuguese pidgin and/or Portuguese foreigner talk, but not to standard colloquial Portuguese as spoken by the Portuguese colonialists among themselves. This pidgin, as spoken by the Portuguese, would have had the phonological system of Middle Portuguese. Secondly, KP exhibits greater influence from the regional Indo-Aryan language than DP, visible in KP in the lack of the velar [l] allophone and the presence of retroflex and aspirated consonants, some of which are present only in loanwords from Marathi. The evidence of greater Portuguese language influence on DP than on KP is seen by a presence of trilled /r/ ( = //), which does not exist in Gujarati. In KP, // was reanalyzed as an aspirated /rh/ (e.g. KP rhnya from Ptg. arranhar). The lack of the phoneme /Z/ in KP and its occasional presence in DP may point to subsequent Portuguese influence in DP, as /Z/ did not exist to the same extent in earlier Portuguese as it did and does in more recent Portuguese (Williams 1938, p. 23). Recall that very often /n/ in Old Portuguese disappeared after nasalizing, as shown in the example Vulgar Latin manu > Old and Modern Portuguese mno (cf. Huber 1933, p. 138). Thus, in Old Portuguese the nasal property in many (but not all) words formed part of a vowel. For KP, a reflection of this is illustrated in the examples in (4), where nasality is part of the stressed vowel, but which corresponds to /Z/ in DP.

(4) Ptg. earlier [v«yu] vs. later [viZu] ‘wine’

KP w«

DP wiZ or viZ

Because of adstrate influence, certain losses of or changes in phonemes have affected KP more than DP, as shown in the examples in (5-7).

(5) Shift of /λ/ to /l/ in KP vs. maintenance of /λ/ in DP⁶

(8) KP DP

a. kata vs. kanta (< Ptg. cantar)

b. s«k vs. sink (< Ptg. cinco)

c. kopra vs. kompra (< Ptg. comprar)

d. nuk and nuk (< Ptg. nunca)

e. mitir and mitir (< Ptg. mentir)

The presence of this reductive phonological process in DP does not appear to stem from KP influence but from the fact that it is also found in Gujarati, though to a lesser extent than in Marathi. Turner (1915, p. 35) documents the historical loss of nasals in stop+nasal clusters from Sanskrit to Gujarati, writing that “in the group, nasal+stop or sibilant, the nasal is lost and the preceding vowel lengthened and nasalized.” Some examples are given in (9).

(9) Sanskrit > Gujarati

nt >t dantaH > dn:t ‘tooth’

nk>k rankaH > rn:k ‘humble’

The lack in KP of /z/and /ž/ could have two causes. In 16^th century Portuguese, these sounds were the affricates /dz/ and /dž/, though it was just around this period that they began to deaffricativize (Teyssier 1984, pp. 49-51). Moreover, Marathi does not have /z/, so even if /z/ were present in the variety of Portuguese of the Korlai area, the Marathi speakers shifting to that Portuguese variety would most likely have perceived /dz/. In DP, both /z/ and /ž/ are present, which is not found in Gujarati (cf. Masica 1991, p. 100-01).

(10) Lack in KP vs. maintenance in DP of /z/

kadz vs. kaz (< Ptg. cas/z/a ‘house’)

medz vs. mez (< Ptg. mes[z]mo‘same’)

Pronominal paradigms

In general, KP and DP pronominal paradigms are quite similar, but there are key differences, not only between the two creoles, but even between different varieties, or registers, of DP. Unless we say otherwise, the lectal differences in DP correspond to the more basilectal variety common in Big Daman (DP1) and the more acrolectal variety heard in Small Daman (DP2) (see map 4). Having said this, however, both pronominal systems are often found in stories told by the same speaker.

Subject Pronouns. The KP and DP subject pronoun inventories are given in (11). DP distinguishes gender in 3sg (il ‘he, gl ‘she’), which is not the case in KP (el ‘s/he’). Moreover, DP2 does not distinguish between 2sg formal vs. informal, whereas DP1 does make the distinction. This non-distinction in DP2 is found throughout the singular pronominal paradigms. Neither KP nor DP makes the formal vs. informal distinction in the plural.

(11) KP and DP Subject Pronouns

KP DP1 DP2 KP DP1 DP2

1sg yo yo yo 1pl n] n]s n]s

2sg w] ]s ----

2pl udzo usez usez

2sg use use use

3sg el il/gl il/gl¹⁰ 3pl elo ilot ez

For comparison, the Middle and Modern Portuguese subject pronouns are given in (12).

(12) Middle and Modern Portuguese Subject Pronouns

sg. pl.

1 eu nós

2 tu vós

2 vocL vocLs

3 ele/ela eles/elas

Both creoles take the first person subject pronouns from their Portuguese counterpart (yo < eu, n], n]s < nós), KP and DP1 derive their second person singular form from the 2pl form (w], ]s < vós), which was also used as an honorific singular in Old and Middle Portuguese. The singular formal form use in both creoles is derived from vocL. KP udzo is from Portuguese vocLs outros ‘you-pl others’, KP el and DP il (el) from ele, DP gl from ela, KP elo and DP1 ilot (elot) from Portuguese eles outros ‘they others’, while DP2 ez is from Portuguese eles. The point here is that KP and DP share the reanalyzed forms elo and ilot, and that pronouns are generally indicative of a relatively deep relationshIP (cf. Schwegler 1999 for a similar argument, based on Arlotto 1972:188).

Object Pronouns. The KP and DP object pronoun inventories, given in (13), contain no vestige of the Middle/Modern Portuguese clitic pronouns o ‘3S.MASC.ACC’, a ‘3S.FEM.ACC, os 3P.MASC.ACC’, as 3P.FEM.ACC’, lhe ‘3S.DAT’, lhes ‘3P.DAT’.

(13) Object Pronouns

KP DP1 DP2 KP DP1 DP2

1sg pari (<parmi) pa(r)mi a mi 1pl p]n] pcn]s an]s

2sg p]r] p]r]s¹¹ ----

2pl pudzo pusez ausez

2sg puse puse awse

3sg pel pirel/ ayil/ 3pl pelo pil]t ayi(l)z

pirel ayel

Instead, they are derived from the combination of a preposition and a disjunctive pronoun. In KP and the more basilectal DP1, the marker is pV(r) (< Ptg. para ‘for’), whereas in the more acrolectal DP2, it is a (< Ptg. a ‘marker of animate DOs and IOs), whereby /u/ of use, usez becomes a glide [w], and the insertion of the glide [y] is found in the 3sg forms. Finally, gender in the 3sg forms continues to be distinguished in DP. Here, DP1 object pronouns again match up more closely with those of KP than with those found in DP2. Given that the object marker a is seldom found in Dalgado’s (1906) Daman texts, the DP2 pronominal system is best considered a more recent development.

Possessive Determiners/Pronouns. Remarkable within the possessive pronoun paradigm is that DP has two systems in the third person. In one of these, the presence of su ‘his/her’ and suz ‘their’ appear to be vestiges of an older system (cf. KP’s system). The fairly frequently attested su in DP may possibly have been reinforced by the decreolizing influence of Portuguese sua ‘his-, her-, their-SG.FEM’; the DP form suz ‘their’ (not equivalent to Portuguese suas ‘his-, her-, their-PL.FEM’) is infrequently heard and only in the oldest people’s speech.¹² The other system in DP uses the marker -d for the non-first person forms, where in KP we find no -d marker at all. Given that this system is well documented in Dalgado (1906), it must have developed well over 140 years ago. In DP, gender is again distinguished in 3sg. The formal vs. informal distinction in 2sg is maintained in KP and DP1, but again not in DP2. As expected, the formal vs. informal distinction is lost in the plural.

(14) Possessive Determiners/Pronouns:

KP DP1 DP2 KP DP1 DP2

1sg mi mi mi 1pl n] n]s n]s

2sg w] d]s ----

2pl udzo dusez dusez

2sg use duse duse

3sg su dil/del/su dil/del/su 3pl sus dilot/suz dez

Deictic Determiners/Pronouns. Both DP and KP have a two-degree system. KP replaced the first-person Portuguese deictic with one from Marathi (KP ye ‘this’ < Marathi he ‘this.NEUT.SG’), while DP es ‘this’ is most likely a conflation ofMiddle Portuguese este ‘this’, and esse ‘that next to you’.¹³ Both KP and DP possess akc, akcl (< Ptg. aquele ‘that over there’) as the other-person deictic. Middle and Modern Portuguese have a three-degree system (este, esse, aquele), while both Marathi and Gujarati display a two-degree system. The creoles emerged with a two-degree system. Note that DP distinguishes number with tud, though not obligatorily.

(15) Deictic Determiners/Pronouns

KP DP

Sg. Pl.

1st ye es es(tud)

2nd/3rd akc akcl akcl(tud)

Interrogative Pronouns. The set of KP and DP interrogative pronouns displays two bimorphemic forms: ky]r, k]r ‘when’ (< Ptg. que hora ‘what hour’) and kilay, kile ‘how’ (< Ptg. que laia ‘what manner’). Bimorphemic wh-words are common in pidgins and creoles (cf. Muysken and Smith 1990, Clements and Mahboob 2000). Whereas other Portuguese-based African and Asian creoles have reflexes of Portuguese que hora (cf. Clements and Mahboob 2000, p.463), the Portuguese reflexes of qui laia are found only in Portuguese-based creoles in India and further east (e.g. Sri Lankan Portuguese kilaay [Clements and Mahboob 2000, p. 470] and Papia Kristang klai [cf. Baxter 1988, p. 190], and even PhilIPpine Creole Spanish [cf. Forman 1972:109]). Thus, the form kile may well have originated in Korlai or possibly further south on the subcontinent (e.g. Cannanore, Cochin). In fact, this could be considered one piece of evidence in favor of DP deriving from KP. Alternatively, the reflexes of que laia could also have their origin in an Asian-Portuguese pidgin, a proposal advanced by Clements (1999).

(16) Interrogative Pronouns

Portuguese KP DP

quem k k

que ki ki, ki koz

quando k]r, k]r ki, ki ky]r, k]n, kwan (acrol.)¹⁴

onde un un, ond (19^th cent.)

por quL pcri (< pcrki) purki

como kile kilay, kom (acrol.)

quanto knt kwnnt

qual kal k]l

TMA Systems

KP and DP TMA markers exhibit a number of noticeable similarities in both form and function, while diverging in key areas as well. We break these down into well-known categories taken from Holm (1988, p. 148; 1989, p. 267), with some modifications needed to discuss the specific TMA features in KP and DP. We focus here on a comparative overview of the TMA markers and combinations thereof. For a more detailed discussion of TMA markers in KP and DP, see Clements (1996, in press-a) and Clements (in prep.) respectively.

Basic stem and past forms. The basic stem and simple past forms are given in (17).

(17)

Unmarked Forms	Basic Stems
KP	kata ‘sing’, kume ‘eat’, subi ‘go up’, tepu ‘heat up’
DP	kanta ‘sing’, kume ‘eat’, subi ‘go up’, bcblu ‘mutter’

The KP/DP basic stem is derived from the Portuguese infinitive (i.e. cantar, comer, subir), whereby the word-final consonant is deleted and syllable-final stress maintained (e.g. [ka 0ta] < [kan0tar]).¹⁵ In both KP and DP, the bare stem functions as an imperative and subjunctive form, as well as an infinitive, and is the form to which elements are added (particles and/or affixes) to express the different tense and aspect meanings.

The simple past, derived in both creoles from the Middle Portuguese preterite form for both -ar and -er/ir verbs in both creoles, is formed by replacing the final vowel with -o for verbs in -a, or by adding the glide -w to verbs in -e, -i. The simple past for the verbs in -u is unmarked. Thus, the basic stem/past tense form for KP tepu ‘heat up’ can refer either to the present tense or the past tense, depending upon the context.¹⁶

(18)

	Simple Past Forms
KP	kato ‘sang’, kumew ‘ate’, subiw ‘went up’, tepu ‘heated up’
DP	kanto ‘sing’, kumew ‘eat’, subiw ‘went up’, bcblu ‘muttered’

In both KP and DP, the simple past functions in a way quite similar to the preterite in Middle Portuguese. The examples in (19) illustrate the canonical use of the simple past in KP and DP: relating fully completed events that took place at some point in the past.

(19a) Korlai Portuguese

[Pay] halo tud dcpcy nigri abriw port.

father said all then girl opened door

‘The father-in-law gave her everything [the code signs], then the girl opened the door.’

(19b) Daman Portuguese

cntnw, rrey saiw pert dc vcran, rrey ko rraín,

so, king went-out near GEN veranda, king with queen

i paso é milyaf.

and passed a eagle

‘So, the king went out onto the veranda, the king and queen, and an eagle passed by.’

We find some variation in the respective KP and DP verb forms. Some verbs in both creoles have lexicalized the erstwhile past marker já, though KP seems to have more of such forms.¹⁷

(20) KP DP PTG

yafoy, yaho(y) jafoy < já foi ‘s/he already went’

yawe jayo < já veio ‘s/he already came’

yade (ja)dew¹⁸ < já deu ‘s/he already gave’

ya fiko, yahiko, jhiko fiko < (já) ficou ‘s/he (already) became/stayed’

jacho acho < (já) achou ‘s/he (already) found/got’

Items such as ‘come’ and ‘go’ are some of the most frequently used verbs. It is not surprising that these verbs are the ones that lexicalized with já. However, for Portuguese já foy ‘s/he already went’ and já veio ‘s/he already came’ to have lexicalized in KP and DP, the element já had to have been a past marker, possibly already in Portuguese pidgin. Indeed, Naro (1978:326) finds evidence for this in portrayals of pidgin Portuguese found in plays of the 16^th century. Although neither KP or DP currently have já as an obligatory past marker, it is still used variably in DP beyond the functional domain of DP já ‘already’, which also exists. In KP, já is used only as a proverb, as in (21), taken from Clements (1996, p. 118).

(21) Question: Use siyo?

you-FORM had-dinner

‘Have you already had dinner?’

Answer: Ja.

‘Yes, I have.’

(Non-)Anterior. KP and DP mark the non-anterior with tc and te, and the anterior with ti and ting (i.e. [tiõ]) respectively. Of note is the similarity between the different sets of forms.

(22) Anterior and non-anterior marking

	non-anterior	anterior
KP	tc (< Ptg. tem ‘s/he has’)	ti (< Ptg. tinha ‘s/he had’)
DP	te (< Ptg. tem ‘s/he has’)	ting (< Ptg. tinha ‘s/he had’)

As will become clear, the non-anterior markers are not used in all contexts, whereas the anterior markers are.

Progressive. The progressive forms in KP and DP, given in (23), are again similar to one another both in form and function.

(23) Progressive marking

	present	past
KP	(te) -n (< Ptg. -ndo)	ti -n (< Ptg. -ndo)
DP	te -n	ting -n

Distinct in (23) is that the DP non-anterior marker te is obligatorily present, while in KP the corresponding form, te, is only found in the speech of older speakers. Although the forms used for the present progressive are slightly different, they match up in function, as shown in (24), where the KP and DP suffix -n marks the progressive.

(24a) Korlai Portuguese

Teru katan.

Teru singing

‘Teru is singing.’

(24b) Daman Portuguese

Joyce te kantan ag]r.

Joyce PRES sing-PROG now.

‘Joyce is singing right now.’

The difference in form between the KP and DP present progressive is explained through the independent evolution of the two languages. Clements (1990) documents the use of present tense prefixal te in the progressive in the speech of older KP speakers. The study suggests that at a prior stage KP marked the present tense as in DP. However, as a result of a typological shift over the last 80-100 years, KP speakers have gradually stopped using the present tense marker te in favor of using only the progressive suffix -n. Stated another way, KP now marks non-anteriority in the progressive with a zero morpheme. The KP present progressive and the DP present progressive can therefore be considered identical in form and function, since it is clear that KP used to mark non-anteriority prefixally and progressive aspect suffixally, as in DP today.

To mark the progressive in the past, the anteriority markers ti/ting are used. Representative examples are given in (25). Marking in KP and DP is identical here, as well.

(25a) Korlai Portuguese

Teru ti kata-n katig.

Teru PAST sing-PROG song.

‘Teru was singing a/the song.’

(25b) Daman Portuguese

Ting kaín chu, nckcl lugar

PAST falling rain in-that place

‘Rain was falling in that place ...’

Habitual. Both KP and DP rigidly distinguish between progressive and habitual aspect in the present tense.

(26) Habitual marking

	present	past
KP	tc	ti
DP	-n	ting -n; er dc, avidi (aydi)

The present habitual is obligatorily marked in KP by preposing the particle tc to the base verb form. In DP, on the other hand, this distinction is made by a zero morpheme (cf. 27), similar to what we saw for the present progressive in KP in (24a) above.

(27a) Korlai Portuguese

Teru tc kata katig.

Teru PRES-HAB sing song

‘Teru sings a/the song.’

(27b) Daman Portuguese

K]n vi temp dc fri, pcn]s

When come weather of cold O-1P

fczen fri, sintin fri.

do-PROG cold feel-PROG cold.

‘When the cold weather comes, we are cold, we feel cold.’

To express the past habitual, KP has only one form, whereas DP has at least three competing forms. In KP, the marker ti is added to the base form of the verb. In combination with the base form, ti here signals past tense, as well as habitual aspect. An example is shown in (28).

(28) Teru ti kata katig

Teru PAST sing song

'Teru {used to/would} sing a/the song.'

In DP, The form, ting -n, as in ting kantan [lit. PAST singing], marks habituality in the past, but is also used for the past progressive in DP. Thus, this form can be considered as having one general function, that of marking ongoingness in the past, whether it be an event in progress at a specific point in the past or a habitually occurring event in the past. An example of the former usage was given above in (25b). The latter usage is illustrated in (29).

(29) yo pc pega pesh, nnw, ting andan pc por s]

1S O/D/P¹⁹ catch fish No PAST going O/D/P put only

‘Me, catching fish, no. [I] would go to put [them on the rope]

pc por as«, prega prega, ting pregan, i

O/D/P put such stick stick PAST sticking and

only, sticking them [together and hanging them on the rope] and

ting vin i kuzinyan. anoyt aydi vi kaz.

PAST coming and cooking night HAB-PAST come house

[then] I would come [back] and would cook. At night I would then come home.’

In (29), we have an example of a second habitual past form, aydi vi ‘used to/would come’. As we will see below, avidi and its reduced aydi (< Ptg. havia de ‘had to...’) are also used to express counterfactuality. Although not found in the stories collected, our elicited data revealed that the DP past habitual has a third marker, er dc (< Ptg. era de ‘was of’), shown in (30).

(30) Estyan kantor kantan tud duming.

this-year choir singing all Sunday

An pasad er dc kanta s] tud mes.

year past PAST-HAB sing only all month

‘This year the choir sings every Sunday. Last year, they sang once every month.’

Although either one of these three forms can be used in DP to express the past habitual, the er dc form is exclusively for the past habitual, whereas, ti + -n and avidi/aydi have other functions along with the habitual.²⁰ Thus, when either of the latter forms are used, the proper interpretation will be determined contextually.

Completive. The situation with the completive in KP and DP is similar to that found for the progressive. First, the absence of the preposed tense marker tc indicates the present tense in KP, where DP has te, though variably.²¹ Additionally, as with the progressive, the completive aspect marker is a suffix in both KP and DP. The use of the anterior ti/ting in combination with this completive suffix indicates past perfect.

(31) Completive marking

	present	past
KP	-d (< Ptg. -do)	ti -d
DP	(te) -d (< Ptg. -do)	ting -d

Just as KP used to have overt marking with te of the present progressive, this was also the case for the marking of present perfect, just as it still is in DP with te.

The examples given in (32)-(33), illustrate the similarity in function of these two forms in the present and the past.

(32) Present

a. Ku Lwidz difludz hika-d. (KP)

OBJ Lwidz cold become-PPART

‘Lwidz has gotten a cold.’

b. Yo nu sab kt te kargad tud

1S NEG know how-much PRES loaded all

ali dent, nckcl kalnw. (DP)

there inside in-that jar

‘I don’t know how much I’ve put inside that jar.’

(33) Past

a. Ali ki jafoy el, akc piken irmnw su

there when went 3S that little brother GEN

amig ti anda-d sirwis. (KP)

friend PAST go-PPART work

‘When he went there, that little brother's friend had already left for work.’

b. ... ki mi fil tin nasid ko ckcl

COMP my son PAST born with that

orel dc burr. (DP)

ear of ass.

‘...because my son had been born with the ear of an ass.’

As regards the forms and functions of the completive, then, KP and DP are again similar to one another, these forms and functions mirroring those of the progressive. The differences found between KP and DP are the result of the independent evolution of the two languages. The clearest example of this is with the overt marking of present tense. KP rarely overtly marks the present tense, whereas DP marks it almost always. Given that KP used to mark this consistently, however, the forms can be considered historically identical to one another.

Irrealis. In contrast to other areas of TMA marking, the irrealis markers of DP and KP, shown in (34), are quite different from one another.

(34) Irrealis marking

future

immediate future

future/conditional

counterfactual

lc (< Ptg. logo ‘at once’)

ted (< Ptg. tem de ‘s/he/it has to ...’)

lc (< Ptg. logo ‘at once’)

ay (< Ptg. havia (de) ‘had to...’)

ater (< Ptg. ha ter ‘s/he/it must/should have)

a (< Ptg. ha de ‘s/he/it has to ...’)

vay (< Ptg. vai ‘s/he/it goes’)

a (< Ptg. ha de ‘s/he/it has to ...’)

avidi/aydi (< Ptg. havia de ‘had to...’)

Whereas the KP future marker is derived from the MP adverb logo ‘at once’,²² the DP irrealis marker is derived from the MP auxiliary verb ha (de) ‘have to...’.²³ Use of the KP marker is illustrated in (35), with a DP example in (36). Note that in the prodosis, after the conjunction ‘if’, the KP and DP forms differ: in KP we find the simple past, while in DP we have the base form of the verb. These forms are obligatory in each case.

(39) el shi jave, el lc kata katig.

3S if came, 3S FUT sing song

'If s/he {comes/came}, s/he {will/}would} sing (the) song(s).'

(40) s] el vi, el a kanta(s) kantig

If 3SF come 3SF FUT sing-IRR song

‘If she comes, she will sing a song.’

In KP, lc in (39) can be interpreted as future ‘will’ or conditional ‘would’, whereas DP a only has a future reading. However, the main verb in the apodosis has an optional final -s (e.g. kantas [< Ptg. cantasse ‘sing-PAST SUBJUNCTIVE’]), which marks irrealis.

This same irrealis marker is also found in DP counterfactual. In both KP and DP statements of counterfactuality, the past perfect form is found in the prodosis (e.g. ti/ting vid ‘had come’) and both KP and DP share one form in the apodosis, namely KP ay and DP avidi/aydi (cf. [34]). Examples are given in (41)-(42).

(41) el shi ti vid, el {ay/ater} kata katig.

3S if PAST come-PPART 3S HYPO-MODAL sing song

'If s/he had come, s/he would have sung the song(s).'

(42) s] el ting vid, el (avidi/aydi) kanta(s) .

if 3SF PAST come-PPART 3SF HYPO-MODAL sing-IRR

‘If she had come he would have sung the song(s).’

Regarding word order in the prodosis, the conjunction in KP must immediately precede the verb, whereas in DP it can but need not precede the verb immediately, and in fact in (42) it does not precede it immediately.

Concluding remarks on TMA markers. KP and DP arose in relatively homogeneous, two-language contact situations between a pidginized variety of Portuguese on the one hand, and Marathi and Gujarati respectively on the other. KP and DP are considered fort creoles, as opposed to plantation creoles which formed under more heterogeneous circumstances and generally involved more than two languages in their respective contact situations (cf. Bickerton 1981). Perhaps it is because of this that the TMA markers (preposed elements and suffixes) as well as the categories, are somewhat distinct from those of plantation creoles. KP and DP share (non)anterior, simple past (both preposed element and suffix), perfect, and counterfactual markers. They differ solely in the irrealis domain in that they have different markers for the immediate future and future, and DP possesses an irrealis suffix (cf. kanta-s) unknown to KP, and KP uses ater, not found in DP. Thus, the extent of the correspondences in the form and function of the TMA markers of the two creoles is remarkable. It suggests the possibility that the two languages may be drawn from the same blueprint. What is less clear is whether the blueprint for both creoles was the variety of pidgin Portuguese prevalent at that time, or whether DP is an offshoot of KP.

The largest differences between the TMA systems of the two creoles involve irrealis marking. These seem to have evolved independently. After comparing other aspects of these two languages, we will return to the question of why this is the case.

Some Syntactic Properties in KP and DP

Typologically, KP and DP are distinct. While KP has a predominance of postpositions and is currently undergoing a transition from a head-initial to a head-final language (Clements 1990, in press-b, Koontz-Garboden 2001) due to the influence of Marathi, DP is decidedly a VO and prepositional language, displaying only the initial signs of becoming a head-final language, in that its pronouns (but not full NPs) may appear with postpositions, as well as with prepositions (except the postposition junt ‘at, with’), as shown in (44).

(44) a. junt de Victor

with GEN Victor

‘with Victor; at Victor’s place’

b. dil junt

GEN-him with

‘with him; at his place’

c. dent dc akcl

inside GEN that

‘inside that’

The postpositional construction ‘pronoun’-junt in DP has existed at least since around 1880, attested in Dalgado (1903), and we assume the other DP adpositions allowed this construction from as early on. As for KP, the head-initial => head-final transition documented in this creole has been underway for at least the last 80-100 years, i.e. it is a fairly recent development.

Both KP and DP mark grammatical relations via adpositions. While the nature of this marking is similar in both languages, the adpositions themselves are sometimes different, as shown in the comparison in (45).

(45) Adpositions marking grammatical relations

KP DP

Acc./Dat.: pV- (pronominal) pV(r)- (pronominal)

ku/pcr (with NPs) pc(r)/(ko) (with NPs)²⁴

Genitive: -su dc

(Pedru su kadz [lit. Pedru gen house] ‘Pedru’s house’ vs. kaz dc Pedru [lit. house gen Pedru] ‘Pedru’s house’)

Instrum.: ku/-su dc/ko

Cause: rhcpcdc pur kawz dc

Location: dc, nc, bc, hcdgc nc (also dc in younger speakers)

(general)

Source: -su dc

Goal: —, dc, nc, be —, pc(r)

itli ‘until’ ate ‘until’

Comitative: -su kosid junt d(c) NP; dc + pronoun + junt

(Lorens su kosid [lit. L.-su-with] ‘with L.’; su kosid ‘with him/her’)

(junt dc Lorens ‘with-dc L’; dil junt ‘with him’)

The only shared markers are the elements KP pV(r)-, ku and DP pV(r)-, ku that mark object case, the instrumental ku, ko, and the locative nc. Otherwise, the markers are distinct in the two creoles.

A comparison of the constructions found in the comparative construction reveals that DP shares the Portuguese structure, while KP exhibits a structure which is neither Portuguese nor Marathi. In (46), we give a comparison of the structures with adjectives, in (47) examples from both creoles.

(46) a. KP: NP1 more than {all/NP2} ADJ COPULA

b. DP: NP1 COPULA more ADJ than {all/NP2}

(47) a. NP1 is smaller than NP2

KP: Luiz mayz ki Pedru piken te

DP: Pedru te may fin dc ki Mario

b. NP1 is the smallest of all

KP: Luiz mayz ki tud piken te

DP: Pedru e/te may fin dc ki tud

The influence of Portuguese, then, is noticeable in DP given that it does not have a construction even near to that of KP, Marathi, or Gujarati.

With respect to negation in KP and DP, nu(n) precedes both the auxiliary and the main verb in both languages (cf. 48-49), as it is found in Middle Portuguese and in southern Gujarati, but not as in Marathi, in which the negative particle appears after the main verb, but preceding the auxiliary if there is one.

(48) Korlai Portuguese

a. Teru nu te kata

Teru NEG-PRES sing

‘Teru does not sing.’

b. Teru nu kato

Teru NEG sang

‘Teru did not sing.’

(49) Daman Portuguese

a. Mercy nu te kantan

Mercy NEG-PRES singing

‘Mercy is not singing.’

b. Mercy nu kanto

Mercy NEG sang

‘Mercy did not sing.’

The lexicon

Out of the 208 core words in KP and DP (using the word list by Comrie and Smith 1977), 182 are derived from the same Portuguese base words. There are 27 exceptions, which fall into two categories: Those in (50), i.e. those pairs where in DP the word is from Middle Portuguese (or Gujarati) and in KP it is from Marathi (MP flor, KP ful, DP flor,‘flower’), or where DP and KP have taken different words from Middle Portuguese (MP [colloq.] gumitar, KP gumita ‘vomit’; MP lanzar ‘throw’, DP lansa [lit. ‘throw, launch’]); and those in (51), i.e. word pairs common in KP and DP which do not correspond to the Middle Portuguese equivalents, such as MP como, KP kile ‘how’, DP kilay (< MP que laia ‘what manner’).

(50) Divergences in Basic Vocabulary KP and DP

English MP KP DP

a. bark (v.) urrar, uivar, hubya ladra

ladrar

b. big grande gran gran, t]moyZ

c. breathe respirar suskar tuma reshpra

d. fat/grease manteiga manteg/adzeyt gordur/manteig

e. flower flor ful flor

f. fog nevoeiro fumas orvaλ

f’. hit bater, dal bate (old: daλ)

da lhe [lit. give

(it to) him]

g. horn corno shing dnnch

i. split,open partir, abrir abri abri

Among the exceptions, then, the differences, listed in (50), are more numerous than the commonalities, shown in (51).

Finally, KP and DP have one copulative verb in common, te/te ‘is/are’ and ti/ting ‘was/were’, from Portuguese tem ‘s/he/it has’/tinha ‘s/he/it had’, with possible influence from Portuguese (es)ta ‘s/he/it is’. In addition, DP has the copula e/er, from Portuguese é ‘s/he/it is’/era ‘s/he/it was’. Interestingly, the semantic distinction between the two DP copulas is akin, but not identical, to that found in Portuguese between estar ‘be (located or in a resultant state)’ and ser ‘be (inherently)’. Thus, in DP, to express location or typically event-related, transitory, physical, emotional, or mental states, te/tin is used and e/er is impossible (cf. [52]).

(52) a. Joyce te/*e kaz.

‘Joyce is (at) home.’ (location)

b. Joyce te/*e durmid.

‘Joyce is asleep.’ (event-related, transitory state)

c. Prat te/*e kebrad.

‘(The) plate is broken.’ (event-related, transitory state)

For any other copulative construction, either te/ti or e/er can be used, generally interchangeably. From the examples in (53), e in (53a) is preferred by some informants. In (53b) ting is for some speakers not interchangeable with er in the clause el er alt. Thus, there is still work to be done to figure out the exact distribution of the DP copulas.

(53) a. Pay e/te vel.

‘Father is old.’

b. El k]n tin/er now, el er alt.

He when was young he was tall

‘When he was young, he was tall.’

In Dalgado (1903), there is mention of the distinction between e (<Portuguese é), ta (< Portuguese esta) and te (< Portuguese tem). The later two were apparently separate forms then, but have now fused formally into te.

Discussion and Final Remarks

In undertaking this comparison of KP and DP using recent KP data and DP material spanning the period from around 1860 to the present, our primary goal has been to determine in which areas these two sister creoles differ and in which they are alike. However, any judgment on how to account for these similarities and differences will be at best incomplete, as we lack the necessary data to determine this unambiguously. Nevertheless, we offer here an analysis of the results of the comparison, which should be seen as an initial assessment of the relation of KP and DP.

Based on the historical events, it is possible that the Daman area was settled by soldiers who had resided in Chaul and who may have brought to Daman dependents who were speakers of the Chaul-area creole. However, it is also likely that such soldiers could have come alone, or come from the Mumbai area, or even from further south. It is clear from the historical records that soldiers from various settlements in the Indian subcontinent particIPated in the attacks on Daman. Consequently, it is reasonable to assume that there was a mixture of soldiers, from various places and with or without dependents, who established themselves in the area. There may well have also been newly arrived soldiers in the mixture given that during the 16^th century there was a yearly influx of 2,000 to 4,000 men into Asia from Portugal. What these men had in common was, arguably, that they spoke some variety of pidgin Portuguese. This variety may well have had a regional flavor. Speakers of pidgin Portuguese would have most likely outnumbered any Portuguese-based creole speakers among the settlers of Daman. Such speakers would have been dependents of veteran Portuguese soldiers.

From a linguistic perspective, in a comparison of the phonologies of the two creoles, we find a stronger Portuguese influence in a number of DP sounds: the presence in DP of /z, ž/, not found in Marathi, KP, or Gujarati, but present in Middle Portuguese (Teyssier 1984, p.27-28), and increasingly present later as the Portuguese affricates [dz], [dž], [ts], and [tš] fricativized around or after after 1580. Note that both KP and DP have chu [tšu] ‘rain’ (< Middle Portuguese chuva [tšuva] ‘ditto’ vs. Modern Portuguese [šuva]). In DP, however, one finds /kaz/ ‘house’ with a historical fricative (< Middle Portuguese /kaza/ ‘ditto’), where in KP one encounters /kadz/. Neither Marathi, KP, or Gujarati have //, present in DP. Moreover, DP exhibits the reintroduction of nasal consonants in words such as viZ (< Modern Portuguese [viZu] ‘wine’) which in KP and Middle Portuguese is [w«] and [v«yu] respectively. What these data suggest, however, is that Portuguese has had a stronger influence on DP than on KP, but it does not shed direct light on the common vs. independent origins of KP and DP.

In the pronominal system, we find an extremely high degree of similarity between KP and DP. They have the originally bimorphemic kile, kilay (< Portuguese que laia ‘what manner’) in common, which is exclusively an Asian-Portuguese creole trait. One key difference is that DP developed a novel possessive determiner/pronoun system based on the Portuguese genitive marker dc not found in KP.

Within the TMA system, again the similarities are wide-ranging and striking. An important difference, however, is found in the irrealis markers. Although the creoles share the counterfactual marker ay and avidi (< Portuguese havia (de) ‘she/it had to...’), for the future and immediate future, KP has lc and ted compared to DP’s a and vai. One could argue that lc may have been DP’s original future marker, and was replaced by a at a later stage. A counterargument to this is that the future marker a is also found in the various Norteiro Indo-Portuguese creoles spoken in the Mumbai area (Dalgado 1906). For its part, KP shares lc with the IP creoles to the south e.g. Sri Lankan Creole Portuguese (Dalgado 1998b, p. 90) and to the east, e.g. Malacca Creole Portuguese (Baxter 1988, p. 126). Thus, the two future markers a and lc belong to two different creole groups, both of which, however, share the same past and present TMA markers. Thus, there is a case, we think, for KP and DP developing different future markers upon creolization.

As for the development of the different immediate future markers, KP ted and DP vai, Holm (1989, p. 268) notes that the form vai ‘go’ (< Portuguese vai ‘s/he/it goes’) is common to all Portuguese-based creoles. Whereas in KP, vai is only found in the imperative (e.g. vai! ‘go!’) and in a request form (e.g. yo vai? ‘Should I go?’), in DP it is the default form of ‘go’, upon which the present and the future are based (e.g. yo vai kaz ‘I’m going home’, yo a vai kaz ‘I will go home’). Thus, vai in DP lent itself naturally to the formation of the immediate future, aided most likely by an identical pattern in Portuguese. In KP, on the other hand, the present tense of ‘go’ is tana(n) (< tc anda(n)), and the infinitive is anda, another significant difference between KP and DP. We consider the KP immediate future form ted an independent development. Again, we find evidence of Portuguese influence on DP to be a factor.

In the area of syntax, negation is virtually the same, and the differences in word order are a recent development. As for adpositions, we find some differences and some similarities. The latter involve the prefix pV(r)- used for marking object case, the instrumental markers ku, ko, and the locative nc, which are found in all Asian Portuguese creoles. The locative nc is present in all Portuguese-based creoles. Thus, it is highly likely that these elements entered KP and DP through pidgin Portuguese. Regarding the more differentiated or specific adpositions expressing cause, source, ‘toward’, and the comitative, these are quite different: cause KP rhcpcdc vs. DP pur kawz dc, source KP -su vs. dc, ‘until’ KP itli vs. ate. In these cases, one could make the argument of Portuguese influence on DP, but in the case of the comitative, KP -su kosid (e.g. Lorens su kosid ‘with Lorenz’) vs. junt d(c) NP or dc + pronoun + junt (e.g. junt dc Lorens ‘with Lorenz; dil junt ‘with him’), we have here an independent development that has nothing to do with Portuguese, but illustrates the influence of the regional languages on the creoles. In both cases, we find postpositions instead of prepositions. We cannot know whether this difference goes back to the formation of the two creoles, but it is possible. As for the construction, the differences here between KP and DP suggest again Portuguese influence in DP.

In the domain of the lexicon, the creoles share 88% of the core vocabulary (182 of 208 items). The items not shared either have to do with Marathi items adopted into KP (e.g. ful vs. flor ‘flower’), more recent Portuguese words finding their way into DP (e.g. KP irgi ‘get up’ vs. DP lawnta ‘ditto’ where irgi is now obsolete, and lawnta (< Portuguese levantar ‘get up’) is more recent), or simply different words taken from Portuguese (e.g. KP ser vs. oyter ‘hill’), both of which were available in the 16^th century. The only item that hints at a possible influence of KP on DP is parse ‘perceive, seem’, which is an amalgam of Portuguese perceber ‘perceive’ and parecer ‘seem’. Although this could possibly have been from Portuguese pidgin, we were not able to find it in other Indo-Portuguese creoles, such as Sri Lankan Creole Portuguese, or in the Portuguese-based African creoles. Lastly, the copula differences seem to be due to the now-common theme of the closer contact between Portuguese and DP.

In conclusion, in the areas compared and contrasted between KP and DP, one encounters a great deal of similarity, which, we argue, is to be accounted for by positing a variety of pidgin Portuguese common to both. Although both creoles developed in rather homogeneous contact situations, the shifting speakers in each situation had pidgin Portuguese as their target. It is because of this that we have creoles in these communities, and not something closer to colloquial Portuguese. Many of the differences between KP and DP are best understood, we believe, by assuming that DP was more strongly influenced by Portuguese than KP. This assumption is justified given the continuing role Portugal and the Portuguese have played in Daman through the centuries, whereas in Korlai Portuguese influence was already waning in the 17^th century and ceased almost entirely after 1740. Finally, the two domains where major differences are found between KP and DP that is not due to stronger Portuguese influence on DP, i.e. the irrealis and specific spatio-temporal markers, are, we submit, a consequence of creolization, that is, of developments that occurred in each creole independently. The only piece of evidence which suggests a direct influence of KP on DP is the verb parse, which could have traveled to Daman in the mouth of a KP speaker, or could have been incorporated into the pidgin of the area and found its way to Daman in that manner.

ACKNOWLEDGMENTS

A version of this paper was presented at the January 1999 annual meeting of the Society for Pidgin and Creole Linguistics. Many thanks go to those who made helpful comments at the meeting. Thanks also go to Mercy Pereira, Dominic Pereira, Paulo Remedios, Joyce Remedios, Maria Jono Remedios, and Victor Jono da Crus Fernandes for their continued help with the Daman Portuguese data. Thanks also to Theresa Titus Fernandes, Jerome Rosario, Angelin Rosario, and Theresa Rosario for help with the KP data. Thanks to Richa Pauranik for proofreading. All errors are, of course, our own.

References

Arlotto, A. 1972. Introduction to historical linguistics. Lanham, MD: University Press of America.

Baxter, A. N. 1988. A grammar of Kristang. Canberra : Dept. of Linguistics, Research School of Pacific Studies, Australian National University.

Baxter, A. N. 1990a. Some observations on verb serialization in Malacca Creole Portuguese. Boletim de Filologia (Lisboa), 31, 161-184.

Baxter, A. N. 1990b. Notes on the Creole Portuguese of Bidau, East Timor. JPCL, 5(1), 1-38.

Baxter, A. N. 1994. Um importante sincretismo no PortuguLs crioulo de Malaca: a preposiçno multifuncional ku. In Celine Cunha (Ed.), Miscelânia Celso Cunha (pp. 1-32). Rio de Janeiro: Melhoramentos.

Baxter, A. N. 1996. Portugues and Creole Portuguese in the Pacific and western Pacific Rim. In S. A. Wurm, P. Mühlhäusler and D. Tyron (Eds.), Atlas of languages of intercultural communication in the Pacific, Asia, and the Americas (pp. 299-338). Berlin: Mouton de Gruyter.

Bernstein, M. and Nimbkar, J. 1982. Structural patterns, book one. Pune: Branthali and Marathi Abhyas Parishad.

Bickerton, D. 1981. Roots of language. Ann Arbor: Karoma.

Boxer, C. 1963. Race relations in the Portuguese colonial empire. Oxford: Clarendon.

Boxer, C. 1975. Women in Iberian expansion overseas, 1415-1815. New York: Oxford University Press.

Charpentier, J-M. 1992. La survivance du créole portuguais “Makaísta” en ExtrLme-Orient. In E. d’Andrade and A. Kihm (Eds.), Actas do colóquio sobre crioulos de base lexical portuguesa (pp. 81-96). Lisbon: Colibri.

Clements, J. C. 1990. Deletion as an indicator of SVO => SOV shift. Language Variation

and Change, 2, 103-33.

Clements, J. C. 1991. The Indo-Portuguese creoles: Languages in transition. Hispania, 74,

637-46.

Clements, J. C. 1992a. Elements of resistance in language contact-induced change. In

Garry Davis and Gregory Iverson (Eds.), Explanation in Historical Linguistics , (pp. 41-58). Amsterdam: John Benjamins.

Clements, J. C. 1992b. On the origins of pidgin Portuguese. Journal of Pidgin and Creole

Languages, 7, 75-92.

Clements, J. C. 1993a. A contribution of an old creole to the origins of pidgin Portuguese. In Francis Byrne and John Holm (Eds.), Atlantic meets Pacific: A global view of pidginization and creolization (pp. 321-331). Amsterdam: John Benjamins.

Clements, J. C. 1993b. Rejoinder to Naro’s ‘Arguing about Arguin’. Journal of Pidgin and

Creole Languages, 7, 119-24.

Clements, J. C. 1996. The Genesis of a language: The formation and development of Korlai Portuguese. Amsterdam: John Benjamins.

Clements, J. C. 1999. Evidencia para a existLncia dum pidgin portuguLs asiático. In E. d’Andrade, D. Pereira and M.A. Mota (Eds.), Crioulos de base lexical portuguesa (pp. 185-200). Associaçno Portuguesa de Linguística: Braga, PT.

Clements, J. C. 2001. Acrolect evolution and stigmatized speech in Daman Creole Portuguese. Paper delivered at the conference Crioulos de base lexical portuguesa e espanhola, 28-29 June, University of Coimbra, Portugal.

Clements, J. C. In press-a. Korlai Creole Portuguese (N] Ling). In J. Holm and P. Patrick (Eds.), Comparative creole syntax. London: Battleridge Press.

Clements, J. C. In press-b. Word order shift and natural L₂ acquisition in a Portuguese

creole. In C. Wiltshire and J. Camps (Eds.), Romance Syntax, Semantics and L₂ Acquisition, John Benjamins, Amsterdam and Philadelphia.

Clements, J. C. In prep. Daman Creole Portuguese: A grammar and texts.

Clements, J. C. and Mahboob, A. 2000. Wh-words and question formation in pidgin/creole languages. In J. McWhorter (Ed.), Language change and language contact in pidgins and creoles, (pp. 459-97). Amsterdam: John Benjamins.

Coelho, F. A. 1880-86. Os dialectos romanicos ou neolatinos na Africa, Asia, e AmLrica.

Boletim da Sociedade de Geografia de Lisboa. Republished in J. Morais-Barbosa (Ed.), Estudos lingüísticos crioulos. Lisbon: Academia Internacional de Cultura Portuguesa.

Comrie, B. Smith, N. 1977. Lingua descriptive studies: Questionnaire. Lingua, 42, 1-72.

Dalgado, S. R. 1903. “Dialecto indo-portugués de Damno.” Ta-ssi-yang-kuo 3, 35-67, 4: 515-22. (Printed also in an offprint of 36 pages).

Dalgado, S. R. 1906. Dialecto indo-portuguLs do Norte. Revista Lusitana 9, 142-66; 193-228.

Dalgado, S. R. 1998a. Estudos sobre os crioulos indo-portugueses. (Edited by M. I.. TomBs). Lisbon: Commissno Nacional para as Comemoraçtes dos Descobrimentos Portugueses.

Dalgado, S. R. 1998b. Dialecto indo-portuguLs de Ceilno. (Edited by I. Smith). Lisbon: Commissno Nacional para as Comemoraçtes dos Descobrimentos Portugueses.

D’Andrade, E. and Kihm, A. (Eds.). 1992. Actas do colóquio sobre crioulos de base

lexical portuguesa. Lisbon: Colibri.

Fernandes, B. A. 1925-26. Antiquities of Chaul. The Indo-Portuguese Review and the

Record of the Indo-Portuguese Association 8, 65-71.

Ferraz, L.I. 1987. Portuguese creoles in West Africa and Asia. In G. G. Gilbert (Ed.),

Pidgin and creole languages. Essays in memory of John E. Reinecke (pp. 337-60). Honolulu: University of Hawaii Press.

Forman, M. 1972. ZamboagueZo texts with grammatical analysis. A study of PhilIPpine Creole Spanish. Cornell University dissertation. Ithaca, NY.

Hancock, I. 1975. Malacca Creole Portuguese: Asian, African or European? Anthropological Linguistics. 17, 211-36.

Hancock, I. 1997. Review of The genesis of a language: The formation and development of Korlai Portuguese by Clements, J. Clancy. Anthropological Linguistics 39, 484-86.

Holm, J. 1988-1989. Pidgins and Creoles. 2 vols. Cambridge: Cambridge University Press.

Honório do Couto, H. 1994. O crioulo portuguLs da Guiné-Bissau. Hamburg: Helmut Buske.

Huber, J. 1933. Gramática do PortuguLs antigo. Lisbon: Fundaçno Calouste Gulbenkian.

Kihm, A. 1994. Kriyol syntax. Amsterdam: John Benjamins.

Koontz-Garboden, A. 2000. On the origins of the Indo-Portuguese creoles: a phonological

comparison of Batticaloa Creole Portuguese and Korlai Creole Portuguese. In J. Auger and A. Word-Allbritton (Eds.), The CVC of sociolinguistics: Contact, variation, and culture. Indiana Working Papers in Linguistics. vol. 2. Bloomington, IN: Indiana University Linguistics Club.

Koontz-Garboden, A. 2001. A stochastic OT approach to word order in Korlai Portuguese. Paper delivered at the 37^th annual meeting of the Chicago Linguistic Society, Chicago, IL. April 20, 2001.

Lopez, M. M. dos Santos. 1950-60. Notas sobre o crioulo PortuguLs de Damno. Dissertaçno de Licenciatura em Filologia Românica. Facultade de Letras. Universidade de Lisboa, Lisbon, Portugal.

Lorenzino, G. A. 1998. The Angolar Creole Portuguese of Sno Tomé: Its grammar and

sociolinguistic history. Munich - Newcastle: Lincom

Feijó, Madureira. 1824. Orthographia ou arte de escrever e pronunciar com ccerto a lingua portugueza (10^th edition). Impressat Regia.

Maharashtra State Gazeteer. 1915. Bombay.

Maldonado, Caetano. 1721. Regras da Lingua Portuguesa.

Masica, C. P. 1991. The Indo-Aryan languages. Cambridge: Cambridge University Press.

Maurer, P. 1995. L’angolar. un créole afro-portugais parlé B Sno Tomé. Hamburg: Helmut Buske.

Meersman, A. 1972. Annual Reports of the Portuguese franciscans in India 1713-1883.

Lisbon: Centro de Estudos Históricos Ultramarinos.

Moniz, A. F. 1923. Noticias e documentos para a historia de Damno : Antiga provincia do norte. Vol. 1. Bastora : TIP. Rangel.

Muysken, P. and Smith, N. 1990. Question words in pidgin and creole languages. In

P.A. M. Seuren and S. S. Mufwene (Eds.), Issues in creole linguistics (=Linguistics 28.4), 883-903.

Naro, A. 1978. A study on the origins of pidginization. Language 54, 314-47.

Schwegler, A. 1999. Monogenesis revisited: The Spanish perspective. In J. R. Rickford and S. Romaine (Eds.), Creole genesis, attitudes and discourse (pp. 235-262). Amsterdam: John Benjamins.

Schuchardt, H. 1889. Allgemeineres über das Indoportugiesische (Asioportugiesische).

Zeitschrift für romanische Philologie, 13, 476-516.

Smith, I. 1977. Sri Lanka Creole Portuguese phonology. Unpublished Cornell University

Ph.D. Dissertation.

Smith, I. 1979. Convergence in South Asia: a creole example. Lingua 48, 193-222.

Smith, I. 1984. The development of morphosyntax in Sri Lanka Portuguese. In M. Sebba and

L. Todd (Eds.), York Papers in Linguistics (Papers from the York Creole Conference, September 24-27, 1983) 11, 291-301.

Teyssier, P. 1984. História da língua Portuguesa. Lisbon: Sá da Costa.

Theban, L.1975. >From creole syntax to universal semantics. Paper presented at the International Conference on Pidgins and Creoles. Honalulu, Hawaii.

Theban, L. 1977. “Indo-rom>nica. Estructura sint<cticas em contacto”. Revue Roumaine

de Linguistique 22.245-48.

Endnotes

1. Koontz-Garboden (2000) compares the phonologies of the creoles spoken in Korlai and Batticaloa (Sri Lanka), using data from Clements (1996) and Smith (1977).

2. The historical information on Korlai and KP is adapted from Clements 1996; the information on the historical sources for Daman and DP are indicated in the text. We note there that there is much more historical information available on Daman than on Korlai. We have incorporated as much as seemed reasonable for the present study. However, much of it has yet to be collected from the archives in Lisbon and elsewhere, an undertaking planned for future research on DP.

3. The linguistic data on KP is taken from Clements (1996, in press-a), while DP data are taken from Dalgado (1903), dos Santos Lopes (1960), and largely unpublished fieldwork carried out by Clements in 1991, 1995, 1999, and 2000 in Daman.

4. In our corpus, evidence of the existence of /c/ involves the pair my ‘but’ and mny ‘mother, more’, whereby we consider the nasalization on the vowels to be a case of progressive nasalization.

5. The phonemes in bold are those not shared by DP and KP, those in parentheses appear only in Marathi loanwords found in KP. The phoneme represented by the symbol // is a trilled vibrant. The phonemes represented by /T/ and /D/ are retroflex counterparts to /t/ and /d/ respectively.

6. Although in KP [λ] is largely not present (one of a number of exceptions is KP uλa ‘look’ [< Ptg. olhar ‘look’]; cf. Clements 1996, p. 75), its presence is stronger in DP, though there is still evidence of its loss in word-final position, as in DP fil ‘child’ (< Ptg. filho ‘son’, filha ‘daughter’) or wel or vel ‘old’ (< Ptg. velho ‘old’).

7. Thanks to a reviewer who pointed out that KP kulcri is more likely to be derived from Middle Portuguese culher(e) ‘spoon’, with an epenthetic postonic vowel than from culherinha ‘teaspoon’. The crucial evidence here is the place of the tonic syllable in the two words: culhe!r(e) > kulc!ri.

8. The aspirated consonant in KP words such as mhara is thought to be a development from an unattested form marha, where we find word-internal [rh] as a reanalysis of []. For a discussion of the phenomenon, see Clements 1996, p. 75.

9. One finds both [w] and [v] in DP speakers’ speech. The labiodental realization [v] may be indicative of subsequent influence of Portuguese on DP, since the alternation [v] - [w] can be found in one and the same speaker, and is possibly an acrolectal feature. Two reviewers suggest that the presence in KP and DP of [w] would be due to dialectal Portuguese. Latin /w/, as in [wenire] ‘come’, became the labiodental fricative in Portuguese, French, Catalan and Italian. Specifically in Portuguese, although Latin word-initial [w] generally became [v], it occasionally became [b], [g], or [f] as well (Williams 1938, p. 59), but there is no evidence of Latin [w] remaining [w] in Portuguese. However, d’Andrade (p.c.), quoting Maldonado 1721 and Feijó 1824, notes that [β], and [v], have been found in European Portuguese for centuries in central and northern areas of Portugal. This, along with the presence of a distinctive [v] allophone of /w/ in Gujarati (cf. Masica 1991, p. 99-100), may have influenced DP.

10. In DP1, the pronunciation of the masc. 3sg. pronoun vacillates between [el] and [il]. In DP2, however, it is more consistently [il].

11. In DP, the key feature singled out as utterly distinctive of the difference between the basilectal t]rt ‘twisted’ variety and the acrolectal pulit ‘polished’ variety is the use of ]s ‘2SFm’ and p]r]s ‘O2SFm’ instead of use ‘2SFr’ and puse ‘O2SFr’. That is, the use of the familiar forms has become the symbol of basilectal DP. Thus, the expression fala ]s, p]r]s ‘talk 2SFm, O2SFr’ is synonymous with the expression fala t]rt ‘talk twisted’. Cf. Clements (2001) for a more detailed discussion.

12. The form suz does not occur in any of the recorded material collected by Clements, but did occur in an 85 year old woman’s unrecorded speech. She was asked about the form and confirmed that suz and dilot are synonymous. Younger speakers were asked about suz. A 55 year old said it was an antiquated form and never heard. When intially asked about the form [suz], the woman first understood it as [sudž] ‘dirty’. Younger speakers do not recognize the word.

Baxter (1990; and in Holm 1989:287) states that Portuguese possessive determiner phrases such as a nossa casa ‘our house’ (cf. Papia Kristang nos sa kaza [lit. we GEN house] ‘our house’ or the antiquated, polite phrase a senhora sua filha [lit. the madame her daughter] ‘madame’s daughter’ may have influenced the reanalysis of su. Extending this argument, possessive determiner phrases such as o nosso filho [u no!su fi!λu] ‘our son’ may have also had an influence in su’s reanalysis. We note, however, that in neither KP nor DP do we find su coocurring with a possessive determiner. That is, there are no forms such as KP *n] su kadz, but rather only n] kadz ‘our house’.

13. In KP, there are vestiges of Portuguese este in forms like estyan ‘this year’ and ez]r ‘(at) this time’, but these are lexicalized chunks.

14. Among more basilectal DP speakers, the perception of what is “pulit” (i.e. more acrolectal) differs interestingly at times according to the age of the speaker. An example of this involves the equivalents for ‘when’, ky]r and kwan. Many younger speakers (adolescents and young adults), regard ky]r as the “pulit” form, whereas for their parents, adults around 45-55 years old, kwan is clearly the more acrolectal of the two forms. What seems to be happening is the following: ky]r is the basilectal form, which for some time now has been yielding ground to kwan, such that when the younger speakers acquired basilectal DP, the default form they learned for ‘when’ was k]n or kwan. Now ky]r is “the other” form, which they have labeled “pulit”. Their parents, however, learned ky]r as the default form, and k]n or kwan as the acrolectal form.

15. In the case of kata, nasal deletion, found throughout KP before stops, has also taken place. See Clements (1996, p. 77-78) for details.

16. It is interesting that both KP and DP incorporated verbs taken from their respective adstrate languages in the same way: they took a form in -u, although DP has far fewer borrowed verbs than KP. In the case of Gujararti (DP’s adstrate language), both the imperative and infinitival forms of the verb end in -u (Masica 1991, p. 260, 300), so DP speakers had/have a fairly clear model from Gujarati. In the case of Marathi, the infinitive, in -e, was not a clear model given its relative infrequency of occurrence. Because the Marathi informal imperative forms often end in a consonant (e.g. kam kcr! ‘do [the] work!), they also were a less than clear form after which to model a base form in KP. KP speakers could have chosen to incorporate the formal imperative verb form in -a (e.g. kam kcra ‘do [the] work!’). However, this may not have been done because of the presence of KP verbs in -a, or because it was not as frequent. By contrast, the negative imperative form in -u —uniformly used in all registers to express the negative imperative as well as to request instruction (e.g. kam kcru nako(s) [lit. work do NEG.IMP.FAM] ‘don’t do [the] work’, and kam kcru? ‘Shall I do the work?’ (Berntsen and Nimbkar 1982, p. 134-46)— is a more frequently occurring form. It may be for this reason that KP borrowed this form of the Marathi verb. It is worth noting that in both creoles a form in -u is used to derive hypocoristic: KP Theresa => Teru; DP Dustin => Dustu.

17. Other Norteiro creoles also display this type of lexicalization. Cf. Clements (1990, p. 130) for more information.

18. The DP form jadew is preferred by some speakers. We take this to suggest that dew has become more acceptable due to superstrate influence.

19. The gloss O/D/P stands for ‘object, directional, and purpose clause marker’.

20. Further study will have to determine whether there is any lectal differences between the three markers. Clements’ informants said there was not, but in the stories collected, which represent mostly basilectal DP, er dc is not found at all.

21. In the stories, there are some instances of te deletion in the present perfect. As yet it is not clear whether there is a difference between the present perfect with vs. without te. This will be the object of further research.

22. A few KP speakers who had a close relation with the Portuguese-speaking priests exhibit ad as a future marker. The example given below, pointed out by a reviewer, is an excerpt of a story told by Francis Martis, a former sacristan of the Portuguese-speaking priests and whose speech was reported on in Clements (1992b, 1993a).

(i) Aka parmi dize depay yo port ad abri; nnwter yo

that O1S say then 1S door FUT open otherwise 1S

nupad abri port.

NEG-FUT open door

‘Tell me that, then I will open the door; otherwise I will not open the door.’

This speech is not representative of KP as spoken by the vast majority of Korlai speakers. One occasionally hears ad used with the deontic meaning ‘should’, but not as a future marker. Thus, it is not the case that KP and DP once shared or now share this particular feature. Other elements from the priest register in this sentence are: dize (in KP we find exclusively hala and occasionally fala), parmi (cf. pari) and nupad (cf. nupa), the KP negative future form (< Ptg. nno pode ‘s/he/it cannot’).

23. Hancock (1997, p. 485) notes that Schuchardt (1889) recorded lo as a future marker in DP. In the relevant passage, it turns out that Schuchardt is referring unambiguously to the Portuguese Creole of Mahé, and not to DP: ‘I [i.e. Schuchardt] obtained data on the Indo-Portuguese from Mahé in 1884-1885 through Mr. W. Schmolck, pastor of the Chombala Mission one and one-fourth hours from Mahé, and he in turn obtained them from an old Portuguese gentleman, Mr. H. D’Cruz, who at an earlier time was an interpreter in Mahé’ [translation by Clements] (1889, p. 516). The piece of data in question is found in the sentence Eu lo vay com vos [lit. 1S FUT go with you] ‘I will go with you’ (1889, p. 517).

24. In Dalgado (1903), we find one example of dative marking with com:

(i) Antno est hom’ pediu com est velh

then this man asked O this old

Qui dixá ficá anôt su caz

COMP let stay his night

‘Then this man asked this old man that he allow him to stay the night in his house.’

One reviewer points out that the ACC/DAT marker is also found in texts in Santos Lopez (1960:149-150), one example of which is given in (ii).

(ii) baí n’ossptal coé filh bL

go L/D-hospital O son see

‘go to the hospital to see their son’

However, in Dalgado and Lopes Santos, these examples are few. In DP of today, ku is not found as an ACC/DAT marker. Given that it is found in KP, as well as in Papia Kristang and Tugu CP (cf. Baxter 1988, 1992), it must be an old trait of the Lusoasian creoles.