Printer Friendly

The slaves' languages in the Dutch Cape Colony and Afrikaans vir(*).

Abstract

The Afrikaans preposition vir ([is less than] Dutch voor `for') is an optional case marker in animate direct objects and in nominal indirect objects, and it is a preposition in recipient and beneficiary indirect object PPs. This un-Dutch feature of Afrikaans is commonly attibuted to influences from Asian Creole Portuguese, which must be narrowed down to Indo-Portuguese in particular. Since there were also many slaves from Indonesia in the Cape Colony, and since Pasar Malay marks animate DOs and IOs with sama `with', one might expect to find an alternative object marker met `with' in (some variety of) Afrikaans. However, this hypothetical met does not exist. This can be explained if we take into account that the majority of the Indonesian slaves came from the eastern parts of the archipelago, while the scarce remnants of South African Malay demonstrate some Eastern Malay characteristics. Object marking in most Eastern Malay dialects happens to agree with object marking in Afrikaans and Indo-Portuguese.

Introduction

The Afrikaans preposition vir, which derives from Dutch voor `for', has a couple of functions that distinguish it from its Dutch counterpart: besides marking beneficiaries and certain prepositional objects, more or less as in Dutch, vir can also mark recipients and animate direct objects, which is an "un-Dutch" feature of Afrikaans.

This expansion of the use of vir for (animate) direct and indirect objects is usually explained as being due to influences from Asian Creole Portuguese, while the evidence seems to point at Indo-Portuguese in particular. However, more languages were spoken by the slaves of the Dutch Cape Colony and in this paper I would like to defend the thesis that some of these languages may also have been instrumental in shaping the use of Afrikaans vir. I will restrict myself to Eastern Malay as spoken by the slaves from eastern Indonesia and to their hypothetical version of Creole Portuguese, although it cannot be excluded that some of the mother tongues of the slaves from India-Sri Lanka may have played a role as well, as will be suggested in section 2 below.

This paper will start with a discussion of the syntax of Afrikaans vir and its Creole Portuguese origin (section 1). Section 2 will discuss evidence to the effect that South African Creole Portuguese may have been Indo- rather than Malayo-Portuguese. However, since Malay was one of the contact languages among the Cape slaves besides Creole Portuguese and (some variety of) Dutch, and since nonstandard (Pasar) Malay would have triggered a different object marker in Afrikaans (section 3), it becomes necessary to scrutinize the scarce data on South African Malay. As section 4 will demonstrate, phonological and lexical properties of South African Malay point in the direction of Eastern Malay, which is in accordance with the provenance of many of the Indonesian slaves in the Cape Colony. Therefore, section 5 will discuss some syntactic properties of the Eastern Malay dialects. It so happens that quite frequently object marking in these dialects is in accordance with object marking in Indo-Portuguese and Afrikaans.

1. Afrikaans vir(1)

Animate, and especially human, direct objects in Afrikaans can, and sometimes must, be marked with a preposition-like element vir, as can be derived from the following examples:(2)
(1) a. Hulle het (vir) Piet geslaan
 they have (FOR) Pete beaten
 b. Ek het nie vir Piet gesien nie
 I have not FOR Pete seen not


In these examples I have glossed vir with FOR, because vir in some of its other uses (i.e. certain prepositional objects as well as beneficiary objects) corresponds to Dutch voor `for'.

Now note that the direct object marker vir hardly behaves as a true preposition, if at all. If it were a preposition vir-marked direct objects should be PPs and the relative "direct object" PPs vir wie and waarvoor as well as their split variant wat ... voor should be grammatical. However, they are not, as the following examples show:(3)
(2) a. ??die man vir wie ek gesien het
 the man FOR whom I seen have
 b. *die man waarvoor ek gesien het
 the man where-FOR I seen have
 c. *die man wat ek voor gesien het
 the man WHAT I FOR seen have


Only example (2a), which I consider an instance of anglicized Afrikaans, seems to be marginally acceptable. The other variants are completely out. This distributional pattern can only be explained if the DO marker vir is not the head of a prepositional phrase but an NP-internal preposition-like case marker.

Furthermore note that DO relatives can best be expressed by means of war `(lit.) what' and that this element may not be preceded by vir, which can be easily explained if we assume that the relative marker war in (2d) is not a pronoun but a specialized complementizer:
(2) d. die man (*vir) wat ek gesien het
 the man (*FOR) WHAT I seen have


This also explains why in split relativized PPs (as used in the ungrammatical example [2c]) the expected relative pronoun waar `(lit.) where' is excluded (cf. [4c] and [6c] below). Apparently, nondependent relative pronouns preferably delete in order to give way to the specialized complementizer wat.

Things are somewhat different with recipient indirect objects (IOs). First consider the following examples:(4)
(3) a. Hy het (vir) ons 'n boek gegee
 he has (FOR) us a book given
 b. Hy het 'n boek vir ons gegee
 he has a book FOR us given


Since (3b) corresponds to the NP PP pattern of Dutch double-object constructions, it is likely that vir in (3b) is the counterpart of the Dutch preposition aan `to' (which occasionally shows up in Afrikaans too), while vir in (3a) may be a case marker. Yet, vir ons in (3a) could, under certain circumstances, also be a PP, since in Dutch an indirect object PP may also precede the DO, instead of following it. Therefore it cannot be excluded that vir in Afrikaans recipient IOs always signals the presence of a PP, although that seems somewhat unlikely.

Evidence from relative clauses shows that vir in recipient IOs can indeed be a preposition:
(4) a. die man vir wie hy 'n boek gegee het
 the man FOR whom he a book given has
 b. die man waarvoor hy een boek gegee het
 the man where-FOR he a book given has
 c. die man wat hy 'n boek voor gegee het
 the man WHAT he a book FOR given has
 d. ??die man wat hy 'n boek gegee het
 the man WHAT he a book given has


The near-ungrammaticality of (4d), which is problematic on all accounts, does not provide any new insights as to the syntactic status of vir in recipient IOs. Apparently dative wh-pronouns may not delete (although they have to).

As regards beneficiary IOs the facts are simpler: these IOs are always marked with vir, which here corresponds to Dutch voor `for'. Compare:(5)
(5) a. Hy het vir ons 'n boek gekoop
 he has FOR us a book bought
 b. Hy het 'n boek vir ons gekoop
 he has a book FOR us bought


Therefore, most probably beneficiary IOs are PPs. This is confirmed by the following relativization facts:
(6) a. die man vir wie hy 'n boek gekoop het
 the man FOR whom he a book bought has
 b. die man waarvoor hy 'n boek gekoop het.
 c. die man wat hy 'n boek voor gekoop het
 d. *die man wat hy 'n boek gekoop het


In conclusion: (1) the DO marker vir is not a preposition; (2) the recipient marker vir usually (and maybe always) is a preposition; (3) the beneficiary marker vir is a preposition; (4) despite functional and syntactic differences there is no variation in phonological shape, quite unlike the situation in Dutch, which lacks the DO marker and which distinguishes aan `to' and voor `for'.

Now received wisdom has it that the expansion of the functions of the Dutch preposition voor ([is greater than] vir) in Afrikaans derives from influences from (Asian) Creole Portuguese as spoken by the slaves. This hypothesis was first put forwrd by Schuchardt (1885: 470, 1890: 227), then repeated by Hesseling (1899, 1923), and later confirmed by Raidt (1976). However, in spite of the fact that Schuchardt is referring to a grammatical property of Indo-Portuguese, both Hesseling and Raidt call this creole Malayo-Portuguese, although Raidt is using Indo-Portuguese examples. Because of this ambiguity it is necessary to establish what kind of Asian Creole was in use in the Dutch Cape Colony. Note that in the literature we also find the simple name "Creole Portuguese."(6)

2. Creole Portuguese in SA: Malayo- and/or Indo-Portuguese?(7)

When Hesseling (1899, 1923) proposed his Malayo-Portuguese theory for the genesis of Afrikaans he meant by "Malayo-Portuguese" a language that was a mixture of Malay and Creole Portuguese, which is not what Hugo Schuchardt and probably any other linguist would mean when using the name "Malayo-Portuguese" (cf. Holm 1988-1989 for Malayo-Portuguese).

Hesseling's opponent Bosman (1916, 1928 [1923]) correctly pointed out that there is no evidence for Hesseling's hotchpotch language and that the historical record shows that there had been two languages in the Cape Colony (besides Dutch and Khoekhoe [Hottentot]): (Pasar) Malay and Creole Portuguese. He furthermore pointed out that this Creole Portuguese must have been variable, since many slaves hailed from India and Ceylon (Sri Lanka) rather than from Indonesia. Yet, for some unclear reason Bosman stuck to the name "Malayo-Portuguese." This infelicitous decision ultimately explains why Raidt (1976) still attributes the object marker vir to Malayo-Portuguese, although she has to resort to examples from Indo-Portuguese. (See below why.)

The Cape Colonial slave population

In view of what follows it is necessary to briefly dwell upon the composition of the Cape Colonial slave population. This overview will be based upon Shell (1994b).(8)

For the greater part of the eighteenth century the total set of newly imported slaves in the Cape Colony can be divided into five subsets: two major groups from India-Ceylon and Indonesia-Malacca, two minor groups from Mozambique-Natal and Madagascar, and a trickle of slaves from West Africa and Angola, imported through smuggling or shipwrecks or as naval "prizes" (Shell 1994b: 41). It, should be noted, though, that alter about 1750 there is a considerable rise in the import of slaves from Madagascar, while in the fourth quarter of the eighteenth century African and Malagasy slaves constituted the overall majority among the newly imported slaves.

In terms of languages this means that besides nonstandard Malay both Indo-Portuguese and Malayo-Portuguese may have been in use among the Cape Colonial slaves, as Bosman (1916, 1928 [1923]) rightly pointed out.

Now, it is quite difficult to distinguish Indo-Portuguese from Malayo-Portuguese, since these varieties (except for SOV varieties such as Batticaloa Portuguese) are mainly identical in their major syntactic properties and it is quite possible that Malayo-Portuguese started out as a variant of Indo-Portuguese.

Therefore it will not come as a surprise that the scarce data we have on South African Creole Portuguese as collected by Franken (1953) for the most part agrees both with Indo- and with Malayo-Portuguese.

First of all, the TMA markers logo for future and ja for perfect in (7)are reminiscent of Indo-Portuguese as well as of Malayo-Portuguese (examples from Franken 1953):
(7) a. Catsioor bos loge more
 dog you FUT die
 (Pieter van de Cust, 1705; 1953: 47)
 b. Jacob ja More
 Jacob PRF die
 (Februarij van Macassar, 1746; 1953: 58)


Furthermore, South African Creole Portuguese made use of the typically Asian Creole Portuguese verb ti(n) `be':
(8) a. tijn bon, tijn bon Augussirie
 tijn bon, tijn bon, Augussirij
 is good, is good(,) August (?)
 (Kleine Marie van Madagascar, 1707; 1953: 48)
 b. nonti platoe
 not-is money
 `There is no money, I don't have money'
 (Imam Rejaldin Sali (80), ca. 1923; 1953: 122)


Finally, negative imperatives were marked with na/nu misti `not must', as is and was the case in Asian Creole Portuguese:(9)
(9) ne miste dali pro mi!
 Na misti dali pro mi
 not must beat OBM me(!)
 `Don't beat me'
 (Absolon van Maleija, 1765; 1953: 73-74)


Note that this feature of Asian Creole Portuguese has given rise to the use of negative imperatives marked with moenie `must not' in Afrikaans.(10)

As regards object marking, however, Indo-Portuguese and Malayo-Portuguese use different prepositions (Holm 1988-1989). In Indo-Portuguese we find a marker per/par/put `for, to', as in the following examples from Ceylon:
(10) a. elle ja olha per elle (Schuchardt 1890: 227)
 3SG PRF see OBM 3SG
 b. Eu ja da acquel per elle (Raidt 1976: 88)
 1SG PERF give that OBM 3SG


whereas in the Malayo-Portuguese varieties ku(n) `with' can be found, as in the following Tugu Portuguese examples taken from Schuchardt (1890):(11)
(11) a. tjoma kun kusir (Schuchardt 1890: 234)
 call OBM coachman
 b. da akel ondra kun eo (Schuchardt 1890: 233)
 give that honor OBM 1SG


(I am inclined to interpret per and kun in [10a] and [11a] respectively as case markers, on a part with vir in Afrikaans DOs, while the same elements in [10b] and [11b] can be analyzed as true prepositions.)

Now the evidence collected by Franken (1953) seems to indicate that South African Creole Portuguese sided with Indo-Portuguese:
(12) a. ne miste dali pro mi!
 Na misti dali pro mi
 not must beat OBM me
 (Absolon van Maleija, 1765; 1953: 73-74)
 b. Maij foedies foedi bos maaij
 mother fuckers fuck your mother
 (Coridon van Malbaar, 1758; 1953: 65)
 c. "die cawallo di pau [...] war koemie agoewa en bibi
 the horse of wood [...] that eats water and drinks
 erroewa [...]"
 grass [...]
 (Imam Rejaldin Sali (80), ca. 1923; 1953: 122)


Apparently animate DOs could be marked with pro (= per/par/pur), which accords with the use of vir ([is less than] Du. voor `for') in present-day Afrikaans. Furthermore, this element could also be used as the head of a PP, as in the following example:
(12) d. Eu nunqua falo por vosse, mas por vos Cammarade.
 ego nouke falla parbosse mee par vos Cammarada.
 I not speak to you(,) but to your comrades
 (Jan van de Capelle, 1723; 1953: 81-82)


By way of conclusion: South African Creole Portuguese seems to have been Indo-Portuguese, and not Malayo-Portuguese. Now, if it is true that Malayo-Portuguese started out as an offshoot of Indo-Portuguese (see above), that might explain why in none of the varieties of Afrikaans objects can be marked with met `with'. That is to say: the early slaves from Indonesia may have spoken a variety of Indo-Portuguese, which only later underwent the changes that transformed it into what we now call Malayo-Portuguese.

Note in passing that this "dative" marking of animate DOs in Indo-Portuguese may have been supported by grammatical properties of (some of) the native languages of the slaves from India and Ceylon, as is suggested by Holm (1988-1989: 608), and it seems to be true for Singalese, witness Gair (1970: 60-62), although I could not detect similar phenomena in Tamil or Malayalam. However, note that my unsystematic findings are representative of a typological pattern, since "dative" marking of animate DOs is a typical feature of the Indo-Aryan languages, which is lacking in the Dravidian languages, as K. V. Subbarao (personal communication) informs me. (Also see Masica [1993: 365-368] on the "dative-accusative" in the Indo-Aryan languages.) Furthermore note that animate direct objects in Sri Lankan Creole Malay can also be dative-marked.(12)

Yet this cannot be the whole story, as will be shown in the next section.

3. The origins of the object markers per/par/pur and ku(n) and the other languages at the Cape

As soon as the origins of the object markers per/par/put and ku(n) are taken into account, it will become clear that object marking in the Cape Colony must have been more complicated than the previous section may suggest.

There cannot be any problem in so far as the slaves from India-Ceylon are concerned (see the end of the previous section). However, as Schuchardt (1890) and Baxter (1988) point out, the use of the Malayo-Portuguese object marker ku(n) `with' must be related to the use of the object marker sama ([is less than] Mal. same `[1] together, [2] with, etc.') in Pasar Malay (also called Bazaar Malay). Compare the following examples from Schuchardt (1890: 229):
(13) a. memukul sama orang
 beat OBM somebody
 b. adjar sama dija
 teach OBM 3SG
 c. berilah sama dija
 give-EMP to 3SG
 d. kasih makan sama andjing itu
 give food to dog that/the


This usage was also known in South Africa, witness a bad Malay translation (by a Dutchman) of a letter written in Buginese (1760; Franken 1953: 67-69). This letter contains four instances of the object marker sama, among which is the following:
(14) saija minta minta sama loe tagal ...
 1SG ask ask OBM 2SG because ...


The historical record provides us with no further cases of sama in South Africa, which is due to the fact that the sporadic Malay sentences in Franken (1953) do not contain objects of the right (animate) type, while the nineteenth-century Malay memorandum by Jan van Boughies uses zero marking or pada (a nonstandard variant of Mal. kepada `to'), where pada is marking indirect objects.(13)

Since Pasar Malay marks human direct and indirect objects with sama `with', a phenomenon that is also attested for the Cape Colony (at least in the abovementioned translation into some sort of Malay written down by a Dutch interpreter), and since it is likely that nonstandard rather than Standard Malay was spoken among Indonesian slaves at the Cape, we have a problem: how is it possible that Pasar Malay and maybe also (an early form of) Malayo-Portuguese have not equally influenced Afrikaans? And why is object marking with met `with' absent even in the dialect of the Cape Muslims (also known as the Cape Malays)?

Furthermore note that Malay was an important language at the Cape, on a par with Dutch and Creole Portuguese (apart from Khoekhoe in the outer districts). Confirming evidence can be found in the court cases up to the year 1772 collected by Franken (1953), upon which the provisional table, taken from den Besten (1997) is based (Table 1). This table indicates how many slaves (or freedmen) from some area X could speak or understand some language Y. The languages mentioned are Dutch, (Creole) Portuguese, Malay, Malagasy, Buginese, the language of the Coast of Malabar (most probably Malayalam, otherwise Malayalam-Tamil), and the language of the area of the Rio de la Goa in southern Mozambique (probably Tsonga). These figures clearly show that Malay was an important language at the Cape during the greater part of the eighteenth century. (And it still was in the early nineteenth century, according to Davids [1990].)

Table 1. Knowledge of languages among slaves and freedmen at the Cape up to 1772
 Dutch Portuguese Malay Malagasy

Madagascar 12 6 1 10
Mocambique 2 1
Terra do Natal 1
Indonesia-
 without-Bugis 27 65 77
Bugis 8 28 44
Malacca 1 1
Chinese 1 3 21
Malabar Coast 18 30 1
Coromandel C. 4 3
Bengal 11 21 5
NW Coast Ind. 1 2
Ceylon 3 2
Cape 3 1 3
Mardijkers 2

Total 91 163 155 10

 Buginese Malabar Rio de
 la Goa

Madagascar
Mocambique 5
Terra do Natal
Indonesia-
 without-Bugis
Bugis 7
Malacca
Chinese
Malabar Coast 1
Coromandel C. 1
Bengal
NW Coast Ind.
Ceylon
Cape
Mardijkers

Total 7 2 5


One could of course try to construct an argument to the effect that interethnic communication had to be in Creole Portuguese since those from India-Ceylon knew hardly any Malay (as the above figures seem to show), while Creole Portuguese was only secondary for those from Indonesian-Malacca, which should have favored Indo-Portuguese. Yet, this does not explain away the presence of nonstandard Malay, and so maybe Pasar Malay, in early colonial South Africa.

Now note that Pasar Malay belongs to the western varieties of Malay, while most Cape Colonial Indonesian slaves seem to have come from the eastern part of the Indonesian archipelago, as can be deduced from their names of origin (Robert Shell, personal communication). Furthermore it is known that the Eastern Malay dialects constitute a variety distinct from the Western (standard and nonstandard) Malay dialects.

The following section will demonstrate that there are also linguistic reasons for assuming that Cape Malay -- some Western Malay influences notwithstanding -- belonged to the Eastern Malay dialects.

4. South African Malay: some data(14)

There are two phonological properties that give South African Malay a distinctly Eastern Malay flavor: the use of a full vowel, more specifically an [a], instead of a shwa and velarization of syllable-final nasals.(15)

The use of an [a] instead of a shwa can be exemplified with data from the eighteenth and nineteenth centuries. First of all, in the Malay translation of the Buginese letter mentioned above (1760; Franken 1953: 67-69) we find forms such as tagal `because' (= tegal), sadiekiet `a little' (= sedikit), panjakiet `illness' (= penyakit). Second, in the (dictated) transcription of Jan van Boughies's memorandum mentioned above (1836; Davids 1990) we find majoeroe `to order' (= menyuruh) and other verbs with a prefix ma- instead of me-. And more forms with <a>, such as paristir `priest' (= Du. priester) and pargi `to go' (= pergi), popped up in a new transcription made by my late colleague Sie Ing Djiang.

Strangely enough no such [a] forms can be found in the remnants of twentieth-century South African Malay collected by Franken (1953: 117-120, 125, 140), but that may be due to the fact that [a] is a substitute for shwa in nonstandard Afrikaans, so that Franken may have heard unstressed [a] while jotting down <e> -- although I have to admit that this is mere speculation.

As for velarization of syllable-final nasals, in the abovementioned Malay translation of a Buginese letter we find boelang `month' (= bulan) and kasiang `pity' (= kasian), and elsewhere there is the case of Lavang `defend yourself' (= lawan; Tallone van Boegies 1751; Franken 1953: 60). Furthermore, in the dictated transcription of Jan van Boughies's memorandum we find da ang `with' (= dengan), whereas the Arabic-Malay original seems to have dangan (according to Sie's transcription). As for the twentieth century, here we find cases like toewang `sir' (= tuan; Franken 1953: 117-118), djallang `path, go' (= jalan; Franken 1953: 119), and malang `evening' (= malam; Franken 1953:118).

By way of a note I mention that there are also functional lexical items in South African Malay that can be related to Eastern Malay. Thus sappe `who' in Sappe itou `who there/that?' (NN 1733; Franken 1953: 53) is reminiscent of Eastern Malay sapa `who' (= siapa), while tra `not' in tramasoe(k) `not enter' (twentieth century; Franken 1953: 140) is also an Eastern Malay negation marker (vs. tidak in Standard Malay).(16)

In view of the above considerations it seems to be justifiable to start consulting grammatical descriptions of Eastern Malay dialects. And in the next section we will see that Eastern Malay dialects often differ from (western) Pasar Malay in that human direct and/or indirect objects are marked with prepositional elements like pa rather than with sama, as has in fact already been noticed for Manado and North Moluccan Malay by Baxter (1988: 170, note 4).

5. The syntax of Eastern Malay dialects with special reference to object marking

Before we start with object marking in Eastern Malay two remarks are in order:

First, in so far as I can see, by redirecting our attention from Pasar Malay to Eastern Malay we do not lose any of the possible influences on Afrikaans that up to now have been attributed to Malay or Pasar Malay (cf. den Besten 1989; Ponelis 1993: 272-273; Raidt 1983: 184-186, 1991 [1971]: 226-227). Reduplication in Eastern Malay seems to be quite similar to what we know about Western Malay, and the Pasar Malay prenominal possessive construction with punya (e.g. Ali punya rumah `Ali POSS house' instead of Standard Malay rumah Ali `house Ali') can also be found in Eastern Malay, although punya usually shows up in a reduced form (pun, pung, pu, pe). These observations are relevant because Malay reduplication is supposed to be the triggering force behind the various reduplication patterns in Afrikaans, while the punya construction is thought to have been one of the forces that reshaped the Dutch pronominal possessive construction (e.g. [Jan.sub.i]/[Marie.sub.j] [z'n.sub.i]/[d'r.sub.j] huis `[John.sub.i]/[Mary.sub.j] [his.sub.i]/[her.sub.j] house') into the Afrikaans se possessive construction (e.g. Jan/Marie se huis `John/Mary POSS house').(17)

Second, while Khoekhoe provides us with a reasonable source for the Afrikaans associative construction (e.g. Jan-hulle `John-3PL'), as is discussed in Nienaber (1994), we finally have an acceptable additional source in Eastern Malay: for example, mama dong `Mother 3PL' in Ambon Malay and S??s Lis dong `Sis Liz 3PL' in Kupang Malay (vs. dong Om `3PL Uncle' in North Moluccan Malay).(18)

This having been said, we can now turn to object marking in the Eastern Malay dialects.

The literature on these dialects confronts us with what seems to be a bewildering variation in object-marking structures, although most of it is tending in the same direction, the exception being the dialect of Kupang (Timor).

This dialect has to be left out of consideration, since it marks its human DOs and recipient IOs with sang (= sama; cf, Steinhauer 1983: 52-53), as in the following examples:
(15) a. Knapa Mia nika sonde undang sang beta?
 why Mia marry without (= not?) invite OBM 1SG
 b.' Dong kirim itu baju sang beta
 3PL send that shirt to 1SG
 b." Dong kirim sang beta itu baju


Although Steinhauer mentions kasi as a substitute for sang in (15b), no mention is made of an object marker pa(da).(19)

In North Moluccan Malay, Manado Malay, and Ambon Malay, however, we find object-marking phenomena that remind us of Indo-Portuguese and Afrikaans.

5.1. North Moluccan Malay

In the North Moluccan Malay texts collected by Taylor (1983) we find several instances of pa marking both human DOs and recipient IOs, although his paper does not mention the phenomenon as such.

First consider the following: examples of DO marking (Taylor 1983: 22-23):
(16) a. Kita pegang pa dia kong dia turus tumbu pa
 1SG grab OBM 3SG then 3SG immediately punch OBM
 kita
 1SG
 b. Dorang pigi di Fayaul, ambel pa itu Pongo pe bini
 3PL go to Fayaul get OBM that Pongo POSS wife
 `They went to Fayal to get Pongo's wife'


(Unlike kita `1PL [inclusive]' in Standard Malay North Moluccan [and Manado] Malay kita is `1SG', `1PL' being kitorang and variants thereof.) It should be noticed that marking human DOs with pa is not obligatory, witness panggel itu ana-ana muda samua ... `call those children adolescent all ...' (1983: 24) vs. panggel pa torang `call OBM 1PU (1983: 27). The following are cases of IOs marked with pa (Taylor 1983: 25-26):
(17) a. Skarang di kampong hampir so kurang orang jual
 now in village almost already few people sell
 kopra pa (A) d??ng (B)
 copra to (A) and (B)
 `There are now only a few people left in the village who
 sell copra to (A) and (B)'
 b. ... dia mulai bikin rekening, kirim pa torang, ...
 ... 3SG start make bill, send to 1PL, ...


Unfortunately, I could not find any beneficiary IO in Taylor's texts. But it goes without saying that the use of a universal marker pa for human DOs and recipient IOs resembles the use of vir in Afrikaans.

Note that according to Voorhoeve (1983:: 6)North Moluccan Malay objects can be marked either with pa (which he derives from (ke)pada; cf. Voorhoeve 1983: 3) or with sama, depending upon the verb, while there is no instance of such an element sama in Taylor's texts. To give an example: where Voorhoeve (1983) has bilang sama `to say to' (p. 6, 9), Taylor (1983) gives bilang pa (p. 25).(20)

This may be due to the existence of different regional dialects or to the intrusion of western (Standard and nonstandard) Malay features into some sociolects of North Moluccan Malay. In fact, Taylor is keen on marking such intrusions in his texts.

5.2. Manado Malay

Now, whatever the solution to the above problem, Manado Malay, an offshoot of North Moluccan Malay, agrees with Taylor's North Moluccan Malay in that both animate DOs and recipient IOs are marked with pa. Consider the following cases of animate DOs marked with pa:(21)
(18) a. ..., kalu neanda' dong bunung pa dia
 ..., if not 3PL kill OBM 3SG (i.e. a dog)
 b. No tu Kapala Pulisi orang Blanda kanaal pa kita
 and that Chief Police person Dutch know OBM 1SG


(Cf. Watuseke and Watuseke-Politton 1981:328 sub [41 and 322 sub [13] respectively.)

The following are cases of recipient IOs marked with pa:
(19) a. Abis mo kase nama apa pa dia?
 and FUT give name which to 3SG?
 b. "Bagini!" kita bilang pa kita pe papa : ...
 this-way 1SG say to 1SG POSS father


(Cf. Watuseke and Watuseke-Politton 1981:328 sub [2] and 330 sub [9] respectively.)

And due to the length of the text in Watuseke and Watuseke-Politton (1981) and sheer statistical luck we also have one case of a beneficiary IO marked with pe, which is a variant of pa and can also be used to mark animate DOs and recipient IOs:(22)
(20) Ta rasa tu Residen neanda' mo beking katjuali pe
 1SG think that Resident NEG FUT make exception for
 satu orang
 one person


(Cf. Watuseke and Watuseke-Politton 1981:332 sub [14].) This use of pa/pe `for' is not an accident however, as Ruben Stoel (personal communication) informs me: beneficiaries in Manado Malay are marked with pa, or with for (< Du. voor, for which also compare for/vur/voor in Karisoh Najoan et al. [1981: 31, 35, 78], a grammar that is otherwise worthless for our purposes).

Therefore, we may conclude that Manado Malay is making use of a general prepositional element pa/pe to mark animate direct objects as well as recipient and beneficiary IOs. (Also compare what is said about IO marking in Ambon Malay below.) So we have found an Eastern Malay dialect that resembles Afrikaans even more closely, in so far as object marking is concerned, than does North Moluccan Malay. Furthermore it should be noted that animate DOs do not have to be marked with pa/pe, which is also in agreement with Afrikaans.(23)

Nevertheless, there are also differences with Afrikaans: Ruben Stoel (personal communication) informs me that pa/pe can also mark inanimate DOs, provided they contain an animate possessor, as in the following example:
(21) Kita da lia (pa) Kiki pe ruma
 1SG REAL see (OBM) Kiki POSS house


Such a structure cannot be found in Afrikaans. However, note that this type of DO marking can be seen as a somewhat quirky extension of the use of pa with animate DOs.

Finally, before we turn to the last dialect to be discussed in this paper, one remark is in order: in Manado Malay there are more variants for pa than just pe. Occasionally we also find par or per, which is in accordance with one of the object markers of Ambon Malay (see below). Therefore, it is unlikely that pa derives from Mal. (ke) pada. Pa/pc/par/per, rather, looks like a Portuguese Creole loan. And this is not unlikely in view of the fact that there is supposed to be some (Creole) Portuguese input in the Eastern Malay dialects, as we can read in the literature.(24)

5.3. Ambon Malay

Ambon Malay, as described by van Minde (1997), may be somewhat disappointing from our point of view because there is no object marking for animate DOs. Yet, Ambon Malay evidences a rich inventory of prepositions for marking recipient and beneficiary IOs as well as what van Minde calls locative-goal NPs, most of which I would like to subsume under a somewhat wider definition of recipient IOs.

The prepositions used are par, fur/for, and buat/bot (van Minde 1997: 183-186, 223, 245-247). As for the etymology of these elements, which seem to be interchangeable but for some sociolinguistic factors: par, the most frequent form, seems to be a Creole Portuguese loan; for/fur, which is a more formal, urban, and older form, is a borrowing from Dutch (Du. voor `for'); while buat/bot is a more recent loan from Standard Malay, that is, Indonesian (Mal. buat `for'). Because of the origin of these elements, all three of them will be glossed with `FOR'.(25)

Furthermore note that Ambon Malay, like Dutch and Afrikaans, also has double-NP constructions as in (22a):
(22) a. Kasi dong bir sadiki (1997: 224)
 give 3PL beer little
 b. Kasi bir sadiki par dong (1997: 224)
 give beer little FOR 3PL


In (22b) we find a recipient IO marked with par. For cases of recipient IOs with the other prepositions, consider the examples in (23):
(23) a. Lalu burung Pombo kasi tongka satu for Kes (1997: 184)
 then bird Dove give stick one FOR Monkey
 b. Jang kasi kepeng buat dia (1997: 186)
 don't give money FOR 3SG
 c. Waktu dong makang papa tu tanya for ana yang tua
 when 3PL eat dad that ask FOR child REL old
 kata ... (1997: 246)
 CONJ
 `When they were eating, the father asked the oldest
 child ...'


Finally, for cases of par, for, and buat marking beneficiary IOs, consider the following examples:
(24) a. Ade kumpul jambu par katong, ... (1997: 184)
 2SG collect rose-apple FOR 1PL, ...
 b. Dong su kas tinggal alamat di kas tu for
 3PL PRF CAUSE stay address at closet that FOR
 om Ka (1997: 247)
 uncle K.
 c. ..., tinggal mata kael perak itu buat ... e ... par dong
 ..., stay fishhook silver that FOR ... e ... FOR 3PL
 dua (1997: 185)
 two


So much for the data.(26)

Now, if we look at these data somewhat more closely, two conclusions can be drawn.

First of all, the use of the beneficiary prepositions voor `for' (Dutch [is greater than] for/fur) and buat `for' (Malay [is greater than] buat/bot) (and probably also Creole Portuguese par) to mark recipient IOs, as in (22b) and (23a)-(23c) above, gives such examples a definitely "Afrikaans" flavor -- especially when for/fur is used, which reminds one of Afr. vir `to, for' ([is less than] Du. voor `for').(27)

However, such a conclusion is nothing but an impression. More important is the conclusion that at some point in the past the speakers of Ambon Malay were able to integrate a foreign beneficiary marker into their language while extending its usage to recipient IOs. This has immediate consequences for the diachronic analysis of the replacement of Du. aan `to' by vir ([is less than] Du. voor `for') in Afrikaans.

Finally, I would like to point out that there may be indirect evidence for the use of par as an animate DO marker in older stages of Ambon Malay, although the evidence is slight and consists of only one example (van Minde 1997: 185):
(25) Niri tikang par Kes pung panta-panta, ...
 Bee sting PAR Monkey POSS buttock-buttock, ...


Van Minde analyzes the par phrase in (25) as a locative goal -- although tikang happens to be a solidly transitive verb (van Minde, personal communication). However, this sentence is reminiscent of the use of pa/pe with inanimate DOs containing animate possessors in Manado Malay, and Aone van Engelenhoven (personal communication) informs me that he recognizes such a use of par from his personal experience with Ambon Malay. So, if my analysis is correct, it is not impossible that par was once used to mark animate direct objects too.

6. Conclusions

The evidence we have collected concerning object marking in some Eastern Malay dialects demonstrates clear correspondences with Afrikaans, except for the quirky marking of inanimate DOs with animate possessors in Manado Malay (and possibly also in Ambon Malay).

In Afrikaans the use of the prepositional element vir ranges over the full spectrum of direct and indirect objects: it is used as a (usually) optional case marker in human/animate DOs, as a preposition (and possibly also as an optional case marker) in recipient IOs, and as a preposition in beneficiary IOs. Afr. vir derives from Dutch voor `for' (for Dutch beneficiary IOs).

In Manado Malay there is an object marker pa/pe/par/per that covers the same range of objects as in Afrikaans, if we discount the quirky extension of animate DO marking. Pa/pe/par/per probably derives from Creole Portuguese per/par.

Since Manado Malay is an offshoot of North Moluccan Malay the same range of facts may be expected to exist in the latter dialect, but for the time being we have to content ourselves with the conclusion that pa in North Moluccan Malay may mark human DOs and recipient IOs, which is a proper subset of the application of vir in Afrikaans.

In Ambon Malay human or animate DOs are not prepositionally marked (although that may be a modern development), but recipient and beneficiary IOs are marked with the some prepositions, which again corresponds to a proper subset of the applications of vir in Afrikaans. At least two of these prepositions, for/fur and buat/bot, derive from languages in which they are used as beneficiary markers.

Therefore, it is probable that the syntax of Afrikaans vir ([is less than] Du. voor `for') has come about through various linguistic influences upon Dutch as spoken in the Cape Colony: on the one hand there was Indo-Portuguese (in cooperation with some of the native languages of the slaves from India-Sri Lanka), on the other hand there were the Eastern Malay dialects as spoken by the majority of the slaves from Indonesia.(28) However, since the Eastern Malay object marker pa(r)/pe(r) seems to be a loan from Creole Portuguese, it is not improbable to assume that the type of Creole Portuguese spoken by these Indonesian slaves must have been a variety of Indo-Portuguese, which is possible since Indonesian Creole Portuguese is supposed to be an offshoot of Indo-Portuguese anyway. In western Indonesia and Malacca, the area of Pasar Malay with its object marker sama, Indo-Portuguese developed into a new distinct variety marking objects with ku(n), which we are used to call Malayo-Portuguese. Since only a minority of the Indonesian slaves came from western Indonesia-Malacca Pasar Malay and (incipient) Malayo-Portuguese were not strong enough to influence Cape Colonial Dutch.

University of Amsterdam

Received 11 November 1999

Revised version received 6 July 2000

Notes

(*) Correspondence address: Instituut voor Algemene Taalwetenschap, Universiteit van Amsterdam, Spuistraat 210, 1012 VT Amsterdam, The Netherlands. E-mail: h.den.besten@hum.uva.nl.

(1.) For vir, the conditions under which it is optional or obligatory and the variables that govern its usage as a direct object marker, such as nominal category, animacy, position of the direct object, emotivity, and style, see Ponelis (1993: 265-272), Raidt (1976), and Donaldson (1993:341-346, 387-388).

(2.) Compare the corresponding Dutch examples:
 (i) a. Ze hebben (*voor) Piet geslagen
 b. ?Ik heb niet (*voor) Piet gezien


(3.) Cf. den Besten (1981), upon which most of the argumentation in section 1 is based. For the data, also compare Donaldson (1993: 145-148, 341-346, 387-388).

Similar diagnostic tests concerning the categorial status of vir, however with a different conclusion, can be found in Ponelis (1993,: 269-271). In so far as I can see, Ponelis is able to argue for a prepositional status for the direct object marker vir by not discussing examples like (2b) vs. (4b) and (6b), while not attributing any value to examples like (2c), which he does mention.

(4.) Compare the corresponding Dutch examples:
 (i) a. Hij heeft (aan/*voor) ons een boek gegeven
 b. Hij heeft een boek aan/*voor ons gegeven


(5.) Compare the corresponding Dutch examples:
 (i) a. Hij heeft voor ons een boek gekocht
 b. Hij heeft een boek voor ons gekocht


(6.) Also see den Besten (1989, 1997), Holm (1988-1989), Ponelis (1993), Raidt (1983, 1991 [1971]).

(7.) This section mainly derives from den Besten (1997).

(8.) Also see Shell (1994a) and the relevant papers in Elphick and Giliomee (1989).

(9.) For examples of na/nu misti in Ceylon Portuguese and Malayo-Portuguese see Jackson (1990: 144, 161,201) and Schuchardt (1890: 75, 76) respectively.

(10.) Cf. den Besten (1989, 1997) and Ponelis. (1993: 460).

(11.) Also compare Baxter (1988: 149-162) for ku in Papia Kristang. Note that there are some stray examples of per, etc., in Schuchardt's Malayo-Portuguese material (Schuchardt 1890).

It is due to the near-absence of per vs. the overall presence of ku(n) in the Malayo-Portuguese data that Raidt (1976) has to resort to Indo-Portuguese examples, while still clinging to the idea of "Malayo-Portuguese" influences -- although "Indo-Portuguese" or "Older Asian Creole Portuguese" would have been a better term in my view.

(12.) Cf. Hussainmiya (1987: chapter 6) and Adelaar (1991). Hussainmiya (1987:163) gives the following examples of case marking in Sri Lankan Creole Malay:
 (i) a. Dey Amat-na pukul
 3SG Amat-OBM beat
 b. Dey Amat-na kasi
 3SG Amat-OBM give


Although Hussainmiya defines -na in (ia) as an accusative marker and-na in (ib) as a dative marker, it seems more likely that -na in (ia) is a dative marker affixed to a human DO. Note that in the early twentieth century speakers of this Creole Malay with SOV word order, when attempting to write Standard Malay, produced sentences like the following (Hussainmiya 1987:155):
 (ii) a. Jangan benci kepada orang
 don't hate OBM people
 b. Anjing itu sudah gigit pada dia orang
 dog that PRF bite OBM3 people/PL
 `That/the dog bit them'


In Standard Malay kepada `to, toward' is used as the head of a recipient indirect object PP. It consists of a directional particle ke and a locative preposition pada and should not be reduced to pada -- at least not in Standard Malay. In (ii) the dative marker (ke)pada clearly serves as a marker of human DOs.

(13.) Jan van Boughies's memorandum in Arabic script (1836), published in Davids (1990) together with a nineteenth-century transcription in Roman script as well as a translation into English, is fraught with difficulties. First of all, Jan van Boughies was not a native speaker of Malay, and a transcription of the Arabic-Malay original made by my late colleague Sie Ing Djiang shows that the nineteenth-century transcription was probably dictated to a Dutch- or Afrikaans-speaking secretary by somebody who could read Arabic script and who made some changes in the text. Finally, the (nineteenth century?) English translation does not completely accord with the original transcription.

(14.) I owe my interest in Eastern Malay to Sander Adelaar, who -- after a talk on Khoekhoe and Afrikaans syntax in Leiden -- questioned me about South African Malay, and who told me that the phonological properties I could mention are Eastern Malay. In this context I would like to point out that there is also work in progress on South African Malay by Cor de Ruyter (George, South Africa), which I hope will appear in a not-too-distant future.

(15.) For these Eastern Malay deviations from the Standard Malay norm, see Voorhoeve (1983), Taylor (1983), Steinhauer (1983), Adelaar (1991), and van Minde (1997), as well as Holm (1988-1989: 582-583). In so far as Manado Malay is concerned, only the use of [a] instead of shwa is discussed in the introduction of Watuseke and Watuseke-Politton (1981: 327), but several cases of velarized nasals can be found in the Manado Malay text published by them, such as tamaang `friend' (= teman; 1981: 328, sub [3]) and ajang `chicken' (= ayam; 1981: 328, sub [4]).

The same phenomena can be observed for Melaju Sini (lit.) `Malay [of] Here', a recent Malay variety with a strong Eastern Malay input, spoken by younger generations of (South) Moluccan expatriates living in the Netherlands, although it should be noted that the relevant sounds are in variation with the corresponding Standard Malay sounds, partly according to phonological position. Cf. Tahitu (1989).

(16.) For sapa `who' see Voorhoeve (1983: 7), Steinhauer (1983: 49), and van Minde (1997: 72).

For tra `not' compare tara `not' in North Moluccan Malay (Voorhoeve 1983: 8, 11; Taylor 1983: 20, 23, 25-26), the negative pretix tar- in Kupang Malay (Steinhauer 1983: 46), and tar/tra `not' in Ambon Malay (van Minde 1997: 273,276-277, 281). In the Manado Malay text published by Watuseke and Watuseke-Politton (1981) tra `not' shows up as a minor negation marker, and always in the collocation tra sala `not to be mistaken', apart from one case of (ka) trada `(or) not', which is also North Moluccan Malay. (Cf. Watuseke and Watuseke-Politton 1981: 328, sub [1], 330, sub [5], etc.) Note however that tar and t(a) ra `not', which show up as t?r and t(a) ra in Sri Lankan Creole Malay, are also negators in Pasar Malay (Adelaar 1991: 31).

In Melaju Sini (cf. note 15) both sapa `who' and a negative prefix ter-/tar- can be found, although the latter is restricted to a couple of lexicalized expressions (Tahitu 1989: 11,105-106).

(17.) For reduplication in Eastern Malay see Taylor (1983), Steinhauer (1983), and van Minde (1997).

For Eastern Malay possessive constructions see Collins (1983b) as well as Voorhoeve (1983), Taylor (1983), Steinhauer (1983), Adelaar (1991), and van Minde (1997). The same construction shows up in Manado Malay, e.g. ta pe papa pe horas kantor `1SG POSS father POSS time office' (i.e. `my father's office hours'; Watuseke and Watuseke-Politton 1981; 330, sub [9]). Also compare Hussainmiya (1987) and Adelaar (1991) on the same type of possessive construction in Sri Lankan Malay Creole.

As for (changes in) reduplication and the punya construction in Melaju Sini (cf. note 15), see Tahitu (1989).

(18.) Cf. van Minde (1997), Steinhauer (1983), and Taylor (1983) respectively. Also see Holm (1988-1989: 583), who is quoting from Collins's dissertation on Ambon Malay. As usual, Manado Malay agrees with North Moluccan Malay, which it is an offshoot of. There are a couple of cases in section 16 of the story in Watuseke and Watuseke-Politton (1981: 334), e.g. dong Eto' `3PL Eto''.

Cases of X-dong in Melaju Sini (cf. note 15), with X being a 1SG or 2SG expression, can be found in Tahitu (1989: 134, 139, 142).

(19.) Kupang Malay marks beneficiaries with untuk `for' as in Standard Malay.

Since the recipient IO marker kasi is also a verb (`to give'), it cannot be excluded that Steinhauer is actually speaking about a serial construction. An argument to that effect could be derived from the fact that the "IO marker" kasi may be followed by sang. Compare the following examples from Steinhauer (1983: 52):
 (i) a. Dong kirim itu baju kasi (sang) beta
 3PL send that shirt GIVE (OBM) 1SG
 b. Dong kirim kasi (sang) beta itu baju


(20.) In this respect the reference to Voorhoeve (1983) in Baxter (1988: 170, note 4) is somewhat unfortunate.

(21.) There is only one case of sama marking a human direct object, and it shows up exactly when the Dutch Kapala Pulisi (Chief of Police) is quoted (Watuseke and Watuseke-Politton 1981: 332, sub [13]):
 (i) Barangkali nona kanaal sama Residen, Residen Morrison
 di Tomohon
 maybe (you)-Miss know OBM Resident, Resident Morrison
 in Tomohon


(A resident was a chief administrator of a district in the Dutch East Indies.) Now the Malay speech used by this Dutchman has some other western characteristics, such as atoran instead of atorang `instruction', while in the immediately preceding context the narrator uses kanaal pa in her own speech (cf. [18b]). Therefore I take it that sama in (i) is Pasar Malay rather than Manado Malay.

(22.) E.g. ambe pe dia `take OBM 3SG' (1981: 344, sub [201) and bilang pe Eto' `say to Eto'' (1981: 338, sub [291).

Note that pa/pe (r) can also be used as a locative marker.

(23.) E.g. lia orang jang ... `see person who ...' (1981: 328, sub [3]) or bawa tu andjing `bring that dog' (1981: 332, sub [10]).

(24.) In kase persen ... per dorang `give (as a) present ... to 3PL' (1981: 338, sub [29]) we have a clear case of an IO with per. The collocation dengar par kita `hear/listen-to OBM 1SG' (1981: 334, sub [19]) is less certain, because par kita might as well be a prepositional object. Note that the corresponding section of the glossary states "par, pa, pe = aan, bij" (1981: 344). The translation provided is the same as the translation given for pa elsewhere in the glossary: Du. aan `to' (for recipient IOs) and bij `with, at', a locative preposition.

(25.) Van Minde (1997: 223--225) also discusses the use of pung (=punya) in double-object sequences of the type IO pung DO, as in (i):
 (i) Kasi dong (pung) bir sadiki
 give 3PL (PUNG) beer little
 `Give them some beer'


Van Minde correctly notes that such sequences cannot represent possessive punya constructions, but he is at a loss how to analyze them. In my view pung in (i) is a serial verb and may be equated with pung `have, possess' (for which see van Minde 1997: 161). This serial verb may in fact be the filler of the functional head of a small clause containing both objects, but this is mere speculation.

(26.) In Tahitu's study on Melaju Sini (cf. note 15) I only found the beneficiary markers buat/bot and voor (Tahitu 1989: 123, 130, 147-148, 150), so that no comparison can be made with Ambon Malay, Melaju Sini's closest relative.

(27.) Compare Afr. gee 'n stok vir Adoons `gives a stick FOR (= to) Adonis/Baboon' with (23a) and Afr. toe vra die vader vir die oudste kind `then asks the father FOR (= zero) the oldest child' with (23c).

(28.) It cannot be excluded that the native (Austronesian and non-Austronesian) languages of these slaves from eastern Indonesia may have played a role as well. Since something, though not much, is known about grammatical correspondences between the Eastern Malay dialects (contact varieties of Malay, after all) and some of the local languages, this is a topic for future research.

References

Adelaar, K. A. (1991). Some notes on the origin of Sri Lanka Malay. In Papers in Austronesian Linguistics 1, Hein Steinhauer (ed.), 23-37. Pacific Linguistics A-81.

Baxter, Alan N. (1988). A Grammar of Kristang (Malacca Creole Portuguese). Pacific Linguistics B, 95. Canberra: Department of Linguistics, Australian National University.

Bosman, Daniel B. (1916). Afrikaans en Maleis-Portugees. Groningen: Noordhoff.

--(1928 [1923]). Oor die ontstaan van Afrikaans, 2nd ed. Amsterdam: Swets en Zeitlinger.

Collins, James T. (ed.) (1983a). Studies in Malay Dialects, part 2 (special issue). NUSA, Linguistic Studies of Indonesian and Other Languages of Indonesia 17.

--(1983b). Syntactic change in Ambonese Malay: the possessive construction. NUSA 17: Studies in Malay Dialects (special issue), James T. Collins (ed.), 28-41.

Davids, Achmat (1990). Words the Cape slaves made: a socio-historical-linguistic study. South African Journal of Linguistics/Suid-Afrikaanse Tydskrif vir Taalkunde 8, 1-24.

den Besten, Hans (1981). Marking WH-Movement in Afrikaans. In Generative Studies in Creole Languages, Pieter Muysken (ed.), 141-179. Dordrecht: Forts.

--(1989). From Khoekhoe foreignertalk via Hottentot Dutch to Afrikaans: the creation of a novel grammar. In Wheels within Wheels. Papers of the Duisburg Symposium on Pidgin and Creole Languages, Martin Putz and Rene Dirven (eds.), 207-249. Frankfurt am Main: Peter Lang.

--(1997). Kreolportugiesisch in Sudafrika: Malaio- oder Indoportugiesisch? In Afrolusitanistik -- eine vergessene Disziplin in Deutschland? Dokumentation des 2. Bremer Afro-Romania Kolloquiums yom 27.-29. Juni 1996, Ruth Degenhardt, Thomas Stolz, and Hella Ulferts (eds.), 317-351. Bremen: Universitat Bremen.

Donaldson, Bruce (1993), A Grammar of Afrikaans. Berlin: Mouton de Gruyter.

Elphick, Richard; and Giliomee, Hermann (eds.) (1989). The Shaping of South African Society, 1652-1840. Cape Town: Maskew Miller Longman.

Franken. J. L. M. (1953). Taalhistoriese bydraes. Cape Town: Balkema.

Gair, James W. (1970). Colloquial Sinhalese Clause Structures. The Hague and Paris: Mouton.

Hesseling, D. C. (1899). Het Afrikaansch. Bijdrage tot de geschiedenis der Nederlandsche taal in Zuid-Afrika. Leiden: Brill.

--(1923). Het Afrikaans. Bijdrage tot de geschiedenis der Nederlandse taal in Zuid-Africa, 2nd revised and expanded ed. Leiden: Brill.

Holm, John (1988-1989). Pidgins and Creoles, 2 vols. Cambridge: Cambridge University Press.

Hussainmiya, B. A. (1987). Lost Cousins. The Malays of Sri Lanka, Bangi, Selangor: Institut Bahasa, Kesusasteraan dan Kebudayaan Melayu, Universiti Kebangsaan Malaysia.

Jackson, Kenneth D. (1990). Sing Without Shame: Oral Traditions in Indo-Portuguese Creole Verse. With Transcriptions and Analysis of a Nineteenth. Century Manuscript of Ceylon Portuguese. Amsterdam: Benjamins; and Macau: Instituto Cultural de Macau.

Karisoh Najoan, J. A.; Liwoso, M. A.; Djojosuroto, K.; and Kembuan, L. (1981). Morfologi dan Sintaksis Bahasa Melayu Manado. Jakarta: Pusat Pembinaan dan Pengembangan Bahasa, Departemen Pendidikan dan Kebudayaan.

Masica, Colin P. (1993). The Indo-Aryan Languages. Cambridge: Cambridge University Press.

Nienaber, G. S. (1994). Pa-hulle is kreools. In Khoekhoe en Afrikaans in gesprek, collection of papers by G. S. Nienaber, 14-67. Linguistic Society of Southern Africa/ Linguistevereniging van Suider-Afrika. South African Journal of Linguistics/Suid-Afrikaanse Tydskrif vir Taalkunde, Supplement 21.

Ponelis, Fritz (1993). The Development of Afrikaans. Frankfurt am Main: Lang.

Raidt, Edith H. (1976). Die herkoms van objekskonstruksies met vir. In 1875-1975. Studies in die Afrikaanse Taal. Uitgegee in opdrag van die Suid-Afrikaanse Akademie vir Wetenskap en Kuns by geleentheid van die inwyding van die Afrikaanse Taalmonument, 72-101. Johannesburg: Perskor. (Reprinted [1994]. Historiese taalkunde. Studies oor die geskiedenis van Afrikaans, 116-147. Johannesburg: Witwatersrand University Press.)

--(1983). Einfuhrung in Geschichte und Struktur des Afrikaans: Darmstadt: Wissenschaftliche Buchgesellschaft.

--(1991 [1971]). Afrikaans en sy Europese Verlede, 3rd revised and expanded edition. Cape Town: Nasou.

Schuchardt, Hugo (1885). Review of N. Mansvelt, Proeve van een Kaapsch-Hollandsch Idioticon, etc. (Stellenbosch: Schroder; Utrecht: van Huffel, 1884). Literaturblatt fur germanische und romanische Philologie 6, 464-470.

--(1890). Kreolische Studien IX: Das Malaioportugiesische yon Batavia und Tugu. Sitzungsberichte der Philosophisch-historischen Classe der Kaiserlichen Akademie der Wissenschaften 122. Vienna: Tempsky.

Shell, Robert C.-H. (1994a). The Tower of Babel: the slave trade and creolization at the Cape, 1652-1834. In Slavery in South Africa. Captive Labor on the Dutch Frontier, Elizabeth A. Eldredge and Fred Morton (eds.), 11-39. Boulder: Westview; and Pietermaritzburg: University of Natal Press.

--(1994b). Children of Bondage: A Social History of the Slave Society at the Cape of Good Hope, 1652-1838. Johannesburg: Witwatersrand University Press.

Steinhauer, Hein (1983). Notes on the Malay of Kupang (Timor). NUSA 17: Studies in Malay Dialects (special issue), James T. Collins (ed.), 42-64.

Tahitu, Bert (1989). Melaju sini: het Maleis van Molukse jongeren in Nederland. Unpublished doctoral dissertation, University of Leiden.

Taylor, Paul M. (1983). North Moluccan Malay: notes on a "substandard" dialect of Indonesian. NUSA 17: Studies in Malay Dialects (special issue), James. T. Collins (ed.), 14-27.

van Minde, Don (1997). Malayu Ambong: Phonology, Morphology, Syntax. Leiden: CNWS.

Voorhoeve, C. L. (1983). Some observations on North-Moluccan Malay. NUSA 17: Studies in Malay Dialects (special issue), James T. Collins (ed.), 1-13.

Watuseke, F. S.; and Watuseke-Politton, W. B. (1981). Het Minahasa- of Manado-Maleis. Bijdragen tot de Taal-, Land-en Volkenkunde 137, 324-346.
COPYRIGHT 2000 Walter de Gruyter GmbH & Co. KG
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2000 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:DEN BESTEN, HANS
Publication:Linguistics: an interdisciplinary journal of the language sciences
Geographic Code:6SOUT
Date:Sep 1, 2000
Words:9159
Previous Article:Orthography and ideology: issues in Sranan spelling(1).
Next Article:Semantic transparency in Lowland Ecuadorian Quechua morphosyntax(1).
Topics:

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters