Die Creol taal

250 years of Negerhollands texts

[p. 19]

5 The origin of Negerhollands

After this short presentation of the main peculiarities of Negerhollands, we now return to its origin, history and development. In section I,2 we have already treated the formative period and the context of its emergence. Before we return to that context, the special multilingual setting of the slave society in a multilingual European settlement, we will discuss the influence of the different source languages and briefly consider the various features of Negerhollands in relation to possible scenarios for its emergence as a language.

5.1 West African influence

The extent of West African influence on Negerhollands still needs to be established, but it probably was not as large as on some of the other Caribbean creoles, particularly those of Surinam. We pointed out the presence of serial verbs and predicate cleft constructions in the creole, constructions which are often claimed to be characteristic of West Africa. It also remains hard to estimate the extent to which slaves were induced by the missionaries to ‘de-africanize’ their language (Stein forthcoming a).

5.2 The genesis of Negerhollands and second language acquisition

On the basis of what we know of the acquisition of Dutch as a second language by adults, we can explain a number of properties of Negerhollands and establish a plausible scenario for the genesis of the language.

Given the demographic facts, we can establish that the first language or mother tongue of the majority of the slaves in the probably decisive initial phase was 17th century Twi and possibly Portugese pidgin and Papiamentu. The possible ‘target’ languages were primarily 17th century spoken Zealandic and Hollandic, and also Danish, English, and French. The resulting creole language shows mainly Zealandic and Hollandic lexical and phonological influence.

The features of Negerhollands similar to those of Dutch as a second language are among others:

(a)	word order, in particular the rigid SVO order of Negerhollands and the absence of postpositions;
(b)	the choice and semantic features of the tense/mood/aspect particles and of auxiliary verbs;
(c)	the absence of Dutch er forms in Negerhollands.

[p. 20]

To exemplify the last point: in Negerhollands there is no reflex of the Dutch R-pronouns:

(22)	a.	wagut	ju	wil	du	mit	di
		what	2SG	want	do	with	DET

‘What do you want to do with it?’

cf. Du.: ‘wat wil je ermee doen?’

	b.	am	no	weet	een	gut	fan	di
		he	NEG	know	a	thing	of	DET

‘He does not know a thing of it’

cf. Du.: ‘hij weet er niets van’

One way of explaining this is by assuming that the er-pronouns, being phonologically weak, did not survive in the process of second language acquisition. On the other hand, the phonologically strong forms, the Dutch demonstrative R-pronoun daar and the interrogative R-pronoun waar, which are only used with non-human NPs, have also disappeared. This suggests that factors of morphological and positional markedness⁶ are involved as well. A tentative conclusion could be that the slaves only used those parts of Dutch that they understood, in building up their new language.

6 See also example 5 where the term ‘marked’ is used in the same sense as ‘out of the ordinary’.

5.3 The Zealandic and Hollandic lexifier language

An important issue is the precise characterization of the target language varieties. From the seminal work of Hesseling (1905) it appears that Negerhollands must have had Zealandic Dutch as its greatest target language, although Hesseling takes recourse now and then also to related West Flemish dialects (see van Rossem Forthcoming a).

Just like Zealandic Negerhollands has /i/ for Du. ij, a diphthong (e.g. ki(k) ‘see, look’ <Du. kijk(en) ‘look’). The Middle Dutch /y/ of Zealandic, which never underwent diphthongization, is an unrounded /i/ in 20th-century Negerhollands, so that a Standard Dutch diphthong ui that derives from a Middle Dutch /y/ generally corresponds to a Negerhollands /i/, as in brin (<Du. bruin ‘brown’). Only where Zealandic, just like Middle Dutch, has a diphthong ui (the so-called ui-2), do we find a diphthong (or a monophthongized variant) in Negerhollands: e.g. reil ‘exchange, change’ (<Du. ruil(en) ‘exchange’), loi (<Du. lui ‘lazy’). Note that in 18th-century (and maybe early 19th-century) Negerhollands, Zealandic /y/ was still present as a variant. We may derive this from the graphemes <y> and <ye> in the Danish texts as well as from data in the Danish ABC booklets (Kingo 1770 and Wold 1770). Furthermore, note that Standard Dutch never diphthongized /y/ (or /i/) before /r/, so that a Negerhollands /i/ out of /y/ sometimes corresponds to an /y/ in Standard Dutch. Disregarding some other complicating factors such as the historical phonetics of ui, we could summarize the

[p. 21]

development of the high front vowels from Middle Dutch to Zealandic as in Figure 1 below:

1. The development of the high front vowels from Middle Dutch to Negerhollands. The symbol [y] represents the IPA front round high vowel, and [ʌy] is a diphthong where the [y] is preceded by the centralized back vowel [ʌ], in Dutch orthography <ui>.

The Zealandic target language also explains the West Germanic /u/ in ju ‘you sing.’ and nu ‘now’ instead of the diphthong or /y/ of Standard Dutch jou and nou, nu resp., so that an English etymology for NH ju is unnecessary. Similarly, the West Germanic /u/ instead of regular Zealandic /y/ (or Negerhollands /i/) in words like hus ‘house’ and muši ‘mouse’ does not have to be ascribed to Danish since such irregular variant forms are also attested for Zealandic (cf. van Ginneken 1913).

Finally, the Zealandic target language also explains the occurrence of southern Dutch words such as kot ‘cabin’, hofi ‘garden’ and rigibe:n ‘backbone’ (all examples from de Josselin de Jong) or, in the 18th century, kachel ‘foal’ and schuif ‘drawer’ (cf. van Ginneken (1913), who, however, was wrong about the status of neusdoek ‘handkerchief’). Furthermore, 18th-century keer ‘like’ may derive from Zealandic keuren ‘like, please’ rather than from Papiamentu ke(r), kie(r) ‘want, like, love’ (cf. den Besten 1989).

Still we have to take into account Hollandic or Standard Dutch as an available target, perhaps in a later stage. The relevant words appear to have to do with town and harbour and with education and religion. De Josselin de Jong (1926) gives a number of words with initial /z/ and /v/ where we would have expected /s/ and /f/ due to Negerhollands devoicing rules, as in se: ‘say’ (<Du. zeg(gen)) and for ‘for’ (<Du. voor), namely: ze: ‘sea’ (<Du. zee), zeil, zeilə ‘sail’ (<Du. zeil(en)), zil ‘soul’ (<Du. ziel), valis ‘valise, hand bag’ (<Du. valies), venstu, wenstə ‘window’ (<Du. venster). This is confirmed by data in the word list of Nelson (1936) and by the recorded use of the word zɔndə ‘sin; what a shame’ by Miss Alice, the last known speaker. The 18th-century religious

[p. 22]

texts often have the regularly derived creolized form sondo, with a devoiced s. This may suggest that changes in voicing reflect changes in the target over time.⁷

Initial /z/ and /v/ are lacking, however, in the list of cardinal numerals in de Josselin de Jong (1926) and Nelson (1936). In the latter source there is an unexpected vowel in the words fɛv ‘five’ and fɛvti:n ‘fifteen’. This corresponds with the use of ei in veif, etc. in Magens (1770) and feif in de Josselin de Jong (1926), which is in accordance with the diphthong.⁸

7 Another explanation may be the fact that texts written by the Moravian missionaries reflect the Northern Standard German voiced [z], which at the beginning of the word is written , whilst the character is pronounced [ts]. It is more likely, however, that the missionaries' was ambiguous between [s] and [z] as was still more or less the case in 18th century Dutch spelling. In Danish the [z] does not have a phonemic status. In any event, the occurrence of the in any early Negerhollands text is rare, except in proper names.

8 The word fikopdibʌl ‘five-headed devil’ in De Josselin de Jong's glossary (1926) appears to give a more popular variant of the numeral, but considering tale 49 in De Josselin de Jong the translation ‘fourheaded devil’ is more probable. (Cf. fi ‘four’ in De Josselin de Jong (1926) and fi ‘four’ in Nelson (1936)).

5.4 A Portuguese pidgin?

In many creoles, even ones lexically unrelated to either Spanish or Portuguese, we find lexical items of Portuguese origin. Examples are the well-known pickaninny (<Port. pequeninho ‘very little’) and save (<Port. saber ‘know’) in English creoles. One hypothesis to account for these has been the assumption that there was one Portuguese pidgin, spoken in the slave trade along the coast of West Africa, from which the Caribbean creoles were formed through the partial replacement of Portuguese by English, French, etc. vocabularies, respectively. This replacement would not have been complete, and hence the lexical traces from the Portuguese pidgin. In the proverbs we noticed a number of Ibero-Romance lexical items. This could be interpreted in the light of this Portuguese pidgin hypothesis. It is much more likely, however, that they derive from Papiamentu, as pointed out by Hesseling (1933). Planters who settled on the Virgin Islands brought over some, presumably Papiamentu-speaking, slaves with them from the island of Curaçao, and these may well have influenced the emerging creole.

5.5 The epithetic vowels: substrate or superstrate influence?

A special problem in connection with the Dutch target dialect (the superstrate or main lexifier language) and with possible African (the substrate language) influences is the question of the so-called epithetic vowels in Negerhollands. In some creole languages (e.g. in Surinam) full vowels appear with great regularity at the end of ‘European’ words, and a very regular consonant-vowel-consonant-vowel-pattern emerges (e.g. buku ‘book’). This pattern could have a universal basis in laws of ease of pronunciation, but it is more likely that it is connected to African patterns.

[p. 23]

Negerhollands also seems to have such epithetic vowels, although we must exclude petrified diminutives as in hofi ‘garden’ (<Du. hof-je, hof-ie) and kalfi ‘calf’ (<Du. kalf-je, kalf-ie), and original full vowels, as in apəlsina ‘orange [lit.: apple china]’. What remains are words such as hogo ‘eye’ (<Du. oog), but it is a sporadically occurring phenomenon and mostly limited to nouns and an adjective or two.

Such a full final vowel also occurs where historically a Dutch schwa may be assumed, as in he:lə, he:le:, hele ‘whole, whole (noun)’ (<Du. hele), here ‘Lord (God)’ (<Du. here), tobo, tubu ‘tub’ (<Du. tobbe) and (a)bini ‘in, into, inside’ (<Du. binnen). If we take the so-called epithetic vowels in krabu ‘crab, lobster’, rigi ‘back’, roto ‘rat’ and stêre: ‘star’ to be original Dutch schwas, then we have the Zealandic forms krabbe, rugge, rotte and sterre with a Middle Dutch final vowel. The sporadic use of epithetic vowels in Negerhollands nouns seems to correspond in large measure to the use of a final schwa in their Zealandic counterparts.

The so-called epithetic vowels of Negerhollands thus appear to be mostly explainable as Dutch diminutive endings (-ie), original full vowels (from Dutch or from other target languages) or as colourings of Dutch or Zealandic final schwas. Whether all forms can be explained like this is a point for further research. A few cases are worth mentioning:

(a) the word duku, dugu, duko ‘clothes, blankets, cloth’ seems to contain a true epithetic vowel, as opposed to a fossilized diminutive or a historical word final vowel. Possibly we have a West African word of Dutch origin here (Norval Smith, personal communication).

(b) According to the early 19th-century Herrnhut grammar of Negerhollands cited by Hesseling (1905), the slaves are claimed to have said grooto ‘big, large’ instead of groot, and de Josselin de Jong (1926) gives gro:to: as a variant of gro:t. His texts give a different picture from the older grammar, however. So far we have only found the adjective gro:t and once the nominalization gro:to: ‘big one’ (<Du. grote) in di twe: gro:to: sinu ‘the two big ones, the two oldest’, with the plural marker sinu. Note that Oldendorp (1767-8) mentions in his dictionary grooto and groot as synonyms for ‘big’ but only the latter occurs as an adjective, for example in the entry ‘greedy’: em ha goe groot hoogo ‘he is greedy, cannot get enough [lit.: he has very big eyes]’.

(c) Voet ‘foot’ is fut or futu, futo, and the final /u/ has not yet been accounted for.