The present work is a lightly revised version of my 1993 publica-
tion of the same title (=SPIL 27). I have corrected a number of
minor but irritating typos and editorial infelicities. I have
also made a few substantive changes to improve clarity and keep
the work in line with my current thinking. Because this work is
being distributed in South Africa and to a small number of fellow
specialist-colleagues abroad, a basic knowledge of Dutch and/or
Afrikaans is presupposed. Readers should note that I have made
no attempt to normalize the spellings in the citations from Van
Rensburg (ed.) 1984. Citations from this corpus are given
diplomatically, and the transcriptions contained therein are
assumed to be accurate in their morphosyntactic aspect.
I am grateful to Dr. Hans den Besten (Amsterdam), Dr. Mark
L. Louden (Austin), and Prof. Sarah Grey Thomason (Pittsburgh)
for sharing with me their thoughts on the original version. As
always, acknowledgment does not necessarily imply agreement with
the positions I have taken, and I bear sole responsibility for
errors of fact, omission, and interpretation, and for remaining
inadequacies. Finally, I am grateful to Prof. Rudolf P. Botha
and the editors of SPIL, whose patience I have surely tested.
What follows remains essentially a working document that may
show, I'm afraid, the hallmarks of the genre. Because this
monograph is intended as a report on research in progress rather
than a definitive statement, I should welcome comments and
criticisms from interested readers.
28 April 1994 Paul T. Roberge
'If we go back in time, the problem of what Afrikaans is becomes
more and more difficult', wrote Valkhoff more than two decades
ago (1972:2), and notwithstanding a far better understanding of
the material facts, his words remain true today. In what follows
I shall elucidate the sociolinguistic nature of the formation of
Afrikaans at the Cape of Good Hope. In section 2 I explore the
social bases of glottogenesis within a pantheoretical framework
in the sense that the parameters I identify will hold for any
theory or model of glottogenesis at the Cape. To paraphrase
Woolford (1983:2): Although there are internal principles that
govern the theoretically possible linguistic paths along which
language may evolve in an extraterritorial setting, it is the
external factors that determine how radically its linguistic
structure will diverge from metropolitan norms. Section 3 is
devoted to a critical overview of both current and selected older
writings on how Afrikaans came into being. No one who has
investigated its history would seriously dispute that the
emergence of the new code was a much a social fact as it was a
purely linguistic one. But not everyone has put equal emphasis
on this truism. In numerous writings on our subject we find
widely varying degrees of concern with sociolinguistic relations
underlying the formation of Afrikaans. Section 4 explores the
implications entailed by adoption of the view that periods of
marked shifts in linguistic patterns are largely congruent with
significant changes in culture. An eminent linguist/anthro-
pologist of another generation, Harry Hoijer, was of the opinion
that in order to understand linguistic change, one must see it as
a part of a wider process of cultural change. Naturally, this is
not to suggest a causal connection between sociocultural trends
and specific linguistic changes. Rather, changes within the
various aspects of culture cannot be regarded as distinct and
unrelated but must be seen as different realizations of a single
process (Hoijer 1948:335). In section 5 I discuss the direction-
al gradience of linguistic items across social class by the end
of the Dutch India Company (VOC) era in 1795, with a view toward
elaborating on my claim (Roberge 1994) that the Cape Colony was a
continuum speech community. More precisely, the Netherlandic
speech community at the Cape consisted of a spectrum of lects
ranging from the 'High' Dutch of the expatriate power elite to a
Cape Dutch Creole. Rather than concern myself narrowly with the
origins of these linguistic items, I focus on their social
transmission and development in a context of interacting social
groups alternating among variants in their linguistic reper-
toires. As such, this essay departs somewhat from the usual
method of historical disquisition in Afrikaans linguistics, which
concentrates on single-feature etymologies and takes for granted
the formation of a socially accepted grammar.
2.0. If there is one parameter that has been regarded as central
to glottogenesis, it would surely be the continuity of language
transmission between generations (e.g., Sankoff 1979:23-25,
Markey 1981, Bickerton 1984:176, Mühlhäusler 1986:94, 255-58, and
especially Thomason and Kaufman 1988:9-12, passim). A priori
there would appear to be only two fundamental, nontrivial classes
of events whereby a new language could come into existence.
These are linear development and catastrophe.
2.1. By 'linear development' I mean gradual, incremental
processes of linguistic innovation (primary hybridization in the
sense of Whinnom 1971) and the social mechanisms by which change
diffuses throughout a speech community. There are neither sharp
breaks in linguistic tradition nor radical restructuring over the
short haul. Low-level rules are added to the grammar over time.
The grammatical core of the language remains intact and ety-
mologically transparent. Thus, discrepancies between succeeding
generations are relatively minor; there are no quantum leaps.
Our conventional understanding of 'normal' linguistic evolu-
tion defines a genetic tree or Stammbaum:
1 2 3 4
A A A A
| | | |
1' 2' 3' 4'
A A A A
| | | |
B C D E
Glottogenesis occurs with the establishment of new speech
communities (language spread) and the achievement of a signifi-
cant degree of Abstand or linguistic differentiation (Kloss
1978:23-30). One group splits off from the ancestral speech com-
munity, and both varieties undergo secondary and separate
evolutive change. We can say that before approximately 874 A.D.,
the island on which Icelandic was going to be spoken simply was
not inhabited. The raw material for what we know today as
Icelandic existed in the Norse dialects of the colonists, who had
been driven out of Norway and later the Scottish isles by Harald
Fairhair. Modern Icelandic has retained most faithfully the
structure and lexicon of Old Norse, although significant changes
in phonology are concealed by a classical orthography (Haugen
1976:32). Inhabitants of the Faroe Islands are probably descen-
dants of immigrants from Southwest Norway. According to Haugen
(1976:34), '[the] form [of Faroese] is . . . intermediate between
Icelandic and West Norwegian dialects, with enough distance from
both to make it unintelligible, unless spoken very slowly'. Of
course, whether such cases represent glottogenesis in any
interesting sense is another matter entirely. The degree of
Abstand required for recognition as a separate language is
ultimately arbitrary; and any attempt to specify a terminus post
quem separating ancestral and daughter languages leads to well-
Under circumstances we would consider 'ordinary', a language
has but one parent. In 'normal' transmission there can be a
certain amount of mixing and discontinuity. We speak of Vulgar
(popular) Latin having fragmented into several Romance vernacu-
lars that in their turn evolved into what we know today as
French, Spanish, Italian, Portuguese, Romanian, and Rhaeto-Ro-
mansch (e.g., Coseriù 1978:265). Yet, simplified and reduced
forms of Latin must have been utilized between Romans and
non-Romans (secondary hybridization in the sense of Whinnom
1971)--especially along the frontier, in the military, in trade,
and in Italy itself in the wake of the Germanic invasions. We
may stipulate that Scandinavian (850-1042) and Norman hegemony in
the British Isles introduced perturbations into the evolution of
English. The later stages of the Danish presence were character-
ized by intimate bilingualism and eventual assimilation into the
indigenous population. The Scandinavian legacy is represented by
lexical borrowings, replacement of the Anglo-Saxon third-person
plural pronouns (OE hie, hem, hiera) with they, them, and their
(ON peirr, peim, peirra), and onomastic elements; cf. Lass
1987:50-54. Following the Norman conquest (1066-70) the English
lexicon absorbed a massive influx of loanwords from medieval
French. At the same time, French never dominated outside the
elite spheres of society, nor was bilingualism pervasive
(cf. Lass 1987:54-61, Thomason and Kaufman 1988:306-15).
Notwithstanding significant contact with other languages, English
and the Romance languages are conventionally seen as a direct
continuations of antecedent languages (cf. Polomé 1983:132-55,
Thomason and Kaufman 1988:263-342). To consider the origin of
these languages possible cases of creolization would be an icon-
Finally, mixing may occur with the migration of a largely
homogeneous speech community that subsequently undergoes language
shift. The first developmental phase of Yiddish commenced when
Jews speaking Loez (Judeo-Romance) crossed the Rhine into Germany
from 1000 C.E. Fishman (1987) points out that the Jewish
community was at no time without a genetically transmitted
language for communication; they could always fall back on their
original language while acquiring German. The result was
presumably a xenolectal (slightly foreignized) form of German
that retained Romance and Semitic lexis and varied in terms of
proximity to German norms according to sphere of usage.
2.2. Nonlinear linguistic development commences with untar-
geted and untutored foreign language acquisition. Social condi-
tions may require communication between people speaking mutually
unintelligible and typologically very different languages. These
are typically: (a) indigenous trade between social equals
speaking a fairly large number of individual languages; (b)
military service involving mercenaries or conscripts of diverse
ethnic and linguistic backgrounds; (c) military or colonial
occupation involving monolingual in-migrants and indigenes (the
latter often speaking several languages collectively) and in
which the respective groups may or may not be of equal social
status (e.g., trade versus domestic service); (d) migrant
(foreign-worker) labor schemes in industrialized countries; (e)
indentured, impressed, or slave labor systems in colonial
settings, in which workers representing a multitude of native
languages must effect communication with the power elite and
The linguistic result of such encounters is likely to be an
auxiliary contact vernacular that arises more or less spontane-
ously and is not the native language of any of its users. These
ad hoc codes are highly impoverished (e.g., very restricted
lexicon, no inflectional morphology, no morphophonemics, deriva-
tionally shallow syntactic structure, etc.), and are suitable for
use only in a limited and rather specialized set of communicative
domains. Jargons (secondary hybridization in the sense of
Whinnom 1971) are ad hoc, individual solutions to the problem of
intergroup communication. They are highly unstable within given
individuals and nonuniform across the learning population. While
jargons exist in innumerable varieties (i.e., the speech of no
two speakers is ever quite identical), the aggregate is usually
easily labeled and stereotyped.
When there is sufficient opportunity for improvement in the
direction of the superstrate language (i.e., of the group holding
socioeconomic power), we no longer speak of a jargon but rather
an interlanguage; that is a developing system that is partially
independent of both the native language (L1) and the target
language. Succeeding generations may then acquire the latter
natively (usually as bilinguals early on) with or without
eventual language shift. However, the social situations listed
above are often defined by linguistically heterogeneous substrate
communities (with little or no power). There are often signifi-
cant barriers to targeted second language acquisition. A pidgin
results from the attempted use of the superstrate language by
substrate speakers sharing no other language in common but under
the influence of the dominant group (tertiary hybridization in
the sense of Whinnom 1971). Like a jargon, a pidgin is a reduced
and simplified form of language; unlike a jargon it has socially
accepted norms of pronunciation, lexical meaning, and syntax.
Stability, of course, is a matter of degree, as the pidgin will
vary somewhat in the mouths of the different L1 groups that use
it. Nevertheless, pidgins (as I use the term here) are quali-
tatively different codes than either jargons or interlanguages,
both socially and linguisitically (see Mühlhäusler 1986:ch. 5).
There is a functional relationship between the exigencies of
communication, on the one hand, and a pidgin's linguistic
elaboration, on the other. If external factors remove the need
for communication outright or favor bilingualism or language
shift, the pidgin is doomed to extinction. It may also happen
that the communicative exigencies remain constant or create new
domains. The latter case requires structural expansion of the
pidgin so that it can maintain itself as a referentially adequate
vehicle of communication within these domains (see Sankoff 1979,
More than any other sociolinguistic setting, plantation
agriculture is supposed to have been especially conducive to the
'catastrophic' development of language (cf. Reinecke 1937:57-63;
Bickerton 1979:7, 1984:176, 1989:17-19; Sankoff 1979:24-25,
Washabaugh and Greenfield 1983, Holm 1988:40-41). Regardless of
whether workers arrived by means of forced relocation (slave
labor) or indenture, they inevitably brought with them a wide
variety of languages (cf. Bickerton 1979:10-11). In the paradigm
case no one language group would be dominant enough so that it
could prevail by means of language shift on the part of other
groups; no second language was shared by enough people so as to
serve as a vehicle of intercommunication (bilingualism). For
almost everyone of the slave or indentured labor class, access to
the superstrate language would be tenuous.
'Catastrophic' glottogenesis presupposes extraordinary
circumstances (Sankoff 1979:24; Bickerton 1981:3, Thomason and
Kaufman 1988:ch. 6). It has been pointed out often enough that
the specifics of labor organization are crucial to whether a
plantation pidgin will actually develop into a creole language;
that is, becomes a native language for most of its users, with a
lexicon and syntax that are sufficiently robust to meet all
communicative needs. Slave labor normally entailed the separa-
tion of speakers from their native-language groups and arguably
created the most severe breaks in the transmission of language.
Sankoff (1979:24) reminds us that an important distinction exists
between 'Pacific' plantations, which used indentured labor, and
'Atlantic' plantations (i.e., in the Caribbean and West Africa),
which used slave labor. In the former case the labor force was
renewed by the continuous importation of workers on short-term
contracts, many of whom stayed on as immigrants. There was a
virtual absence of child language learners (at least in the early
period). In the latter case the slave labor force increased
continually due to natural human reproduction. This resulted in
large numbers of children, whose only means of intercommunication
was to nativize the pidgin. For Bickerton (1979, 1981, 1984,
1989), the presence of children--specifically an early generation
of children--is pivotal to his definition of "exogenous" (plan-
tation) creoles, which, ideally, arose out of a prior pidgin that
had not existed for more than a generation and in a population
where not more than 20% were native speakers of the superstrate
language and where the remaining 80% were substrate speakers of
diverse languages (1981:4). In a subsequent publication Bicker-
ton (1984:176) writes that the ratio of superstrate to substrate
speakers in a given creole community is only one of several
factors that determine the severity with which language trans-
mission could be disrupted (infra).
2.3. The relative continuity of language transmission
implies a continuum along which individual cases of glottogenesis
can be plotted, the theoretical poles of which are absolute
linearity and virtual nonlinearity. According to Bickerton
(1984:176-78), disruption will be most severe in cases of early
nativization of a minimal pidgin or jargon. Early nativization
is a hallmark of maroonage; the creation of communities of
escaped slaves virtually precluded the effective transmission of
preexisting languages (Saramaccan, Djuka). The early withdrawal
of the original lexifier language of a planation creole due to
political change cuts off further influence from native speakers
of the dominant language (e.g., Sranan). The rupture in genetic
transmission is somewhat less severe in the case of 'endogenous'
(fort, maritime) creoles, which, by contrast, remain in contact
with their substrate languages.
In the mid range of our continuum would be languages that
are frequently referred to as 'semicreoles' (Holm 1988:9-10) or
'convergence creoles' (Gilbert 1993a, b). The need for an
intermediate construct arose out of the empirical observation
that a number of languages exhibit many of the structural
properties of creole languages (e.g., simplification), even
though they appear not to have originated in the nativization of
a pidgin (cf. Mühlhäusler 1986:10, Holm 1988:10). Whereas 'true'
creoles develop where there is a radical break in language trans-
mission, many languages appear to have developed with only a par-
tial break (Mühlhäusler 1986:10). Scenarios that could conceiv-
ably yield results structurally similar to creole languages in-
volve multilingual societies in which a continuum of lects de-
velops between a superstrate language and several substrate lan-
guages, each lect reflecting varying degrees of substrate influ-
ence (Romaine 1988:160). Alternatively, 'next to mixing between
fully developed linguistic systems one also finds mixing between
full systems and developing systems' (Mühlhäusler 1986:10); that
is, a 'language that grew out of the close contact of a creole
with a non-creole, without itself ever having had a basilectal
stage' (Holm 1991:22). In Gilbert's model (1993a, b) members of
a colonial society have transmitted the metropolitan language
without interruption to their descendants. One or more groups of
speakers learned this language first as a pidgin and subsequently
as a creole, with varying degrees of convergence between acro-
lectal and basilectal varieties.
Putative cases include inter alia Singapore English (Platt
1975), South African Indian English (Meshtrie 1992), Brazilian
Vernacular Portuguese (Holm 1989:299-303, Gilbert 1993a, b),
African-American Vernacular English (Holm 1989:498-503, 1991),
and Yiddish following the expansion of Judeo-German-speaking
Ashkenazic Jewry into Slavic Eastern Europe from the 12th century
(Louden 1993). When reading the literature in creolistics and
even lately in Afrikaans linguistics in South Africa, one is
struck by the diversity of terminology positioning Afrikaans
within the intermundia between creole and noncreole: 'rudimentary
creole' (Hancock 1971:518), 'creoloid' (Trudgill 1978:49n.,
Makhudu 1984:96), 'fusion creole' (Markey 1981:25, 1982:
201-2), 'acrolectal creole' (Ponelis 1988:126), '(nonradical)
fort creole' (Den Besten 1989:226), in addition to 'semicreole'
(Thomason and Kaufman 1988:148, 251-56; Holm 1989:339-40, 1991;
Bruyn and Veenstra 1993:30) and 'convergence creole' (Gilbert
3. On the Genetic Transmission of Dutch in Southern Africa:
Major Positions and Issues
3.0. Neither the social situation nor the linguistic facts would
support a claim that Afrikaans is a 'true' creole language. The
circumstances that led to catastrophic breaks in the transmission
of language between generations--such as were created par
excellence by plantations using slave labor--were not present at
the old Cape. The slave population in the Colony never greatly
exceeded that of the Europeans. Numerical parity was not reached
until ca. 1730; and by 1798 the colonial population consisted of
25,754 slaves versus ca. 20,000 free settlers (Elphick and
Giliomee 1989:524). Nor were there any large slave-holders save
for the VOC itself (cf. Raidt 1983:14). The topography of the
Cape was ill-suited for plantation agriculture, and in any event
the VOC did not at the outset envisage colonization as an end in
itself. Linguistically, Afrikaans appears more creolelike than
metropolitan Dutch but in turn displays far fewer prototypical
characteristics than Negerhollands (Virgin Islands Dutch Creole),
Berbice Creole Dutch or Skepi Creole Dutch of Guyana, all three
'true' creoles (cf. Markey 1982, Ponelis 1988, Bruyn and Veenstra
3.1.1. Afrikaans has maintained a fundamental typological
feature of continental West Germanic languages (i.e., Dutch and
German) that is virtually unknown in creoles; namely, underlying
SOV word order with verb-second (V2) phenomena. Despite signifi-
cant innovation in some systems (infra), the syntax of standard
Afrikaans does not diverge radically from Dutch in essential as-
pects. Afrikaans has preserved the Dutch periphrastic perfect
with the auxiliary het 'have' (Dutch hebben) and morphological
marking on the verb (viz. ge+present verb stem). It is true that
Afrikaans has lost zijn as a perfect auxiliary, but so has
English. And while it is also true that Afrikaans retains only
vestiges of the Dutch preterite, this particular absence is not
without precedent in other West Germanic languages (southern
German dialects, Yiddish). There is case syncretism in the
first- and third-person plural pronouns (Dutch wij/ons, zij/hun,
beside Afrikaans ons, hulle) but the rest of the pronominal sys-
tem shows the same inflectional oppositions as Dutch. Pluraliza-
tion is achieved by means of suffixation rather than anaphora or
reduplication. Curiously, attributive adjectives are frequently
inflected in Afrikaans, the criteria for inflection being phono-
tactic and in part semantic. Donaldson (1993:163) is quite cor-
rect in noting that adjective inflection is one area of Afrikaans
grammar where simplification has not occurred; see Lass 1990a.
The lexicon is mainly Netherlandic in origin despite the
fact that Portuguese and Malay have left their mark. Khoikhoi
lexis in Afrikaans obtains chiefly from adlexification for plant
and animal names, expletives, and some cultural items. Nether-
landic patterns of word formation remain virtually unchanged
(Raidt 1983:160), save for the addition of reduplication (infra).
Phonemically, standard Afrikaans is derivable in the main
from vernacular and dialectal Early Modern Dutch. Theoretically,
its phonotactic divergences could have resulted from metropolitan
Dutch, universals of untutored second language acquisition,
substratum influence, or the interaction of some, if not all of
these factors. At first glance, the apocope of final /t, d/
following a tautosyllabic obstruent (Dutch nacht, hoofd, Afri-
kaans nag, hoof) is consistent with the kind of coda simpli-
fication that can result from intensive language contact;
cf. Hesseling 1899:152- 53, 1923:126; Holm 1988:110. However,
the probative value of parallel phenomena in other contact
situations is diminished by the fact that cluster reduction is
widely attested in colonial Dutch before 1700 and in contemporary
Dutch dialects (Kloeke 1950:284-87, passim; Raidt 1974:99-101,
1983:80-82, 1991:198-200; Ponelis 1991:68-72). It is certainly
reasonable to sense some relevance in the fact that 'Khoi syl-
lables have the canonical form CV (Hagman 1973:21)' and speculate
that 'phonotactic rules of Dutch dialects converged with those of
Khoi languages to simplify final consonant clusters in Afrikaans'
(Holm 1991:9). Empirically, the question of convergent phonotac-
tic processes is moot. More difficult is Afrikaans initial [sk-]
(skryf) for Dutch initial [sx-] (schrijven). Replacement of a
marked syllable onset with a less marked one would be consistent
with what we should expect to find in creole languages (cf. Valk-
hoff 1966:199, Den Besten 1987a:74). It is also true that
initial sk- < WGmc. *sk- is present in many metropolitan dialects
(Hesseling 1899:152; Kloeke 1950:225; Scholtz 1972:85-86,
1980:56; Raidt 1983:87; Ponelis 1991:55-56), even though sch- and
not sk- is evidently preserved in New Netherlands Dutch in North
America, which shares the same dialectal base as Afrikaans
(Buccini 1992). The velarization of posttonic final -n (Dutch
doorn, Afr. doring) has eluded unified explanation. Schonken
(1914:172-74) attributed it to substrate (Malay) phonology, but
not even Hesseling (1899:150-52) was prepared to go quite that
far. Dialectal antecedents in the metropole have been argued
(Scholtz 1963:208-14, 1980:60; Raidt 1983:88-89; Ponelis
1991:38-39), though a linkage between Netherlandic and Malay
tendencies is not unthinkable (Den Besten 1987a:83-84). Another
salient phonological divergence from Dutch is the nasalization of
nonhigh, nonfinal vowels, as in Afr. kans [k :s] beside kas
[k s]. Citing Hagman 1973:13, Holm (1991:8) draws attention to
phonemically distinctive nasalized vowels in the phonology of the
Khoikhoi language Nama. But this connection is offset by the
nasalization of vowels observed in metropolitan Netherlandic
dialects (Scholtz 1972:84-85, Ponelis 1991:53-54) and by its
presence in New Netherlands Dutch (Buccini 1992). Whether
nasalization in Afrikaans represents a generalized Netherlandic
dialectism, autochthonous innovation, the influence of a sub-
strate language, or some combination of all three forces is
difficult to determine.
Although Afrikaans has not moved as far from Dutch as 'true'
creoles have from their lexifier languages, the changes that
distinguish it from metropolitan forms are still fairly exten-
sive. Afrikaans has divested itself of the bulk of Dutch
inflectional morphology. Case distinctions, such as they existed
in spoken seventeenth-century Dutch, are nonexistent except in
singular personal and anaphoric pronouns and as relics in fixed
expressions (ten slotte 'in conclusion', destyds 'at that
time'). With specific reference to deflexion we note the
following losses: (a) grammatical gender in nouns; (b) a distinc-
tion between nominative and oblique cases in the first and third
person plural pronouns (supra); (c) a separate reflexive pronoun
corresponding to Dutch zich (German sich); (d) the Dutch demon-
stratives deze/dit, die/dat, which have been superseded by
neologistic hierdie, daardie; (e) personal agreement in verbs
(even in the copula) and a formal distinction between finite and
nonfinite forms (save for is/wees and het/hê); (f) grammatical-
ized apophony (Ablaut) to mark categories of the verb, save for
in attributive past participles, where it is strictly morphophon-
emic; (g) past participle suffixes (weak -t/-d, strong -e(n)) in
periphrastic tenses and concomitant reanalysis in attributive
past participles; (h) the preterite as an inflectional category
(except in modal auxiliaries, weet, wees, and dink) and also the
pluperfect in the active voice; (i) all trace of a morphological
subjunctive; (j) weak allomorphs of pronominal forms (Dutch
ik/'k, wij/we, etc.) and of the adverbial daar (Dutch daar/er).
There is a uniform relative pronoun for all antecedents regard-
less of number, definiteness, or humanity (Afrikaans wat, Dutch
die/dat/ wat). The possessive particle se (phonologically
derived from the weak allomorph of sij(n), Dutch zijn/z'n) does
not vary according to gender and number of the possessor: Afri-
kaans ma se hoed, die kinders se skool beside Dutch vader z'n
hoed, moeder d'r (haar) hoed, de kinderen hun school.
Significant innovations in Afrikaans grammar include
reduplication (die dokter vat-vat die swelsel), the associative
construction (Piet-hulle), the verbal hendiadys (hulle het 'n
glas water gestaan en drink), use of vir before personal objects
(hulle het vir my geslaan), conversion and unification of
function (ek is honger), and the double negation with nie in
sentence-final position (die son gaan nie vroeg onder nie). If
one looks to nonstandard varieties of Afrikaans, then one finds
still greater divergences. Particularly noteworthy is the use of
positional verbs as preverbal aspectual markers, e.g., hierie
mense loop gee solke snaakse name, nou sit sing hulle virie
heeldag (note SVO order).
3.1.2. Nowadays, virtually everybody agrees that the trans-
mission process was 'bent' but not broken in the early years of
the Cape Colony (cf. Thomason and Kaufman 1988:253). There is
general agreement, too, that by 1740 an extraterritorial variety
of Dutch had come into existence in the Cape Colony, and that by
1775--certainly no later than 1800--we may speak of a separate
but cognate Netherlandic language (cf. Raidt 1983:6-8, 15,
27-28). There are good reasons to be skeptical of the received
termini ad quem (cf. Roberge 1994), but we shall accept them here
as a working hypothesis. Older theories generally posit the
formation of Afrikaans by the end of the seventeenth century,
although assignment of such an early date has consistently failed
to gain lasting acceptance.2
At the same time, 'the drastic inflectional simplifications
and consequent remodelling of Dutch structures in Afrikaans are
not typical, as a set of changes, of any European Dutch dialect
or dialect group' (Thomason and Kaufman 1988:255). Moreover,
they are much too extensive to have occurred solely by means of
internal evolutive change within the elapsed time (Kloss 1978:
151, Thomason and Kaufman 1988:255). Exactly how far the
transmission of grammar was 'bent' has been an enduring crux in
Afrikaans historical linguistics. The various positions fall
into three categories, with varying degrees of overlap and
difference in emphasis. They are not necessarily incompatible
(cf. Kloss 1978:151); and it is important to bear in mind that
the questions asked are often not the same.
3.2. Models in the first category proceed from the assump-
tion of more or less normal transmission of language within the
socially dominant European community at the old Cape. To the
extent that basilectal forms of Afrikaans are considered at all,
they are seen as the result of untutored second language acquisi-
tion on the part of the indigenous Khoikhoi and slaves of African
and Asian origin, followed by language shift on the part of their
descendants. Their varieties are seen as separate developments.
3.2.1. One early school of thought placed the genesis of
Afrikaans squarely at the normal transmission end of our line-
arity continuum. This is the so-called 'spontaneous development'
model initially proposed by Kruisinga (1906) and with which the
Afrikaner linguists Boshoff (e.g., 1921:78, 1959), Bosman (1923,
1947), and Smith (1927, 1952) were in substantial agreement.
Briefly, currents of change present in Dutch, Flemish, and Low
German dialects rapidly became diffuse in an extraterritorial
setting. Advocates of spontaneous development assumed minimal
language contact and fixed the origins of Afrikaans in the late
seventeenth century. The one exception is Smith (1927:19,
1952:20), who opined that Afrikaans did not reach its modern form
until the year 1750.
3.2.2. For his part, Bosman (1923, 1947) conceded that
spontaneous development alone could not account for the relative-
ly rapid transformation of Dutch into Afrikaans (1923:43), even
though he appears to have had no quarrel with the prevailing
terminus post quem. Insofar as Afrikaans is not the spontaneous
development of Dutch on foreign soil, he wrote, its creolelike
features (e.g., deflexion) are attributable to the influence of
nonnative speakers of Dutch (1923:101-4).
This 'foreigners' Dutch' or 'adaptation' model is indeed
preponderantly eclectic (Reinecke et al. 1975:323); to call it
'evasive', as Markey 1982:169 does, would hardly do justice to
this writer; cf. also Zimmer 1992:355. True, Bosman made no
attempt to apportion the influence of the various linguistic
determinants (Reinecke 1937:572). Yet, he drew a distinction
between European and creolized varieties (e.g., 1923:83), and
indeed many of his views strike the alert reader as prescient in
the light of subsequent research. Nienaber (1934:54,
1949:121-32) thought that Bosman was essentially on the right
track even though the empirical underpinnings to his formulations
did not run deep. As we shall see, it is but a short step to
expand Bosman's notion of 'foreigners' Dutch' and recast it into
contemporary second-language acquisition terms.
3.2.3. The Dutch dialectologist G. G. Kloeke (1950) is
conventionally included in the spontaneist camp (Reinecke et
al. 1975:323, Combrink 1978:86, Markey 1982:169, Makhudu
1984:13), and it has perhaps been all too easy to overlook the
fact that he saw his book as a reaction to some of the more
fanciful assertions of both creolists and spontaneists (346). He
inferred from the relative uniformity of (Euro-) Afrikaans over a
vast territory of South Africa that the new language must have
gained its most characteristic features before 1700--well before
contact with other languages could have played any significant
role. According to Kloeke, Afrikaans shows some striking affini-
ties with dialects in the southern part of the province of South
Holland. He attributed a strong 'founder effect' to the language
of the outpost's first commander Jan van Riebeeck and his
entourage, the bases of which must have lain in South Holland
(1950:289-302). At the same time, Afrikaans is not the pure
development of a single Netherlandic dialect, for one can discern
compromises with the 'High' Dutch of the period. In the absence
of prescriptive norms reinforced by education and with the rapid
assimilation of German and French immigrants, 'the younger
generation must have "murdered" the language' (1950:346, 363).
3.2.4. In an extreme revival of the spontaneist model Van
der Merwe (1963, 1964, 1968, 1970) went so far as to claim that
the emergence of Afrikaans took place within a scant four to six
years (1656-58) after Van Riebeeck's arrival (1968:66). Glotto-
genesis was essentially predetermined by latent tendencies toward
change (taalneiging) inherent in the structure of sixteenth- and
seventeenth-century Dutch--in other words accelerated drift.
Geographic displacement of Dutch speakers along with an influx of
new European immigrants upset the equilibrium of Dutch grammar,
unleashing a wave of structural readjustments in which purely
internal factors governed the succession of changes. He cate-
gorically ruled out any possibility that people of color
contributed significantly to the shaping of Afrikaans (e.g.,
3.2.5. If by 'spontaneous development' we are given to
understand that Afrikaans arose some time between 1658 and 1750
through a series of 'perfectly ordinary internally motivated
changes from Dutch', then the time factor itself becomes an
explanandum. After all, such a chronology 'flies in the face of
everything we know about ordinary rates of internally motivated
change. We do not suggest that we can specify precise rates of
change, but rather that the changes from Dutch to Afrikaans,
apparently during the early years of the Cape Colony, were much
too extensive to have arisen solely by internal means within the
elapsed time' (Thomason and Kaufman 1988:255). Contrary to the
impression one might glean from an uncritical reading of Valkhoff
(e.g., 1971:464, 466n.; 1972:1-2, 12-13, 34-41), the idea of
spontaneous development was long obsolete even in South Africa by
the mid 1960s. Kloeke's book has intrinsic value by virtue of
the data that his dialect-geographical approach makes available;
not surprisingly, it earned serious international attention. By
contrast, Van der Merwe's views were deservedly passed over in
silence by serious students of Afrikaans historical linguistics
(cf. Raidt 1977:73).
3.2.6. The prevailing model of linguistic development at
the Cape has emerged out of what Den Besten (1987a) has called
the 'South African philological school', the two most prolific
writers being J. du Plessis Scholtz (1963, 1965, 1972, 1980) and
Edith H. Raidt (1974, 1983, 1984a, 1991, etc.). Scholars working
within this paradigm have not concerned themselves with theor-
izing the origins of Afrikaans, which they regard as a dubious
enterprise (Scholtz 1963:74, 1980:29-30; Raidt 1976b:163n.).
Instead, they have concentrated on the history of specific
Underlying the philological paradigm is a developmental
model that presupposes continuous and linear development within a
contact situation. Accordingly, Afrikaans evolved from Early
Modern Dutch by a series of internally motivated changes effected
by the generalization of Netherlandic dialectisms and secondary
autochthonous development. The Netherlandic tongue imported to
the Cape existed in several dialects. The provinces of North and
South Holland seem to have been especially well represented at
the outset, although there followed speakers from Utrecht,
Brabant, Flanders, Zeeland, and the eastern regions (Raidt
1983:17). After 1700 there is a discernible slope toward
deflection and regularization. Our source material indicates a
transition period between 1740 and 1775. Raidt (1983:15)
proposes the term 'Cape Dutch' (Kaaps-Nederlands, Kapniederlän-
disch) if a designation is deemed desirable or expedient. Some
changes that define Afrikaans were already in place, while others
were still in progress. By 1800, however, we can assume a more
or less uniform and stable vernacular (Raidt 1983:6-8, 27-28),
somewhat different in the mouths of the Khoikhoi, slaves, and
subsequent generations of mixed descent.
A century and a quarter is still a fairly short period for
glottogenesis to occur solely by means of 'ordinary' linguistic
change. According to the philologists, the large number of
nonnative speakers using the Dutch target language in a multilin-
gual society--speakers of High and Low German dialcts (Grüner
1982) and French (Pheiffer 1980) as well as Khoikhoi and slaves--
accelerated the pace of change, all the more so given the absence
of strong normative pressures.
Theoretical discussion of language contact does not play a
part in the philological literature. The proffered accelerating
factor would seem to resuscitate Bosman's idea of 'foreigners'
Dutch' as a secondary mechanism of change (1923:56-60). Native-
language (L1) interference and imperfect approximation of the
superstrate resulted in 'broken language' (roughly jargons and
interlanguages) but not outright pidginization, much less creoli-
zation (Raidt 1978:119, 1983:24-28, 1991:124-31, 176-77; Pheiffer
1980:1-11). At first, speech 'errors' were random and unsys-
tematic. Eventually, they coupled with the gradual diffusion of
internal linguistic change in progress, or else introduced
perturbations into patterns of variation inherent to the fledg-
ling speech community. A case in point is the replacement of the
Dutch relative pronouns die, dat, etc. with wat in Afrikaans,
which according to Ponelis (1987:69-70) was abetted by invariant
que in Creole Portuguese and to a lesser degree Malay (as
Hesseling 1923:121 speculated).
Empirically, scholarship conducted within the philological
paradigm has much to recommend it. One proceeds inductively from
a thorough investigation of the documentary evidence to a compar-
ison of Afrikaans features with what we know of Early Modern
Dutch and what we can impute to that period on the basis of
modern Netherlandic dialects. The heuristic procedure is not
precisely the one that comparatists have always applied to gen-
etically related languages. The comparative method leads to a
uniform protolanguage without any dialectal variation (except as
necessary to accommodate irreconcilable differences in the daugh-
ters). In the case of Afrikaans the relationship is quite the
opposite: synchronic uniformity has resulted from diachronic
After a tabulation of unambiguous continuities from Dutch
(which require no explanation), one searches for forms in
metropolitan Netherlandic dialects that are similar enough to the
divergent Afrikaans features. Features that can be paired off in
terms of these correspondences are considered 'explained'. It
would be naive to think that an overseas territory settled
originally from one area of the metropole should always and
exclusively show dialect features from that area. Lass (1990b)
has stressed that when a cluster of related but markedly differ-
ent dialects move into an extraterritorial setting, there are two
developmental options. Either one particular input type will
dominate; or there will be varying degrees of mixture and re-
codification, with compromise outputs stemming from a variety of
inputs. Since Kloeke 1950, there has been nearly unanimous
agreement that South Holland occupied a special position in the
early days of the Cape colony, but immigrants from that region
never constituted an absolute majority (Kloeke 1950:229-88). We
should therefore hardly expect all defining features to be trace-
able to South Holland.
The development of Afrikaans is complicated by the fact that
as a general rule it did not entail multiple migrations in which
the component waves had a distinctly regional character. It is
not possible to separate 'archaic' from 'advanced' linguistic
patterns stemming from successive layers of continental Dutch
dialects. Invoking dialectal substrata in order to account for
the unexplained residue requires that two conditions be
satisfied. First, there must be a metropolitan dialect that
provides a plausible linguistic model. Second, there must be
independent evidence for the presence of speakers of the dialect
in question 'at the relevant time and in sufficient numbers'
(Thomason and Kaufman 1988:255). The replacement of nominative
wij by oblique ons in the function of subject has been recorded
in the Dutch province of Zeeland (Raidt 1983:155). We also know
that there was significant emigration from Zeeland between
roughly 1685 and 1700 (Katzen 1982:198). So what at first blush
appears to be an obvious creolism (Valkhoff 1966:222, 1972:45-47)
may find its origin in an old dialectism. To infer from these
facts a Netherlandic origin for subjectival ons is a defensible
(but not necessarily unchallengeable) position. That there is no
neutralization of the nominative/ oblique distinction in the
singular (ek/my, jy/jou, etc.) would lend some slight support.
Mühlhäusler (1986:123) has warned that as regards the
determination of substratum influences in creoles, nothing is
more misleading than a simple static comparison between two
languages in their synchronic states. This caveat applies
mutatis mutandis to a comparison of metropolitan and Cape
varieties of Dutch. Problems arise when one or neither condition
is met for a specific feature, and they become acute when that
feature is not attested until fairly late in our Cape corpora.
The etymologically opaque Afrikaans double negation shows a
superficially striking resemblance to a negation pattern that
Pauwels (1958) found in the modern dialect of Aarschot in
Brabant. Yet, Den Besten (1985:13-30, 1986:199-206) has shown
that the Aarschot pattern is structurally and pragmatically quite
different than what we find today in Afrikaans. One cannot in
any event draw a connection here unless one can show on in-
dependent grounds that the Cape settlement included a significant
dialectal substratum with this kind of negation. As it happens,
there were Flemish speakers in the service of the VOC during the
seventeenth century. But they were very much in the minority and
interspersed among the Hollanders; their linguistic influence was
otherwise marginal. In the case of the double negation, only the
first condition is met, and even at that very imperfectly.
It is certainly not enough to suppose that if a linguistic
feature can be Netherlandic, it must be Netherlandic (cf. Den
Besten 1986:191, Thomason and Kaufman 1988:255), although such
kneejerk opinions hardly characterize the scholarship cited in
the present section (§3.2.6). Reduplication and the object
particle vir do not find clear cognates in any continental
variety of Dutch, and Raidt has fixed their origins (respec-
tively) in Malay (1980, 1981) and Creole Portuguese (1976a).
More usually, linguistic data from the formative periods at-
testing to the usage of slaves and Khoikhoi is at best scanty,
and one simply cannot do adequate philology. Nonetheless, only
in the last resort does one look to the substrate contact
languages. From the philologists' point of view, a remotely
plausible Netherlandic prototype must prevail when the evidence
is in equipoise. To my mind, the conspicuous absence of double
negation in the Afrikaans pattern from our Cape Dutch source
material before the early nineteenth century flatly contradicts
the received opinion that some facultative or discourse-dependent
dialectism imported from the metropole became grammaticalized at
the Cape. If the Afrikaans double negation is a Netherlandicism,
where has it been lurking during 150 years of Dutch hegemony?
True, avoidance of highly stigmatized variants in writing could
explain the absence. But that explanation invites tautology and
raises the awkward question of why a supposedly diffusing feature
of dialectal Early Modern Dutch should command such a negative
sociolinguistic evaluation among a rural, insular, and semiliter-
ate settler population. Not without interest is the fact that an
Afrikaans-like double negation is entirely unknown in another
former territory of the Dutch colonial empire; namely, New
Netherlands Dutch in North America (Buccini 1992).
It should be obvious by now that the perspective adopted by
the 'philological school' is preponderantly Eurocentric. It
seems to me that what the philologists have posited by ca. 1775-
1800 is an idealized Euro-Cape Dutch, a composite of all defining
features. Contemporaneous forms of basilectal Cape Dutch are
seen basically as ancillary and epiphenomenal. Note how the
philological approach spreads the 'interference' factor more or
less equally among several groups of European and non-European
learners of Dutch. Raidt (1983:155-56) is too astute a linguist
to claim that Zeelanders with subjectival ons in their dialect
simply imposed this feature on everybody else, and so she allows
that subjectival ons advanced at the Cape in part 'durch den
Einfluß der Fremdlinge'. That is as far as she takes us.
Idealization of Euro-Cape Dutch and the separation of basilectal
forms represent a useful diachronic abstraction, but it could
hardly have been sociolinguistic reality (see Roberge 1994).
3.3. Glottogenetic models of the second type proceed from
the same postulates as those in the foregoing discussion, except
that they stress the leveling of grammatical systems between
closely related West Germanic dialects in contact (rather than
internal linguistic change). Early proponents of this model were
Wittmann (1928) and Louw (1948). Wittmann's contention (1928:43,
57) that a variety strongly divergent from Dutch had come into
existence already by 1685 is antiquated and untenable. Louw
(1948:87) saw convergence as a far slower process, and he allowed
a century for leveling to run its course. It seems gratuitous to
add that like proponents of the older spontaneous development
model, both writers reserve the mechanisms of change exclusively
for the European settler community.
The basic principle is that when genetically closely related
dialects come into contact in a foreign environment, they will
coalesce into a uniform code (koine), often with a greatly
simplified morphology (Mühlhäusler 1986:12, Holm 1988:10). Kotzé
(1991) has stressed that the history of Euro-Afrikaans satisfies
all of Siegel's (1985) requirements for koineization: (i) mixing
of mutually intelligible codes; (ii) the gradual nature of this
simplification; and (iii) sustained intensive contacts and
gradual assimilation of social groups.
In a similar vein Combrink (1978:72-77) explains the demise
of personal agreement in the Afrikaans verb as the linguistic
consequence of mixing between similar but nonidentical inflec-
tional systems. The Netherlandic and Low German dialects
imported to the Cape were structurally and lexically so similar
as to be mutually intelligible. The main barrier to communica-
tion lay in the interdialectal 'channel noise' created by inflec-
tional disparities. Because the exigencies of efficient communi-
cation implied greater reliance on syntax and lexical roots,
verbal inflections became completely redundant and thus dis-
posable. In this way Cape Dutch could be morphologically
stripped even while preserving intact its continental West
Germanic syntactic typology.
3.3.1. For Van Rensburg (1983:138-39, 1985, 1989), the
history of Euro-Afrikaans represents a continuous process of
koineization; namely, leveling of inherent variation (taalver-
vanging) coupled with the generalization of erstwhile variable
rules (reëluitbreiding). The language of the Dutch rank and file
at the Cape (viz. uneducated peasants and ordinary VOC employees)
is equatable with nonstandard varieties already spoken on the
continent and in the Dutch colonies (1984:514). Oosgrens-Afri-
kaans represents a convergent form that can be historically
identified as the language of settlers who established themselves
along the eastern frontier. The northern varieties of Afrikaans,
as spoken in the Transvaal and the Orange Free State, are based
As before, the accelerating factor is seen as developing
second languages among the colony's non-Dutch adult population,
for whom nonstandard Netherlandic varieties imported to the Cape
served as the target. Whenever such cases can be observed
directly, nonnative varieties of a target language are typically
characterized by reduction, simplification, overgeneralization,
and transference of structure from the native language. Creole-
like features in the contemporary varieties of their descendants
are in the main attributable to the interlanguages of adults
mastering a foreign language in the conventional way vis-à-vis
the creations of children.
Within the Afro-Asian substrate, language shift was preceded
by imperfect code switching on the part of adult language
learners in the early years of the colony, with succeeding
generations acquiring Cape Dutch natively (as bilinguals for an
indeterminate period of time). The contemporary Afrikaans of
people of color still bear the imprint of the interlanguages of
their forebears but are not sensu stricto creole languages
(cf. Van Rensburg 1985:138-54, 1989:137-38, 1994; Webb 1993; Van
der Merwe 1993). Kaapse Afrikaans (i.e., of the Cape Malay and
Cape 'Coloureds') is based on the varieties of the early slaves
and Khoikhoi communities in the Western Cape. Oranjerivier-Af-
rikaans (i.e., spoken by the Griquas, in the Richtersveld, and by
people of color in Namibia) represents a form of Afrikaans that
shows a greater influence of Khoikhoi languages and was spoken in
the regions along the Orange River.5
Standard Afrikaans is a relatively late developmental phase
(roughly 1870-1930) in which the settler vernacular spoken along
the eastern frontier (Oosgrens-Afrikaans) provided the dialectal
base and which proceeded under the influence of Dutch prestige
norms (vernederlandsing); cf. Van Rensburg 1983:139-41. During
the standardization process, East Cape Afrikaans made several
inroads into the other varieties, whereas the latter had no
measurable involvement in the overall process.
3.3.2. According to Van Rensburg, it is the convergence of
preexisting variants rather than evolutive change in the neo-
grammarian sense that was the 'driver' in the formation of
Euro-Afrikaans. As such, this postulate qualifies merely as a
shift in perspective, for traditional diachronic formulations are
easily translatable into variationist terms. One generally looks
in vain for explicit statements on how specific variables
coalesced over time, space, and social class. It is one thing to
offer programmatic allusions to a kind of linguistic stew during
the early VOC period that somehow managed to sort itself out; it
is quite another to reconstruct sociostylistic variation or early
koine, difficult though this will be.6
3.3.3. In what proportion language contact combined with
dialect leveling to produce Euro-Cape Dutch can of course be
debated (see Hesseling 1923:112-13; Ponelis 1993:29). That
disagreement aside, Ponelis (1988, 1993:27-30) proceeds from
essentially the same fundamental theses as the variationists: (i)
The Cape Colony was a heterogeneous, multilingual society in
which Dutch was a minority first language in the early years and
was approximated in a haphazard, untutored way on account of its
extensive use as a lingua franca. Furthermore, the colony
continually received new interlectal speakers (basically a
cocoliche situation). (ii) There was no withdrawal of the
superstrate language. Dutch continued as the first language of a
significant portion of VOC personnel and of the free settler
population. (iii) There was a spectrum between '(spoken) matri-
lectal Dutch and . . . a whole range of interlectal varieties'.
The interlectal codes within this continuum were characterized by
varying degrees of substrate transfer, simplification, and hy-
bridization, 'depending on closeness of contact' (1993:30). (iv)
Afrikaans today exhibits many structural properties attributed to
creole languages generally due to interlectal modification.
3.3.4. It is easy to get the impression from the views just
summarized that the explanatory power of interlanguage is
supposed to finesse the question of pidginization/creolization by
rendering these concepts derivative and ultimately dispensable
(cf. Van Rensburg 1989:142). For Ponelis (1993:27, 30) re-
structuring due to secondary proficiency is creolization.
Central to the 'interlectalist' position is the claim that the
developing second languages of adult Khoikhoi and slaves in the
initial contact generation (G ) became a viable primary lan-
guage. The disruption of language transmission would have been
minimal, for the substrate populations could avail themselves of
their native languages (e.g., Khoikhoi) and/or on imported lingue
franche (Creole Portuguese, Pasar Malay). This brings up the
problem of stabilization; that is, how to account for the
elimination of individual solutions to intergroup communication
and the establishment of social norms (Whinnom 1971:99, Mühlhäus-
Collectively, interlanguages would have represented a very
open system, having neither shared norms nor stability in given
individuals. Theoretically, the interlanguage continuum in the
Cape Colony between 1652-1700 could have ranged from the most
rudimentary jargon to fluent, nonnative Dutch. With regard to
Orange River Afrikaans, Van Rensburg explicitly rules children
out as agents of innovation: 'It does not seem that the children
who learnt this . . . language from their parents (next to their
mother tongues in the beginning) affected the original version in
a substantial way' (1989:138). He seems to presuppose that adult
interlanguages in G developed at roughly the same rate--and
further that fossilization (the point where learning in second
language acquisition permanently ceases) set in more or less
simultaneously in given individuals. That innumerable varieties
could have been reduced to comparative uniformity within G is
By definition, the existence of an interlanguage continuum
implies that second language acquisition in G was targeted
toward superstrate Dutch. With continuing improvement in per-
formance in achieving communication with Europeans, highly
individualized L1-transfer and spontaneous interlingual hybrid-
ization would have become increasingly ephemeral over time.
True, social distance between superstratum and substratum would
have perpetuated bilingualism and prevented total convergence.
Yet, the end result could only have been a somewhat indigenized
variety of Cape Dutch and surely not one that 'is widely recog-
nized as a pidgin or creole, or a language with distinct pidgin
or creole characteristics' (Van Rensburg 1989:136).
If the requisite measure of stability came about through the
nativization of intermediate forms of language by a succeeding
generation of children (G ), then it is unclear to what extent
the difference between the positions of Van Rensburg (loc. cit.),
Ponelis (1988, 1993:27-30), and Den Besten (e.g., 1989) is not
merely one of terminology. The only remaining avenues to
stabilization are the fusion of speech communities (which nobody
has claimed) or withdrawal of superstrate Dutch as the target
language of G . A compelling sociolinguistic reason for indi-
genes and slaves to maintain and nativize Dutch in the latter
circumstance is not obvious. Nor is the stabilization issue even
apprehended. As concerns the genesis of nonstandard varieties of
Afrikaans, many interlectalist ideas (which seem drawn from
simple bilingual situations) conflict with their own presup-
positions and so cancel themselves.
3.4. Glottogenetic models in the third category assume
still greater 'bending' in the transmission of language.
3.4.1. For the Dutch linguist D. C. Hesseling (1899,
1923:59), the transmission of language was disrupted by the
sudden encounter of two completely different peoples and lan-
guages. With the introduction of slavery in 1658, Creole Portu-
guese with an admixture of Malay is supposed to have become so
widely spoken in the Cape Colony during the period 1658-85 as to
leave a very strong impression on the Dutch language. Slaves
spoke this 'Malayo-Portuguese' (as Hesseling called it) among
themselves; it was for some a native language, for others a
previously acquired language. Colonists used it with them and
with other Europeans who did not know Dutch. Sailors and
officials were conversant in the lingua franca as well, for it
was in widespread use in the Dutch East Indies and in various
ports of call. The slaves at the Cape learned Dutch from their
masters, albeit defectively under the influence of their 'Malayo-
Portuguese'. European interlocutors were unable to avoid
absorbing a number of their linguistic patterns into their own
speech. Miscegenation was also a factor, as the emergence of a
half-caste population is supposed to have greatly accelerated the
erosion of the Dutch inflectional system and introduced a number
of creolisms into the syntax and lexicon. Hesseling (1923:ch. 1)
thought that the 'breaking down' of Dutch into Afrikaans was
largely completed by the end of the seventeenth century.
Ultimately, however, creolization was only partial due to regular
arrivals of VOC officials and new immigrants from the Nether-
lands, and also to the conserving influence of the Dutch church
and Bible (1923:59-60, 128).
3.4.2. In essential respects this was also the position of
Du Toit (1905), who sought to elaborate on Hesseling's views, and
Van Ginneken (1913:210-13), who regarded Afrikaans as essentially
creolized. Though primarily of antiquarian interest today,
Hesseling's views on Afrikaans have been most influential. They
are discernible in such widely disparate linguistic writings as
Kainz 1943:571-72 on reduced languages and Lockwood's thumbnail
3.4.3. For Franken (1927-31, collected 1953), the proto-
typical Afrikaans is that which has evolved as the spoken
language of the 'Coloureds' (1953:202-3). He concluded from his
study of early archival materials that Afrikaans evolved from
'broken' forms of Dutch that emerged already during the first
fifty years of Dutch occupation as the vernacular of slaves,
Khoikhoi, and their descendants of mixed race. It was during
this time also that the speech of European children came under
the influence of these varieties (1953:26, 95). Thus, Franken
followed Hesseling in favoring the late seventeenth century as
pivotal and stressing contact with people of color (even while
deemphasizing somewhat the latter's construct of a mixed 'Malayo-
Portuguese' lingua franca; cf. 1953:43).
3.4.4. Bosman (1923:53-60) rejected Hesseling's theory on
the grounds that the number of Asian slaves was insignificant
before 1715. Weighing both sides of their debate, Reinecke
(1937:568) commented that the argument in favor of the influence
of some adstratal 'Malayo-Portuguese' might have been stronger if
Hesseling had allowed that the transformation of Dutch continued
after 1685. Valkhoff (1966, 1972) appears to have recognized
this fact when reasserting the 'quick birth' of Afrikaans under
the influence of Dutch spoken by 'foreign peoples, namely Indon-
esians, Malayans, Indians, Hottentots, Bushmen . . . , Malagasy,
Negroes, and White foreigners' (1966:206). According to him,
Cape society from the second half of the seventeenth century and
still in the first half of the eighteenth century was so much
integrated that there was a very close intercourse between
Europeans, indigenes, and slaves. Valkhoff assumed the emergence
of a 'proto-Afrikaans' among the latter groups during the first
fifty years of Dutch occupation (1966:204-7; 1972:48-49). During
these 'linguistic encounters' Creole Portuguese provided the flux
in the semicreolization of Dutch, though Malay, the other lingua
franca of the East Indies, gradually overtook it as a slave
language in Southern Africa in the eighteenth century and left
its mark as well (1972:72, 83). For Valkhoff, then, the process
of transformation 'was started at an early date in the bosom of
the Coloured community, for whom Afrikaans is still the mother
tongue. As a matter of fact the situation was more complex than
it appears at first sight, and the Whites, too, had their share
in the transformation of the language . . .; and so did the
Dutch-speaking slaves and Hottentots' (1966:206-7). By the mid-
dle of the eighteenth century this proto-Afrikaans 'had probably
developed into a dialect so different from "Hollands" Dutch that
most people would regard it as a new language' (1972:72; cf. also
pp. 48-49, 83). Thus, Valkhoff followed Hesseling, Bosman, and
Franken in dating the 'actual origin' of Afrikaans from the late
seventeenth century (roughly 1685-1710). But he clearly hedged
his position by making allowance for a subsequent period of
development and convergence lasting several more decennia. Thus,
'pure' Afrikaans--the one truest to type--was the language par
excellence of the 'Coloureds' in the first half of the nineteenth
century, a little different according to whether it was spoken by
the Cape Malay or the 'Coloureds' proper (1972:7). The Boers
must have used a more or less similar language, but one that was
somewhat nearer to 'High' Dutch.
Valkhoff's first book (1966) provoked a heated quarrel with
old-guard spontaneists (e.g., Van der Merwe 1966) that did
nothing to advance the field. Valkhoff 1972 is a deeply reactive
monograph that offers little in the way of direct engagement with
existing historical research programs. For their part, the
philologists tend to dismiss Valkhoff's work as unscientific,
speculative, preoccupied with social conditions to the exclusion
of 'hard' linguistic analysis, and generally uninformed with
regard to the history of Netherlandic (Lubbe 1974:94-98; Raidt
1975, 1976b, 1977, 1983:42-46). As a Romance specialist, Valk-
hoff was by his own admission out of his element with respect to
Early Modern Dutch, Netherlandic dialectology, and Afrikaans
philology (1966:217). One may grant Valkhoff's claim that
Portuguese Creole and to a lesser extent (Pasar) Malay were used
as lingue franche among slaves, as also between Europeans and
slaves. However, Raidt (loc. cit.) has argued that he greatly
overestimated the impact of these languages on Cape Dutch.
Furthermore, Valkhoff staked much of his case on circumstantial
evidence involving the social setting in which Afrikaans em-
erged. It is not enough to suppose that if a feature can be a
creolism, it must be a creolism. Although simplification and
reduction are hallmarks of prior pidginization, their appearance
in Afrikaans is misleading. The continuity of our textual
sources militates rather strongly against regarding Afrikaans as
a creole. Nearly 200 years had elapsed before all of the
defining variables were fully in place. The abrupt and simulta-
neous structural changes typical of pidgins and creoles is
generally absent (Raidt 1983:191).
Mühlhäusler (1974:13) correctly observes that Valkhoff seems
to treat pidginization and creolization as the same process. The
latter's vaguely formulated notion of the 'breaking down' of the
original Cape Dutch (e.g., 1971:456) with miscegenation as a
causal factor was anachronistic and totally inadequate by the
late 1960s (cf. Mühlhäusler 1974:43). Yet even if nothing else
in Valkhoff's work turns out to be of lasting value, his insis-
tence that from the outset Afrikaans was subject to the continuum
principle (that is, we should speak of more and less 'advanced'
forms of Afrikaans) ought to be a serious consideration for any
student of its history.
3.4.5. One simple alternative to reconstructing actual
sociolinguistic processes is a comparison of Afrikaans with
recognized creoles having Early Modern Dutch as their lexifier
language. Markey (1982) proffers a gradient evaluation of
Afrikaans in terms of a roster of features that are supposed to
be universally present (or nearly so) in creole languages. As an
'unmistakably creole form of Dutch that is well recorded in the
literature' (175), Virgin Islands Dutch Creole serves as the
standard of comparison. Whereas Virgin Islands Dutch Creole has
all eleven diagnostic features, Afrikaans tests positively for
only two, with two other features being weakly present. This
leads Markey to conclude that 'in typology . . . as throughout
human language there are no absolutes. Afrikaans, for whatever
inexplicable reasons of environment and structure, is a transi-
tional language located on a continuum somewhere between creole
and non-creole' (204).
At one level, it is impossible to disagree with either
aspect of this conclusion as far as it goes, even if one does not
fully endorse the efficacy of his diagnostic features (cf. Bender
1987:44-45), or the very idea of characterizing creoles in terms
of selected features (cf. Le Page 1987:115). Markey's finding--
again as far as it goes--is entirely consistent with the emerging
recognition among creolists that Euro-Afrikaans is linguistically
much closer to Dutch than either Hesseling or Valkhoff averred;
cf. Mühlhäusler 1974:18; Makhudu 1984:57; Thomason and Kaufman
1988:256, Den Besten 1989:227, Holm 1989:339. At another level,
it is vacuous, amounting essentially to a mere restatement of the
problem it seeks to address; cf. also Zimmer 1992:352.
Makhudu (1984:3-4, 54-59) is rightly critical of Markey for
having culled his information chiefly from a normative grammar of
standard Afrikaans and for factual inaccuracies. His determina-
tion (1984:60-95) that 'coloured' Afrikaans and Flytaal show sig-
nificantly higher indices of creolicity provides a necessary
corrective to the woefully incomplete picture that the unwary
reader will adduce from Markey's article (or from its mostly un-
critical summary in Romaine 1988:55-62, although this author does
alert her readers to Makhudu's principal objection, p. 62). By
utilizing a modified version of Markey's checklist to test non-
standard varieties of Afrikaans in addition to the standard lan-
guage, Makhudu carries on with the enterprise of typological
classification grounded in the contrastive analysis of syncrhonic
states--even while offering a far more revealing examination of
the continuum of lects that are constitutive of Afrikaans. At
one level, Makhudu's general conclusion is entirely in line with
current thinking among those who adopt the creolist perspective:
'It now seems likely that pidginization and creolization did
indeed occur in the non-native Dutch communities of the early
Cape'. Racial separation preserved creolisms in the 'coloured'
community, while the Afrikaans of Europeans developed under the
'conserving influences of Dutch immigration and the promotion of
the High or standard variety of Dutch until the early 20th Cen-
tury' (1984:96-97). At another level, the 'creoloid nature of
the Afrikaans lectal continuum' is only suggestive; apparent
creolisms in synchronic grammar do not explain themselves. While
I am entirely sympathetic to Makhudu's conclusion, I do not agree
fully with the argumentation by which he reaches it.
Well-informed comparison of the endpoints of glottogenesis
can be instructive to the extent that it shows what developments
are possible as the result of intensive language contact (see
more recently Ponelis 1988:132-42; also Bruyn and Veenstra 1993,
who explicitly limit their discussion to standard Afrikaans).
Though often forced upon us by a lack of reliable documentation,
inference of prior processes of linguistic hybridization on the
basis of static comparison is methodologically suspect (cf. Mühl-
häusler 1986:206). Moreover, 'there is considerable disagreement
as to the world-wide similarities of creole structures' (Mühl-
häusler 1986:222). Yet, all this has done little to discourage
attempts to classify Afrikaans--or specific varieties thereof
(e.g., Kotzé 1989, Van der Merwe 1993)--on typological grounds:
How creolelike is it? Is it a postcreole? A fort creole? A
semicreole or creoloid? The decisionistic and asocial foundation
upon which such classifications rest undermines their usefulness
and leads me to view them with increasing skepticism. The end
result of this exercise is virtually guaranteed by its own pre-
suppositions ('in typology as throughout human language there are
no absolutes'); Afrikaans will inevitably fall somewhere within
the mid range of any reasonable scale of creolicity.
3.4.6. Today, most linguists who take a creolist view
concerning the genesis of Afrikaans will readily stipulate that
Dutch colonists at the Cape can reasonably have been expected to
pass their language on to their descendants in a continuous and
unbroken process of 'normal' transmission (Makhudu 1984:95;
Thomason and Kaufman 1988:252; Den Besten 1989:226; Holm
1989:343, Gilbert 1993a, b). Pidginization and subsequent
creolization took place within the Afro-Asian substrate. The
Afrikaans of Whites--particularly their colloquial speech--bears
the imprint of this former creole, but the antecedent vernacular
of European settlers was only peripherally involved in the
According to Den Besten (1978, 1986:224 et passim, 1987b,
1988, 1989), the Khoikhoi were the primary substrate community
during the early years of the Cape Colony. From as early as
1590, when the Dutch and English started calling at the Cape of
Good Hope, there came into existence a jargon used between
Europeans and indigenous Khoikhoi. From 1658 slaves were brought
in from West Africa, Mozambique, Madagascar, India, Ceylon, and
Indonesia. The slaves acquired this trade language in their en-
counters with the Khoikhoi and contributed their own modifica-
tions; it became stabilized as a pidgin during the last decennia
of the seventeenth century (Franken 1953; Valkhoff 1972:50; Den
Besten 1986:192-201, 1989:217-24). Creolization occurred first
in the Western Cape around 1700 following the withdrawal of
Khoikhoi into the interior to escape European domination and in
the wake of the smallpox epidemic of 1713 that decimated their
population. Nativization of the Cape Dutch Pidgin was effected
by slaves, the mixed offspring of Khoikhoi who remained behind,
and other free people of color. Modern Cape Afrikaans (supra) is
traceable to the pidgin and creole Dutch formerly spoken widely
in the Western Cape and now almost completely decreolized.
According to Kotzé (1984:42), the nonstandard characteristics
that are typically associated with Cape Afrikaans are quanti-
tatively most prevalent in the speech of the Cape Muslim communi-
ty. The retreating Khoikhoi took with them their own variety of
Cape Dutch Pidgin (possibly itself on the fringes of creoliza-
tion), which they later deployed in their encounters with trek-
boeren along the eastern frontier. The expansion of Cape Dutch
along the northern frontier coincided with the migration of mixed
Khoikhoi, slave, and European populations. Its synchronic
reflex, Orange River Afrikaans, is widely thought to have
descended from a creole ancestor (cf. Van Rensburg 1989:135),
though its prehistory remains poorly understood.
Den Besten (1989:226) regards Afrikaans as a 'fort creole'
in the taxonomy of Bickerton (1989), which differs less radically
from its lexifier language than a plantation creole. That
Afrikaans has remained linguistically close to Dutch is attribut-
able to three factors: (i) The population of the Cape Colony was
comprised of a high percentage of Europeans, who accepted
individual features from the Cape Dutch Pidgin/Creole but did not
adopt it in its entirety; in fact the European superstrate
exerted reciprocal influences of its own on hybridized forms of
Dutch; see further Ponelis 1988 on this point. (ii) The Cape
Dutch Pidgin/Creole was a second or third language for the many
slaves who could avail themselves of Creole Portuguese and/or
Pasar Malay. The availability of these lingue franche mitigated
ruptures in the overall transmission of language and limited
somewhat the importance of Cape Dutch Pidgin for intragroup
communication. (iii) The legally free Khoikhoi were in a better
position than the slaves to improve their performance in the
direction of the superstrate by virtue of their greater access to
that language (cf. Den Besten 1989:227).
The creolists are certainly correct to assert that simpli-
fied, reduced, and restructured forms of Dutch must have arisen
among the Khoikhoi and other Africans dealing with the Europeans,
as also among the slaves. Den Besten (1987, 1988, 1989:229-234)
has marshalled evidence for a Cape Dutch pidgin (or pidgins) from
the seventeenth and first half of the eighteenth centuries. He
has attributed the instantiation of certain Afrikaans features to
substratum influence. This means the direct transfer of struc-
tures belonging to the speaker's first language (implemented by
Netherlandic exponents) or the creation of new structures on the
basis of L1 interference. Examples of direct transfer would
include the associative construction in -hulle, -goed (Den Besten
1993) and the uniform possessive particle se, which is phono-
logically derivable from Early Modern Dutch sij(n) but syntactic-
ally patterned after possessives in Khoikhoi (Nama/!Ora di),
Creole Portuguese (sua) and Pasar Malay (punya) (Den Besten
1978:28-38). The Afrikaans double negation exemplifies the
grammaticalization of an interference neologism (Den Besten
1978:40-42, 1985:32-35, 1986:210-24). Den Besten ascribes other
Afrikaans features to universal strategies for the encoding of
meaning, examples of which are the neologistic demonstrative
pronouns hierdie, daardie (1988:26-27).
I do not wish to deal generally here with the substance of
Den Besten's claims; many of the phenomena he examines have been
massively discussed in the literature and are long-standing
etymological cruxes. My point is that all such hybrids would
also be characteristic of haphazard approximations of the Dutch
target language (Mühlhäusler 1974:17-18), or 'rapid and drastic
linguistic change due to imperfect learning' (Ponelis 1988:119).
There are important structural parallels between pidginization
and the interlanguage stage of language shift 'because both
reflect cognitive and linguistic universals at play in the
acquisition of another language' (Holm 1991:21). At increasing
time depths, their results may become indistinguishable. In
principle, then, the issue that divides the creolist and intelec-
talist positions is not so much one of the actual mechanisms of
hybridization, but rather what type of code was nativized within
the Afro-Asian substrate: a stable pidgin or unstable, transient
interlanguages emanating from gradient degrees of multilingualism
and language shift. Ponelis (1993:28) believes the likelihood of
a stable Cape Dutch pidgin having existed to be rather small, and
he challenges Den Besten's assertion that Khoikhoi pidgin Dutch
supplied the foundation for subsequent developments: 'He con-
siders no sociohistorical evidence . . . [and] his position is
based entirely on shaky linguistic evidence' (1993:33-34).
Whether one agrees with this assessment or not, the prevailing
occupation with identifying the source of individual features
(substratum transfer, universal strategies) does not reveal the
full story. It does not tell us whether suspected pidgin
features observable in our early source material reflected
socially accepted norms of syntax and lexical meaning or were ad
hoc, individual solutions to the problem of interethnic communi-
cation. The stabilization question presents itself as usual.
In order to sustain Den Besten's hypothesis, one will first
have to demonstrate the existence of a stable Dutch pidgin within
the Afro-Asian substrate (or a relexified 'Malayo-Portuguese',
improbable though this may be) and the subsequent addition of
grammatical rules and lexis (cf. Mühlhäusler 1974:18). To
consider nonstandard Afrikaans a postcreole would by definition
imply the existence of an erstwhile basilectal creole that is now
extinct but has nonetheless left its imprint. But so far nobody
has been able to show by the dint of direct evidence or con-
vincing argumentation that this is in fact what happened. We
have no linguistic documentation whatsoever of a first-generation
Cape Dutch Creole.
4. From Initial Contact to Social Convergence in Cape Colonial
The three groups primarily responsible for the formation of
Afrikaans--European settlers, Khoikhoi, and slaves--were quite
distinct during the first decades of the Cape Colony. This
distinctness was defined by physical appearance, culture,
religion, and language. By the end of Dutch East India Company
era in 1795, a number of processes had eroded these boundaries:
'(1) the incorporation of the Khoikhoi into the European-domin-
ated society as wage labourers subject to Dutch law, (2) the
conversion of slaves and free blacks to Christianity or Islam,
(3) miscegenation and intermarriage among groups, (4) the
manumission of slaves and the consequent emergence of an im-
portant new group--the free blacks, (5) cultural exchanges among
groups' (Elphick and Shell 1989:184). The present section
briefly examines the sociolinguistic implications of these
In 1652 Jan van Riebeeck and his party landed at the Cape of
Good Hope with the limited objective of establishing a refresh-
ment station on behalf of the Dutch East India Company. At the
outset, the population of the outpost consisted almost entirely
of the 100 or so Europeans in the original expedition. The
presence of VOC officials, ordinary servants, soldiers, and
sailors afforded some social stratification. Beyond this, social
and racial divisions were ill-defined and fluid during this
initial period. After 1657 the VOC allowed 'free burghers'
(vrijburger) to settle at the Cape to produce commodities needed
by the station. This group of European arrivals was comprised of
many Dutch, as also some 200 Huguenots, who had fled France after
the revocation of the Edict of Nantes (1685), and large number of
Germans in the service of the VOC. The Huguenots, who arrived in
the Cape Colony from ca. 1688, assimilated into the existing
Dutch-speaking population with little in the way of a direct
linguistic legacy. Within a generation, their French language
had disappeared (cf. Pheiffer 1980). In the main German free
burghers were able to achieve fluency in Dutch within a decade of
their arrival. Typically, they married Dutch-speaking women;
children of their issue were brought up to speak Dutch (cf. Grü-
ner 1982). These new immigrants merged with the existing Dutch
population to constitute the core of the Afrikaner community.
Buccini (1992) has demonstrated that the colonial Dutch
dialects (Cape Dutch and New Netherlands Dutch) reflect in all
general respects the spoken Dutch of the lower and middle classes
in the provinces of Holland and Utrecht during the early and mid
seventeenth century. Because of the founder effect, the tinge of
South Holland is perhaps more prominent in the Cape Colony than
the number of immigrants from that region might otherwise predict
(Kloeke 1950). Traces of other Netherlandic dialects may be
discerned (Raidt 1983:16-17), but they are in the main relatively
superficial (chiefly lexical). The prestige variety of Dutch
imported to the Cape was the educated speech of the urban upper
class of Amsterdam and The Hague (Raidt 1983:16). But it was
spoken by a minority of the population (the commander and
transient officials of the VOC), an elite to which the rank and
file--particularly migrant farmers along the frontier--would not
Already by the end of the seventeenth century, there had
emerged a distinction between Africaanders, people of European
descent who regarded the Cape Colony as their permanent home, and
expatriate VOC officials and other personnel in temporary resi-
dence there. (In the eighteenth century Africaander did not bear
the semantic burden of its modern cognate, but that is a topic
for a separate essay.) By the eighteenth century, the settler
population was no longer an undifferentiated community. The most
salient socioeconomic division arose between the colonists in the
southwestern Cape and the migrant farmers (trekboeren) in the
interior. The former were dominant politically. They consisted
of a 'small and fairly prosperous bourgeoisie' (Du Toit and
Giliomee 1983:5); namely, civil servants (some of whom were born
at the Cape), prosperous burghers in Cape Town, and a few wealthy
wine, grain, and vegetable farmers in the Boland. Beneath this
group were the majority of Capetonians and smaller farmers in the
surrounding areas, who lived respectably but enjoyed no great
wealth. After 1717, a growing number of individuals moved inland
to join the few free burghers who, since the turn of the century,
had established themselves as migrant farmers on land leased from
the VOC. This migration was due in no small measure to lack of
opportunities in the monopolistic official economy. Many colon-
ists could not afford to become entrepreneurs in Cape Town or
farmers in the southwestern Cape. They were not sufficiently
educated to enter the limited ranks of the Dutch colonial
administration or professions. Reliance on slaves and an in-
digenous labor force precluded the development of a white working
class (cf. Du Toit and Giliomee 1983:7-10). As these people
migrated further away from Cape Town, the authority of the
government and cultural influence of the city (such as it was)
Prior to 1658, there was only a handful of personal slaves
at the Cape, including a few in Van Riebeeck's household. The
first significant numbers arrived in that year from Angola and
Dahomey. Excepting a few individuals, they were the only West
African slaves who were brought to the Cape during the VOC period
(Armstrong and Worden 1989:111-12). Slaves were thereafter
imported from Madagascar, from Mozambique and entrepôts along the
East African coast, as well as from the Indonesian archipelago,
India, and Ceylon. The slave population also increased naturally
by procreation. The children of liasons between slave women and
European or Khoikhoi men were de iure slaves (Elphick and Shell
1989:202). By 1834, when the institution was abolished at the
Cape, as in other British colonies, the slave population had
risen to 36,169 (Armstrong and Worden 1989:109).
The diversity of the languages represented in the Cape slave
community guaranteed the deployment of extraterritorial lingue
franche. Valkhoff (1966:146-91) was certainly correct in his
assertion that Pasar Malay and especially Creole Portuguese were
readily available as means of interlingual communication among
slaves. Merchants, officials, sailors, magistrates, and other
VOC personnel who had lived in the East Indies could be expected
to be proficient to some degree in Portuguese. This has never
been a matter of controversy; see Franken 1953:15-27, 41-79,
116-43; Raidt 1983:20; Den Besten 1989:224; Davids 1991:44-46.
At the same time, Raidt (1983:20) is no less correct in her
assertion that these languages could not have been in general use
as lingue franche throughout the entire colony because too few
rank-and-file colonists knew them. The emergence of a Dutch
jargon to effect communication between slaves and master can
scarcely be open to doubt; see Den Besten 1987b, 1989.
The Khoikhoi were the first South Africans to confront the
Europeans at the Cape of Good Hope. Within 60 years of Dutch
occupation, the traditional Khoikhoi economy, social structure,
and political order had almost entirely collapsed in the south-
western Cape (Elphick 1977:ch. 11, Elphick and Giliomee 1989:18-
21). Smallpox, stock disease, and the advance of European
settlement during the eighteenth century destroyed some inland
Khoikhoi groups, realigned other groups, and drove still others
deeper into the interior. Many destitute Khoikhoi became hunters
and robbers; others became servants to the trekboers, living in
virtual serfdom. By 1800, there were few Khoisan in the colony
who were not in the service of the Europeans as laborers,
herdsmen, and nursemaids (Elphick and Giliomee 1989:35-43).
VOC policy was formulated with a view toward preserving the
Dutch character of the settlement. Europeans appear to have
placed little value on knowledge of the languages of groups that
were subordinate in status. Irrespective of official constraints
and attitudinal barriers, Khoikhoi languages were considered
impossibly difficult and all but unlearnable. Aside from two
officials who evidently had a working knowledge, 'it is not until
1711 that we hear of a few white children picking up Khoikhoi,
and not until the isolated frontier conditions of the mid-eight-
eenth century that such skills were common among settlers'
(Elphick 1977:210; cf. also Elphick and Shell 1989:229). The
upshot of all this is that in general if Khoikhoi wanted to
understand the settlers and be understood themselves, they had to
acquire Dutch. Fluency was rare before the early eighteenth
century. Elphick (1977:211) infers from fragments of Khoikhoi
Dutch recorded by Wilhelm ten Rhyne (1673), the German astronomer
Peter Kolbe (1705-12), and in judical records (1706-8) that 'even
those Khoikhoi who had regular dealings with the Dutch spoke in
broken dialects'. The historian may of course be excused for
writing in a general way of 'broken dialects' and shunting over
to the linguist the problem of whether the Khoikhoi Dutch jargon
developed in the direction of an interlanguage continuum or a
stable pidgin. That jargonized Dutch arose among the Khoikhoi
can also scarcely be open to doubt (cf. Den Besten 1986:193-99,
1987a:85-89, 1987b). The Khoikhoi learned languages other than
Dutch--among them French, English, and Portuguese (Franken
1953:28-40), and there is no reason to think that with few
exceptions, these, too, were jargonized.
The decline of Khoikhoi identity as it had existed prior to
1652 was exacerbated by attendant language shift. To be sure,
cultural change was more gradual than structural change. The
Khoikhoi continued to speak their own language among themselves
until the mid eighteenth century, at which time their dialects
began to disappear from the western Cape. One can envisage a
generational continuum in which fully fluent L1 speakers of
Khoikhoi were limited to the oldest members of the population of
the colony proper or those living on its fringes. Semispeakers
with varying degrees of fluency probably occupied the middle-age
group, none of whom transmitted Khoikhoi as a first language to
their own offspring.
Social differentiation of the Cape society was furthered by
emergent groups of people of color. The so-called 'free blacks'
(vrijzwarten) were of wholly or partially African (but not Khoi-
khoi) and Asian descent. This group came into being primarily in
Cape Town through the manumission of slaves, although a sizeable
number of free blacks traced their origins to miscegenation and
to Asian settlers, political exiles and convict laborers (Elphick
and Shell 1989:216). In the course of the eighteenth century
miscegenation between European men and Khoikhoi women produced
another new group, the Bastaards. This took place mainly in the
remote outlying districts (especially along the northern fron-
tier), where European women were few and where because of their
comparative poverty, trekboers relied less on slaves than on
Khoikhoi labor. The term Bastaard (or Bastaard-Hottentot) could
also denote the offspring of unions between Khoikhoi and slaves
(cf. Elphick and Shell 1989:202, 231; Armstrong and Worden
Elphick and Shell (1989:225-30) make the very important
point that the direction of cultural history during the VOC
period was towards convergence. By 1795, the various European
and slave cultures were merging with one another and with the
culture of the Khoikhoi.
In Cape Town a mixed European and Asian culture was
shared by Company officials, some burghers, and slaves,
though some of the latter managed to retain more
traditionally Asian traits, especially through con-
version to Islam. In the agrarian southwestern Cape,
slaves and Europeans seem to have shared in a culture
which was predominantly of European origin. In the
trekboer regions . . . the culture of slaves and
colonists was a composite of European and Khoikhoi
influences, appropriate to a livestock economy (Elphick
and Shell 1989:230).
While there were demonstrably strong pressures toward cultural
merger, there were clearly ecological and ethological barriers to
total convergence: 'Slaves had comparatively little opportunity
to become free, even in Cape Town. Khoikhoi never enjoyed the
rights of burghers, and free blacks and baptised Bastaards
gradually lost the privileges they once had' (Elphick and Shell
1989:232). Notwithstanding mutual poverty, free people of color
and poor landless Whites did not form a coherent class. The
latter tended to identify rather strongly with more prosperous
colonists, who often accepted them as tenant farmers (bijwoner).
Though legally free, the former remained proletarian (Du Toit and
There was always a need for communication between the
various segments of a polyglot society: between Dutch, Germans,
and French; between Europeans and indigenes; between Europeans
and their slaves; between slaves of varying ethnolinguistic back-
grounds; and between slaves of whatever background, the Khoikhoi,
and free Blacks. Dutch was the dominant language within the
limits of the Cape Colony between 1652-1795 (cf. Ponelis 1988),
though by no means the sole means of interethnic communication.
Given European hegemony and cultural convergence, it was in-
evitable that the descendants of the various groups would
eventually come to share in a common Cape Dutch vernacular.
Individual speakers would not be uniform in their experience of
this vernacular. That experience would vary according to the
socioeconomic relations outlined above.
5. Sociolinguistic Stratification of the Cape Colony at the End
of the VOC Era (1795)
5.1. It has long been the practice of Afrikaans historical
linguistics to divide the Cape Dutch speech community into
discrete compartments. The distribution of linguistic variants
is assumed to be roughly isomorphic with ethnic or status groups
(Europeans, Khoikhoi, slaves). Accordingly, linguistic forms
assumed whatever social valuations that were associated with the
respective groups by other members of the speech community.
Since Nienaber 1950, the literature has given the impression of a
speech community consisting of four or five Cape Dutch varieties
during the eighteenth century: the 'High' Dutch of the power
elite, the Cape Dutch of the settlers (slightly different in the
mouths of Capetonians vis-à-vis rural white speakers), plus the
Cape Dutch vernaculars of slaves and Khoikhoi. However, such
compartimentage is artificially static. Responsible Afrikaans
language historians have always acknowledged this, if only
fleetingly. Lectal boundaries were fluid and in flux in response
to dynamic social forces. Like anywhere else, the use of
linguistic variants at the Cape was determined by patterns of
social and stylistic norms (cf. Nienaber 1953:163, Loubser
1961:2, Scholtz 1965:101, Raidt 1984b:265-66).
5.2. I have argued elsewhere (Roberge 1994) that what the
philologists have posited in their studies as 'Afrikaans' was by
the turn of the nineteenth century an abstract and ideal type--a
composite of all features--that was probably spoken by few (if
any) South Africans during the period in question. In other
words there was no clear separation at some particular point of
the superposed standard and the vernacular. There were instead
any number of lects intermediate between the superstrate and the
most extreme form of the Cape Dutch vernacular. Let me il-
lustrate this point with data from what I consider 'acrolectal'
Cape Dutch (i.e., the variety of Cape Dutch closest to the
metropolitan language) from the end of the VOC era. This is
represented by the diary fragment of a prosperous Cape Town
resident, Johanna Duminy (née Nöthling) from 1797, which can be
shown to be extremely close to metropolitan Dutch even while
containing many extraterritorial features. Because her diary is
a personal document kept for her own private purpose, its lan-
guage cannot be regarded as pure orthographic fiction. In other
words we cannot make the simplistic assumption that the diary is
merely a failed attempt to write 'correct' Dutch in which her
true spoken language (more or less the same as what we know as
Afrikaans today) leaks through; cf. Roberge 1994.
Duminy uses the first person plural pronoun wij throughout;
subjectival ons is unattested. Sij and sulle alternate as the
nominative third-person plural pronoun; the oblique case forms
are haarluij (which can also serves as a possessive pronoun),
sulle, and the weak allomorph se (Modern Dutch ze). Gender in
the noun has virtually disappeared (de huijs, Dutch het huis).
De and die appear to be free variation as the definite article,
with het preserved only vestigially in the function of an
anaphoric (2a) or dummy pronoun (2b):
(2) a. ik wilt het [een boulte vee] niet enkelt verkopen
(Franken, ed., 1938:83).
b. het was vraaij weer (idem, 88).
The demonstrative pronouns are deese (proximal) and die (distal);
dit is unattested. The relative pronoun is uniformly die; dat
occurs only anaphorically with a sentential antecedent:
(3) tou was rijnoo genootsaak om de varkens op de wage te laaden
dat hat ook versuijmt (86)
Postnominal possessive forms show no sign of simplification:
(4) a. en een groote c(er)misbet lee redelikhuyse sijn vrouw
b. in die ouwers haar droufhijt (109)
Finally, Duminy makes frequent use of the weak form of the
adverbial daar, e.g., er was ook een mooiie bastert bul (88).
The pluperfect tense is intact, as is the usage of both
hebben and zijn as auxiliary verbs in periphrastic construc-
tions. Duminy appears to have captured even a subtle distinction
with regard to vergeten in (5e):
(5) a. hij see niet minder als die ander man heeft gekreegen
(Franken, ed., 1938:83)
b. ik bin buyte geweest (86)
c. ik hat ook een groote caatel gekogt (86)
d. sij ware de voorige dagt al na de vandiesie gereeden
e. ik heeft vergeeten om te sege (85)
Cf. Modern Dutch:
e'. ik heb vergeten te schrijven (i.e., did not think of)
e''. ik ben je naam vergeten (i.e., gone from memory)
For the relation of anterior events, the preterite is Duminy's
tense of choice. The distinction between 'strong' (ablauting)
and 'weak' (dental suffixal) inflection is preserved in preter-
ital conjugation: kwam, liet, gong/ging, kogt, schreefte,
bestelde, etc. (Modern Dutch komen, laten, gaan, kopen, schreeu-
wen, bestellen); similarly, in the past participle gekreegen
(krijgen), opgebragt (opbrengen), etc.
Duminy consistently maintains a distinction between finite
and nonfinite forms of the verb, but her usage vacillates between
inflected and endingless forms: wij sliep beside wij sliepe
(Modern Dutch wij sliepen). Cluster reduction is evident in the
diary (direk, Dutch direct), and one would think that it would
have brought additional pressure to bear on second- and third-
person singular verb forms and on the weak past participle
(gewerk for gewerkt). However, cluster reduction may not have
been as general in acrolectal Cape Dutch as one might suppose.
Several idiosyncrasies of Duminy's usage are hardly consonant
with the usual assumption of a fully diffuse cluster reduction
rule by the end of the eighteenth century: the presence of a
paragogic dental stop after a tautosyllabic velar obstruent in
other categories of words (dagt for dag, nodigt for nodig);
ahistorical -t in the present-tense first-person singular (ik
heeft, ik komt) and plural (wij komt, sulle heeft), and in the
strong preterite (wij sagt 'we saw', ik gaf/gaft); the fact that
a t-less variant of the auxiliary hebben (heef) does not occur at
all. The direction of change in acrolectal Cape Dutch seems to
be toward an invariant inflectional opposition: finite versus
nonfinite. As concerns personal agreement, it is the singular
(the exponents of which could be either zero or -t) that is in
the process of supplanting the plural termination -en. Note that
the weak preterite has already attained a stable state of uniform
finite inflection through the apocope of final nasals (ik maakte,
pl. wij maakte, Dutch wij maakten); the rest of the verb system
has not yet progressed this far. It is surely no coincidence
that Duminy writes wij sal [Dutch zullen] aan hem vraage (81),
but never zulle(n) where a singular form would be called for;
that she utilizes ik wilt (cf. Dutch ik wil, jij wilt), ik bin
buyte geweest en is in de verkeerde caamer gekoomen (86), but
never wille(n) or sijn. Nor does she introduce excrescent -t's
into the more suppletive paradigms (viz. *zalt, *ist). The ex-
crescent -t's in the strong preterite (krijgen, kreegt) reflect
the same kind of analogical projection as the nonce form kreegte
(strong preterite kreeg reinflected with the weak preterite
ending -te). As such, they have nothing to do with either
phonology or with hypercorrection but reflect instead the kind of
phonomorphological indeterminacy that foreshadows paradigmatic
leveling (cf. Roberge 1985, 1987).
The Duminy diary is no less important for the hallmark
Afrikaans features (§3.1) that it does not show: the double
negation, the demonstratives hierdie/daardie, reduplication, the
verbal hendiadys, subjectival ons, hulle etc. Nevertheless, the
features just summarized--together with divergences in pronunci-
ation, pluralization, and lexis (cf. Franken 1953:169-74)--would
be more than sufficient to mark the speech of even upper-class
Capetonians as an extraterritorial variety of language, albeit
one that is still recognizably Dutch.
5.3. For their part, interlectalists and creolists have
focused on the instantiation of jargons in a single generation of
speakers but have given little attention to subsequent develop-
ments over a number of generations. As we have already seen
(§4), social conditions did not remain constant during the VOC
era. At some point in the process of cultural convergence, a
socially accepted (stable) grammar of the Cape Dutch vernacular
had to have emerged in the substratum among succeeding genera-
tions. That jargonized Dutch became a stable pidgin by virtue of
its expansion into the heterogeneous slave community and then
later creolized (Den Besten 1989:226, Davids 1991:44, 1994) is
entirely plausible in principle but exceedingly difficult to
ascertain. The degree to which a Khoikhoi Dutch jargon would
have stabilized into a pidgin would accordingly depend on the
intimacy of their linguistic encounters with slaves, about which
Mentzel (1785 [1921:49]) has left to posterity a tantalizing
clue: 'Since the arrival of the Europeans the inhabitants of
these kraals [the Khoikhoi] that were near the new settlement
greatly enriched their vocabulary by contact with the newcomers;
they learned still more from the slaves, and borrowed some [my
emphasis] of the so-called Portuguese, or more accurately, of the
lingua franca, common among all Eastern slaves'.
I proceed on the assumption that Mentzel's observation, is
fundamentally correct at face value; that is, the slaves and
Khoikhoi, sharing no common language, used Dutch as their primary
medium of intercommunication augmented by adlexification from
Creole Portuguese. By the early eighteenth century,
stabilization of Khoikhoi and slave jargons into a Cape Dutch
Pidgin occurred in the colonial service community, albeit with
regional and ethnic variation. In the trekboer regions Portu-
guese and Malay elements in the pidgin became diluted. One
should bear in mind that the Creole Portuguese of the slaves had
to have been rejargonized (i.e., subject to individual language
acquisition strategies) in the mouths of the Khoikhoi, for whom
it would have been entirely novel and more foreign than Dutch.
Traditional Khoikhoi with very restricted contact with slaves
probably did not speak the Pidgin but instead retained their
jargonized Dutch, whence the greater Khoikhoi element in modern
Orange River Afrikaans. This Cape Dutch Pidgin was transmitted
as a native language to the children of interethnic unions--most
notably the Bastaards and Free Blacks at opposite ends of the
colony, but surely within the slave population as well. The
first generations of creole speakers came into existence during
the period 1680-1750. The superstrate community did not speak
this Cape Dutch Pidgin, except possibly in the frontier areas.
The transient population of officials, sailors, and soldiers in
Cape Town could avail themselves of Portuguese, as could perhaps
other Company personnel with prior service in the Far East, many
of whom took their discharges at the Cape. Outside Cape Town
Europeans addressed their interlocutors of color in the dominant
language--Dutch--doubtless in foreigner talk registers.
The Cape Dutch Creole did not diverge as radically from
superstrate Dutch as 'true' creoles do from their lexifier
languages. The rate of dilution of Dutch across the entire
colony was far less extreme than in the case of Virgin Islands
Dutch Creole, where the language of slave masters was vastly more
remote from the majority of enslaved Africans. In the Cape
Colony continuous interaction between the inchoate social strata
afforded more opportunities for targeted language learning on the
part of individual slaves and acculturated Khoikhoi. This fact
alone all but assured an end product that would be much closer to
the metropolitan language than what we find in the Caribbean.
I am writing as if basilectal forms of Cape Dutch (those
furthest removed from the superstrate) were directly observable
for the period in question. The reality is that Dutch in the
mouths of slaves, Khoikhoi, and other people of color are very
sparsely attested before the mid nineteenth century. Den Besten
(1978, 1987b, 1988, 1993) has culled the evidence from the early
periods and has attributed the source for a number of Afrikaans
features to the erstwhile Cape Dutch jargons. Limitations of
space do not permit me to review his findings here. What I shall
I do instead is supplement them by reconsidering the linguistic
variables examined in the Duminy diary and showing how they might
have been realized in the Cape Dutch Creole.
The only other access to the Cape Dutch Creole ca. 1795 is
through internal reconstruction on the basis of contemporary
nonstandard varieties. To that end I should like to consider
some exemplary data from Orange River Afrikaans. Our point of
departure is the Afrikaans of the Griquas in the 1980s. The
Griquas are descendants of 'early Boer frontiersmen; of the
remnants of Khoisan tribes . . . ; of escaped slaves from the
wine and wheat farms of the south-west Cape; of free blacks from
the colony who could find no acceptable place for themselves in
it; and of African tribesmen' (Ross 1976:1); in short, their
Bastaard forebears were a creole community. Two methodological
precepts long accepted in traditional historical linguistics and
dialect geography are: (i) Older forms of language are preserved
in and are retrievable from its nonstandard varieties; and (ii)
the speech of nonmobile, less-well-educated, and especially older
speakers of these varieties will be conservative and least
affected by the normative influence of the standard language. I
shall assume that the Griqua form of Orange River Afrikaans has
been in a gradual but inexorable state of decreolization over the
course of the past two centuries. Nevertheless, a creole element
should be traceable in the speech of its oldest speakers as
recorded in the invaluable Afrikaans van die Griekwas van die
tagtigerjare (Van Rensburg, ed., 1984, hereinafter GA).
Among the first casualties during jargonization were gender
distinctions, personal agreement in the verb, the preterite,
periphrastic tenses with hebben and zijn, the weak forms of the
pronouns and daar, and the demonstratives deze, dit, and dat.
Possession was signalled by means of invariant se, which
could also be used to indicate close relationships other than
possession in the strict sense. The particle was juxtaposed to
the right of nouns, adjectives, and adverbs of time and place:
(6) a. En nou nou sal ik vir baas sê, die grootste ding ook
nog wat daar is, nou dié se Griekwas, nie die ou
Griekwase, die jong Griekwa, hulle is meer in die
Hollands (GA, 2.333).
b. Daai tyd se grootmense (GA, 2.306)
c. daarie se tyt (GA, 2.102)
d. Sôs die vroeër se noois gedra het (GA, 2.276)
e. die nou se kjêners is slim (GA, 2.81).
f. Die perskebome, die druiwebome--is maar goete wat
later se tyd eers ingekom het (GA, 2.321).
g. Watsie plek se nam, meneer, hierso, hier oner hierso
hierse skool? 'Nabeni', 'n kaffernam. 'n Groot skool.
Of course hierse in (6g) could be hierso. However, reduction of
the vowel in so would seem unusual, to say nothing of a possess-
ive function of the adverbial. The use of se in (6) is demon-
strably old, and we may impute it to the earlier Cape Dutch
Creole. Wikar (1779 [Mossop (ed.) 1935:66]) reported that the
Khoikhoi along the Orange River referred to the baboon as de oude
tyden zijn mens, which would be quite infelicitous in Dutch
As regards the personal pronouns, the form hy could refer
not only to masculine singular antecedents, but also to feminine,
inanimate, and plural entities.
(7) Die Gift al gedaan dood, wie kan hy meer wat schaden?
(Khoikhoi speaker, cited from Kolbe 1727:2.114)
'This/that poison has died, whom can it harm any more
even a little?'
Consider the following patterns in contemporary Orange River
(8) a. En dan kaptein Kok se vrou. Hy's mos nou ook 'n Griekwa
b. Daar is saad gewees wat die plant mee gesaai is, en as
hy groei, dan kom hulle lote, nou die blaaie
c. Hy [piesangs] kom baie skaars (GA, 2.322)
d. Hier's nie piesangs nie, baas. . . . Maar 'n mens
kan hom darem in die winkel koop, nè (GA, 2.321).
See further Links 1989:78-80. Overt pronominal marking for
plural referents was made possible by the expansion of the Dutch
pronoun hulle (etymologically oblique hun + lui) for both subject
and object. The pronominal system of the Cape Dutch Creole had
no formal means of distinguishing between subjective and oblique
cases in the first person plural. This was arguably the state of
affairs already in the stabilized Cape Dutch Pidgin:
(9) a. . . . waar om ons die goeds niet weder beitum en op
vretum (Khoikhoi speaker, cited from Kolbe 1727:2.66).
' . . . why don't we bite these/those things [lice]
back and eat [them] up?'
b. Ons soek kost hier, ons al gedaen wegloopen
(slave, ca. 1706, cited from Franken 1953:89).
'We seek food here; we have run away'.
Den Besten (1987b:passim) has culled the relevant data on this
feature from the eighteenth-century source material, and I refer
the interested reader to his study for a full discussion.
At first blush, it would appear that Cape Dutch Creole
possessive pronouns were formed by the placement of se to the
right of the corresponding personal pronoun (10). Upon closer
scrutiny, the history of this pattern is somewhat opaque.
(10) a. Maar daarie tyd toe hy se vader mos nou daai jare in
die Boere-oorlog gegaan het (GA, 2.294).
b. Nou dis nou ons se speletjie wat ons geleer speel
het (GA, 2.309).
c. Hulle se rokke dra hulle hiersô wys (GA, 2.275).
The emergence of a gender distinction in the third-person
singular pronoun is arguably secondary and due to the influence
of Euro-Cape Dutch. Rademeyer (1938:66) reported that 'onder die
Rehoboth-Basters word die bestilike sy en haar nou en dan gehoor,
dog die meer gebruiklike vorm is hy se' for both masculine and
feminine singular antecedents. A half century later, Links
(1989:82) would call attention to the existence of sy se for
'her' in Namaqualand (Sy se [=haar] man is aan die ploeg); but
this must be a late development given the absence of a parallel
form in other varieties of Orange River Afrikaans (cf. Rademeyer
Le Roux (1923:98) pointed out a syntactic parallelism
between the Orange River Afrikaans personal pronouns with se and
a Khoikhoi construction exemplified by ti di khóin = lit. 'ek se
vriende', sa di hab = lit. 'jy se perd'. Standing alone,
however, substratum transfer is inadequate as an explanation in
light of the fact that neither Rademeyer (1938:66) nor Links
(1989:82) found instances of ek se and jy/jou se. Sparse as they
are, our eighteenth-century attestations of pidginized Cape Dutch
do not flatly contradict Le Roux's hypothesis; but they do little
to support it, either. We should expect a developing system to
have considerable variation in forms standing for the same
concept. The second-person singular pronoun 'you' could appear
as jij, gij, je, and even the oblique form jou.
(11) a. . . . ik zoo lang zal by u blyven, tot jou Husing de
dubbeltjes betaalt hemme (Kolbe 1727:1.121-22).
'I shall remain with you [in your service] until you
have paid Husing the silver coins [i.e., money]'.
b. ons denk jou ook soo (slave, 1706, cited from Franken
1953:93) 'we think [ you (are), too]'14
Only jou is attested as the second-person singular possessive
pronoun; see example (19c), infra. In our older source material
the first-person singular pronoun 'I' is often the emphatic form
ikke, which would seem attributable to holophrastic speech during
the jargon phase; ik occurs, too, and both types are preserved in
Orange River Afrikaans (ik/ek, ekke). We find no direct evidence
at all for 'I' being expressed by oblique mij. In contrast to
the singular, the first-person plural form wij is unknown in the
fragments of pidginized Dutch that have come down to us from the
seventeenth and eighteenth centuries; 'we' is consistently
rendered by oblique ons. These same materials suggest that both
nonneuter and neuter forms of the first-person plural possessive
pronoun (respectively onze, ons) were available in pidginized
Dutch during this same perdiod: onse groote Kapiteyn (Ten Rhyne
1673 [1933:140]) beside ons bloed (Kolbe 1727:2.66). Although
their usage in these sentences is superficially consistent with
the rules of Dutch grammar, this is surely coincidence. The loss
of nominal gender can only mean that they were free variants.
My sense is that the use of se with personal pronouns was
originally limited to the anaphora (hy se, hulle se). This
innovation was creatively generated on analogy with the pos-
sessive/associative construction for nouns (supra), to which
anaphora refer. The first-person plural possessive onze, (which
occurred alongside of ons), was decomposed (ahistorically) into
two unbound morphemes, whence ons se. At some point in the
history of the Cape Dutch Creole, the plural paradigm was filled
out by the addition of a second-plural pronoun. As with the
first and third persons plural, julle is formally identical in
subjective and oblique environments. Projection of se into the
possessive of this pronoun, as we see in Orange River Afrikaans
(12) julle se skool het nôu gesluit (GA, 2.271),
produced a symmetrical and transparent plural inflection in the
Cape Dutch Creole pronominal system. A further step in the
expansion of se would be the first person plural possessive ons
se having become a forme de fondation for a first person singular
my se. However, Griqua Afrikaans offers but one very marginal
possibility of such a `founded form':
(13) Dis hierie hierie boetie van my se oorle vrou se skoonpa
gewees (GA, 2.345).
i.e., if not [boetie van my] se vrou.
Die was both the definite article and the sole demonstrative
pronoun ('this, that') in the Cape Dutch Creole. The focus of
the article could be sharpened by means of a preceding deictic
adverb (viz. hier, daar, and doer) if a given discourse called
for narrower specification of (respectively) proximal, distal, or
far distal location of the referent in relation to the speaker.
Eventually, this pragmatic combination would lexicalize into the
demonstratives we know today: hierdie, daardie, and nonstandard
doerdie (Roberge 1992). It is quite probable that the adverb
alone could function as an adnominal deictic element in the early
stages of the Cape Dutch Creole:
(14) a. doer onderste draad (lit. 'yonder lowest wire')
b. Dáár tyd toe't hy die skrywe, toe skryf sit my
Vader (GA, 2.39)
In his Travels in Southern Africa of 1815, Hinrich Lichten-
stein made the following cryptic comment on the Dutch of Khoikhoi
along the frontier: 'Farther, there are no auxiliary verbs; and
the Hottentots, even in speaking Dutch, do not know how to make
use of them. . . . The want of auxiliaries to express the time,
is often transferred by the Hottentots into the Dutch language'
(1815 [1930:2.467]). The accuracy of this observation for
Khoikhoi is of far less interest than the allusion to the omis-
sion of the tense auxiliaries hebben and zijn in their Dutch. In
the Cape Dutch Pidgin a preverbal particle ge, together with a
phonological variant ga, marked events that are situated in the
past. The fact that Afrikaans developed in a multilingual
contact situation raises the possibility of multilevel syn-
cretism, in which phonological, syntactic, and semantic proper-
ties of morphemes can be traced to multiple sources. The use of
ge/ga as a past tense marker closely corresponds to the Dutch
past participle prefix ge-. There is also evidence to suggest
that Khoikhoi preverbal preterital particles with a similar
canonical shape may have reinforced the observed usage; Nama gye,
go (Kroenlein 1889:101, 106); kò (recent past), kè (remote past)
(Hagman 1977:62). One cannot help but wonder whether Lichten-
stein stumbled onto precisely this feature, construing it as L1
We can be reasonably confident in this reconstruction on the
basis of what we find sporadically in Orange River Afrikaans:
(15) a. Nee, e, speel mos mar met hierie bal en sô an, mit die
voetbal en sô an en krieket ok gespeel (GA, 2.213)
b. Baas hulle hom geroep die galsbôom (GA, 2.267)
Although this morpheme is joined to the verb in standard ortho-
graphic practice, it appears to have been an unbound grammatical
formative in the Cape Dutch Creole. In (16a) the ge particle
precedes the lexical verb while the modal in the V2 position
remains in the present tense; in (16b), by contrast, ge is
attracted to the modal auxiliary.
(16) a. Ou Paul Krieër my bas, Paul Krieër. Dankie my bas.
Bas ek kjen hom nie gasiene, mar sy kop, ek had hom
oppe kop gasien (GA, 2.274)
b. Interviewer: Dan't julle, julle het seker mooi gelyk?
Informant: Ons het mooi gelyk, mooi gelyk. Ons het
sommer mooi gelyk, want dit was mooi gewerk ook. Die
ou Griekwatantes het mooi gewerk, en mammase en
oumase, wat nou gekjen daai broeke werk (GA, 2.306).
As a sidebar to the discussion of the Cape Dutch Creole past
tense form, it is worth recalling adjectives frequently become
creole verbs (Holm 1988:85). The use of the adjective dood
'dead' to mean 'die' is well known from the fragments of Khoikhoi
Pidgin Dutch recorded by Ten Rhyne (Mashy doot 'when I die', 1673
[1933:140]) and Kolbe (Die Gift al gedaan dood, 'this/that poison
has died', 1727:2.114)). That contemporary Orange River
Afrikaans also shows dood in place of sterf/sterwe suggests that
such conversions may have been at least marginally present in the
prior Cape Dutch Creole:
(17) Hierie plek Bultfontein sal ik nooi:t afgan nie.
As ek moet hier weggan, da dood ek (GA, 2.168).
This usage of dood is of course by no means unique to Orange
River Afrikaans (cf. WAT 2.245), but it cannot be a Netherland-
icism. It can hardly represent Dutch doden 'kill', which
requires a patient and is unknown in Afrikaans; see Den Besten
1987b:17-18 on its replacement with doodmaak. Note how the tense
particle ge/ga attaches itself to dood in (18):
(18) Die Griekwas het twintagduizend vee gehat. . . . En baie van
die vee het gevrék. Die Basoetoes het baie vannie vee
gesteel want hulle het oorie Drakensberg gekom. En toe't
hulle hier kom, die land s' gras dieselfde gras vannie
Vrystaat nie, toe dôod die vee, baie van die vee, beeste is
gadôod (GA, 2.40)
From the outset, all events in (18) occur at a single point in
time prior to the time of utterance. Their relation by means of
the narrative past is consistent with superstrate usage and is
what we should expect in a decreolizing variety. The qualitative
difference in grazing land between the Orange Free State and
Kokstad establishes a new time center. The tense nucleus
implied by the particle ga in gadood signals that the process of
dying occurred in the past; the copula refers to a complete
action by the time of speaking.
As concerns modality, I believe the irrealis marker of the
Cape Dutch Pidgin was kamma, which Nienaber (1963:373) has
derived from Khoikhoi //kamüh or =kamüh (cf. Nama =homi 'lie').
I take this to be the kam(m)e recorded by Kolbe (19) and Mentzel
(20) during the eighteenth century. Den Besten (e.g., 1986:217-
18, 223-24) thinks kam(m)e represents pidgin or creole Dutch, and
I am in agreement with him on this point. What I do not concur
with is his resolution of the form as 'be able to' (supposedly
from a Khoikhoi verb--cf. Nama //xáa--or a nasalless variant of
kan) plus a postvocalic variant -me of a 'Hottentot Dutch' ending
-um/-om/-me that is affixed chiefly to verbs but also adjectives
and nouns (1986:213-14, 1987a:88, 1987b:33-37 et passim).
Problems arise with the second constituent -me, for which there
is no clear semantic motivation. Our source material neither
affirms nor denies Den Besten's interpretation of kam(m)e.
Nonepistemic modality involving ability does make sense in
glossing the utterances that Mentzel and Kolbe have preserved for
posterity; but so too does expression of unrealized action or
lack of certainty.
(19) kam(m)e attributed to Khoikhoi:
a. Kamme niet verstaan (Kolbe 1727:1.504)
'won't/would not understand'
b. Ey Vrouw die Tovergoeds ja zoo bytum, ons ik kame niet
verdragen (Kolbe 1727:1.528).
'Ayee, woman, the medicine stings so, we [click] shall
not endure it'
c. Vrouw, jou Tovergoeds bra bytum, dat is waar, maar jou
Tovergoeds ook weer gezond makum, dat is ook waar.
Ons Tovermanns kame niet helpen, maar die Duits
Tovervrouw ja bra, die kame helpe (Kolbe 1727:1.528)
'Woman, your medicines sting very much, that is true,
but your medicines also make healthy again, that is
also true. Our medicine men will not help, but the
Dutch medicine woman is good, she will help'
(20) kam(m)e attributed to slaves:
a. Kammene Kumi, Kammene Kuli (Mentzel 1787 [1944:99])
Mentzel: 'If I have nothing to eat, I cannot work'
Analyzed: 'IRREALIS food-not, IRREALIS work-not'
b. Kammene Kas, Kammene Kunte (Mentzel 1787 [1944:99])
Mentzel: 'If you have no money, I have no ---'.
Analyzed: 'IRREALIS money-not, IRREALIS sex-not'
The use of kamma to give a nuance of pretense or ostensibility in
Orange River Afrikaans (21) and in other forms of spoken Afri-
kaans favors the irrealis interpretation of the kam(m)e attested
in our source material from the eighteenth century.
(21) a. Die klippe is bontes, wittes, dis kamma onse veëens.
Nou maak ons kraletjies, kraletjies, kraletjies. Nou
so maak ons dis kamma onse veëns wat ons inja in die
krale (GA, 2.310).
'The stones are colored ones, white ones, they are
our cattle. Now we make little kraals, kraals,
kraals. Now we do as if they were our cattle that
we herd into the kraals'
b. Dis maar sommer 'n klein poppie. . . . Maak hulle maar
net sit hulle kamma ook nou 'n kappietjie. . . .
Hulle maak hom kamma ook 'n kappietjie, so (GA, 2.311).
'But it is just a little doll. They just do as if they
were putting a little bonnet on. They also make for it
[the doll] a pretend bonnet'.
In the Cape Dutch Pidgin the basic--in the sense of prototypi-
cal--meaning of the irrealis marker is that the action of the
predicate is not (yet) part of reality (cf. Holm 1988:164-66).
The semantic range of this particle subsumed counterfactuality,
future time reference, and prediction.
Completion of an action in the Cape Dutch Pidgin appears to
have been expressed by (al) gedaan lit. '(already) done, finish-
ed' within the middle field before the main verb or adjective
(gedaan being the past participle of the Dutch verb doen 'do').
Den Besten (1987b:19-20, 22; 1989:238) cites (22a-c) in support
of this reconstruction, to which we can add (22d-e) in Orange
(22) a. Ons soek kost hier, ons al gedaen wegloopen . . .
(slave, 1706, cited from Franken 1953:89)
'We seek food here, we have run away'
b. de Clercq heeft gesegt jij mijn Cameraat gedaan vast
maken . . . (slave, 1720, cited from Franken 1953:50)
'De Klerk said you have tied up my comrade'
c. Die Gift al gedaan dood, wie kan hy meer wat schaden
(Khoikhoi speaker, cited from Kolbe 1727 2.114).
'This/that poison has died, whom can it harm any more
even a little?'
d. Klaar gesaai het, nou om-, nou as dit klaar omgeploe is
e. en lôp sê vir hulle . . . lat hulle hom regmak want
hulle-t al klaar gebetaal (GA, 2.222).
Expression of completive aspect by words meaning 'already',
'finished', or 'done' should hardly surprise us, given the
linguistic scene at the old Cape and what we find in pidgins
generally (cf. Tok Pisin pinis < English finish). As Den Besten
(1989:238) points out, Creole Portuguese ja 'already' and Malay
sudah 'finished' could indicate completion of an action in the
respective languages. In (22e), for example, a completed action
(payment) has resulted in a state (someone being made whole).
Durative aspect in the Cape Dutch Creole was marked by a
preverbal element derivable from the Dutch verb leggen 'lay', its
contrast with liggen having been neutralized (Afrikaans lê).
(23) Maar orwaat sou vullekie vanaand so lê hardloop?
'But why would the little foal be running so this evening?'
So far as I have been able to determine, this marker has largely
disappeared in contemporary Griqua Afrikaans. Preverbal loop
signalled the beginning stages of events, as we see in (24):
(24) As hy [die grond] eers loop nat word
In Orange River Afrikaans the verb-form loop shows a spectrum of
finely graded shades of meaning, ranging from its etymological
lexical meaning of 'go, walk' to a punctual, purely inchoative
function (cf. Du Plessis 1984:132-41).
Finally, another possible creolism sparsely attested in
contemporary Orange River Afrikaans is verb topicalization:
(25) Sit, ik sit lekker, lekker, lekker
It is not clear to me at this point whether (25) is a nonce form,
the result of general pragmatic principles (topic-comment order)
or is a vestige of some earlier grammaticalized pattern.
5.4. The linguistic items that defined Cape Dutch did not
fall into nonoverlapping domains. Rather, they were organized
into a continuum of lects in which the speech of individuals took
on superstrate or Cape Dutch Creole features--or avoided them--to
varying degrees. One group may have used one particular variant,
and another group the other. But we should expect to find mutual
exclusivity only when comparing the extremes of the sociolin-
guistic continuum. Between acrolectal Cape Dutch (Duminy) and
the basilect (Cape Dutch Creole) we should expect to find
linguistic forms that were subject to both social and stylistic
variation. The rise of a composite culture implied a convergence
of variables leading to compromise (mesolectal) varieties that
underlie today's forms of Afrikaans. These mesolects were
partially independent to the extent that the defining variables
have hybridized leading to new structures not found in either
acrolectal or basilectal Cape Dutch. Thus, mesolects cannot be
considered simply as composita of variants selected from the end
points of the continuum. I have elsewhere investigated two cases
of convergence leading to hybridization: the Afrikaans double
negation (Roberge 1991) and the verbal hendiadys (Roberge 1993).
I shall now very briefly discuss three additional cases that I
intend to treat elsewhere in detail.
5.4.1. During the process of convergence, the tense auxili-
aries hebben and zijn were reintroduced. But as the Cape Dutch
Creole moved closer to the superstrate, not all inflectional
categories were fully restored, even though their exponents
managed to survive. The result is a residue of allomorphs that
are used more or less interchangeably in Orange River
Afrikaans: het, 't (< Dutch heeft), had (< Dutch had):
(26) Die goue pondtjie bas? Ja, ek had hulle gakjen bas. Ek het
die tiensielings ok gakjen. (GA, 2.275)
Vestiges of zijn as a perfect auxiliary are also to be discerned
in Orange River Afrikaans:
(27) a. Die boere was báie laat hier gakom (GA, 2.40)
b. Ek moet so sê want êk is maar op hom [=plaas]
grôotgeword (GA, 2.219)
Merger of the semantic import of the creole temporal marker ge/ga
with that of the Netherlandic periphrastic perfect facilitated
additional hybridization. We find in Orange River Afrikaans the
use of both 'have' and 'be' as tense auxiliaries with the copula
(28) and in the expression 'to be born' (29):
(28) a. war dit baie gras gewees het, baie (GA, 2.138)
b. my anner seester se nam is gewees Fytjie (GA, 2.248).
(29) a. Ek het eintlik daar gabore (GA, 2.232).
b. die Van der Westhuise is mos hier in die Griekeland
gebore (GA, 2.282).
With the restoration of tense auxiliaries, the ge/ga particle
(30) a. En toe't hulle hier kom (GA, 2.40).
b. Nee bas, my ôupa kan ek'ie, wan dis toet
ek bore was, toe't my ôupa al voortsoorg gemak,
toe's hy weg (GA, 2.248).
Interesting, too, are pairings of the temporal particle ge/ga
with etymologically finite forms of the lexical verbs 'have'
(het) and 'be' (was) (31). These hybrids, which are manifesta-
tions of the creole pattern just described, also show up in
Orange River Afrikaans with one or both tense auxiliaries:
(31) a. mar oupa se pá was 'n leraart gawas (GA, 2.79)
b. Ennie oumênse het mos nou jaaare siek gawas
c. Ons het intlijk 'n plek da:r gehet (GA, 2.189)
In (31a, b) gewas (ge + preterital was) cannot be a phonological-
ly conditioned variant of standard past participle gewees. Were
this so, we should expect to find other instances of lowering,
monopthongization, and shortening, say *bas(te) for bees(te).
Given that convergence involved the progressive acquisition
of noncreole features by substrate speakers, and metropolitan
prestige norms were inaccessible to them, it follows that the
language of the permanent European settler community from the
late eighteenth century provided the linguistic model. Remnants
of a metropolitan feature such as the tense auxiliary zijn
alongside salient creolisms in Orange River Afrikaans strongly
indicate that the Cape Dutch of European speakers occupying the
same geographical space displayed greater affinity with the
language of the Duminy diary than with modern Afrikaans. This
would render untenable the conventional view that a more or less
uniform Cape Dutch vernacular strongly resembling modern Afri-
kaans had come into being between 1770 and 1800.
5.4.2. Save for some diachronic speculations in Slomanson's
(1993) analysis of infinitival complements in Orange River
Afrikaans, very little has been said about complementation in the
Cape Dutch Creole in current creolist literature. This may well
reflect the genuine paucity of data in our source material.
However, in the Orange River Afrikaans corpus complied by Van
Rensburg and his collaborators (1984), variation in the formal
signalling of complements provides some basis for internal
In colloquial forms of Afrikaans the conjunction lat can
introduce a sentential complement that would begin with dat in
the standard language. Le Roux (1910:107-8) took the former to
be an allegro variant resulting from different releases (lateral
versus oral) from a common point of occlusion. Since then, few
scholars have seriously questioned this view, even though are
reasonable grounds to do so; cf. Rademeyer 1938:53. Paardekooper
(1990) is rightly skeptical of an autochthonous phonological
development d- > l- in a single lexical item, but his own
suggestion of a Netherlandic dialectism is no less unlikely.
First, dat is unique in Afrikaans in showing an alternation
between dental and lateral onsets. Second and even more puzzling
is the existence of covariant laat; that is, with both the
lateral onset and a long vowel. This additional allomorph is
first reported in Le Roux (1910:107) but has generally slipped
beneath the radar, even though it can be heard in nonstandard
Afrikaans as late as 1965:
(32) a. laat di kaffers fer my moet fermoor
(S. J. du Toit, Di Koningin fan Skeba, 1898,
p. 15, cited from Le Roux 1910:107)
b. elke keer laat Poena die geld gekom haal het
(Adam Small, Kanna hy kô hystoe, 1965)
Le Roux (loc. cit.) attributed the long vowel to confusion with
the verb form laat (Dutch laten) 'cause (to be done), have
(done), let, allow'. From a purely synchronic viewpoint there is
no obvious grammatical or semantic motivation for secondary con-
tamination of the complementizer lat from an auxiliary verb.
Diachronically, however, in the context of intensive language
contact, Le Roux's intuition may not to have been unfounded.
Let us begin with one of Le Roux's example sentences
illustrating the usage of laat in place of dat.
(33) hoe kom het julle ni laat ons die goue goed uit
di klip uithaal ni (S. J. du Toit, Di Koningin fan Skeba,
1898, cited from Le Roux 1910:107).
'Why didn't you [pl.] have us remove the gold from the
If the matrix verb het is equivalent to wil hê (dat) 'want
something done', then a 'true' complementizer would be called
for. However, the WAT subentry for wil hê (4.119) does not
indicate the existence of a bare-stem variant. Yet, het in (33)
cannot be the tense auxiliary, which would preclude a comple-
mentizer. Under these circumstances laat would be interpretable
only as the causative verb, and its leftward dislocation would be
inexplicable. Ruling out these possibilities leaves 'to have/
order somebody to do something' as the only reasonable interpre-
tation of het in (33). If this reading makes sense, then laat
must again be seen as introducing a subordinate clause.
I want to make a case that this laat is the causative verb
functioning in a way somewhat comparable to serial 'say' meaning
'that' in Caribbean creoles (cf. Holm 1988:185-88). So far as I
have been able to determine, substrate influence does not seem to
have been a factor. I suspect this structure emerged during the
transition from less to more complex systems, when barriers to
mixing are greatest (Mühlhäusler 1986:127) and independent modes
of syntactic innovation come to the fore.
I proceed on the assumption that complementation in the
early Cape Dutch Pidgin was effected proponderantly by means of
the simple juxtaposition of sentences. Consider example (34)
from contemporary Orange River Afrikaans. Note in particular the
(34) Klaas, tel gou vir ons.
Ek het gehoor, baas, 0 die ou grootmense gesê het:
[informant counts in the Griqua language] (GA, 2.295)
The omissibility of the complementizer was doubtless retained in
the subsequent creole and from there percolated into acrolectal
forms of Cape Dutch. As structure was added to the convergent
creole grammar, a modal verb could be used to introduce a
quotation or a complement clause after matrix verbs the meaning
of which was epistemic, involving modes of knowledge or belief:
(35) Nou, kan hy Griekwa praat?
Ik weet nie sal hy kan praat nie (GA, 2.336).
'Now, can he speak Griqua? I don't know whether he
can speak [it] (lit. 'he will be able to')'.
If the matrix predicate involved either an imperative or deontic
modality (modes of obligation, necessity, commission), the modal
verb signalling the subordinate status of the following string
was laat. Thus, sentences of the type in (36) are taken to be
(36) a. nee, meneer du Toit moet nou fluks fertaal,
laat ons ferder kan hoor, en di Boesmans moet werk,
laat ons di gat ope grawe (S. J. du Toit, Di Koningin
fan Skeba, 1898, cited from Le Roux 1910:107).
'No, Mr. Du Toit must now translate energetically,
(so) that we can hear further, and the Bushmen must
work (so) that we can dig the hole open'.
b. Broer, jy moet lat ons lê troei hoeistoe want daar kom
'n verskriklike weer aan, opdat die weer ons nie beseet
nie (Rademeyer 1938:125). 'Brother, you must allow
[that] we head back home because a dreadful storm
is coming, so that the weather does not possess us'.
The sentences in (37) demonstrate the semantic proximity between
clauses introduced by hortative laat and etymological complement-
izers following an imperative or deontic modal in the matrix
(37) a. Wag dan, ou broer, laat ek hom eers betrag
'Wait, old brother, let me first look at it [the storm
b. Bring daai skaap lat ik hom slag (GA, 2.362).
'Bring that sheep (so) that I (can) slaughter it'.
Both (37a) and (37b) are purposive. Put another way, the
semantics of hortative laat 'let' and of a 'true' complementizer
'(so) that' share in common the speaker's expectation that the
following proposition will become reality.
Whether shortening of the vowel in serial laat was due to
lexical realignment with acrolectal dat or was part of a more
general phonological trend cannot be considered here. Whatever
the case, the weakened allomorph occurs for the first time in
1830, not coincidentally in the mouth of a person of color;
namely, C. E. Boniface's 'Hottentot' character Hendrik Kok:
(38) Probeer lat hy maar voor my een soopie geeuw
(De Zuid-Afrikaan, 13 August 1830)
'try let/that he gives me a tot (glass of liquor)'
i.e., 'Just have him offer me a tot (and I won't drink it)'
The nineteenth century saw an ongoing merger between serial
laat/lat and the acrolectal complementizer dat. By the 1880s,
there is syntactic evidence that serial laat/lat, which in the
stable Cape Dutch Pidgin was conditioned by imperatives and
deontic modals has gained acceptance in Euro-Afrikaans, where it
has become fully grammaticalized and associated with dat:
(39) Ek is jammer, lat hij ni kom ni (Mansvelt 1884 [1971:156].
In Orange River Afrikaans lat is fully grammaticalized as a
conjunctive element and functionally on par with dat.
(40) En die manne het met 'n nuus angekom dat hulle verbas
is om kom vind lat hier nog ê Griekwa-kjerk is (GA, 2.68).
Whether the earliest forms of Cape Dutch Creole maintained a
distinction between infinitival and sentential complements lies
beyond reconstruction. We do know that there must have been
considerable variability in infinitival complementation, all of
which is preserved in Orange River Afrikaans. At some point in
time, basilectal Cape Dutch Creole acquired a complementizer om
that corresponds to om . . . te in standard Afrikaans and in
(41) a. om die warheit sê (GA, 2.72)
b. Die daarop wôon en ok maar 'n plekkie het om sit
c. Cf. (40), supra:
hulle verbas is om kom vind . . .
Om was not available to introduce sentential complements,
incidentally, for it functioned as a causal conjunction in both
acrolectal and creolized forms of Cape Dutch. We can ascertain
these facts respectively from the usage of Duminy (42a) and from
that of the slave in Teenstra's zamenspraak (42b).
(42) a. om het een caapschee bul was
(Duminy diary, Franken [ed.] 1938:88)
b. hij niet spreek, om hij geen boodschap doen wil nie
(Teenstra 1830 [1943:240]).
cf. Modern Afrikaans:
c. Ek verlaat jou om jy niks vir my oorhet nie
Mesolectal forms of Cape Dutch added the particle te in various
configurations under the influence of acrolectal Cape Dutch. One
such configuration positions the particle te immediately to the
right of the complementizer om:
(43) a. Hulle was vorbas om tê Griekwa-mênse kry (GA, 2.68)
b. en pa vra vir oorlede oupa om te die pert dar die berg
in sit lat hy bietjie groei (GA, 2.138)
The question arises as to whether this type of complementation
involves attraction of the particle to om (generated in SPEC, CP;
cf. Slomanson 1993), or the actual fusion of morphemes into an
unanalyzable whole, as Rademeyer (1938:71) seemed to think; see
further the remarks of Du Plessis 1984:159-62. One factor that
lends support to the latter position is that infinitival comple-
ments introduced by om te in Orange River Afrikaans often show a
the particle te before the infinitive as well.
(44) Orange River Afrikaans:
a. Jy moes rontval om te pakkie kjers voor te betaal
b. Ai meneer, wee jy, om te jou die waarheid te sê
c. Die oumase en die mammase het ook die kappies so geleer
werk vir hulle om te op te sit (GA, 2.307)
a'. . . . om vir 'n pakkie kerse te betaal
b'. . . . om die waarheid vir jou te sê
c'. . . . om op te sit
If I am correct in the above reconstruction, convergence between
acrolectal Cape Dutch and the Cape Dutch Creole resulted in two
hybridizations. Basilectal complementation with om was made
formally more similar to its acrolectal counterpart through the
introduction of te first in tandem with om and then secondarily
into the position directly before the infinitive itself.
5.4.3. The distinction between predicate nominatives and
predicate adjectives was much less obvious in the Cape Dutch
Creole than in the acrolect. It is not clear whether the prior
pidgin possessed a verb corresponding to Dutch zijn in its equa-
tive function. The evidence, such as it exists, is in equipoise:
(45) a. Dat is doet (Ten Rhyne 1673 [1933:140])
'that is good (?)'
a. Ons denkum, ons altyd Baas, maar ons ja zienom,
Duytsman meer Baas (Kolbe 1727:1.477)
lit. 'We think we always master, but we indeed see
Dutchman greater master'.
However one interprets doet in the well-known datum from Ten
Rhyne (cf. Den Besten 1987b:17-18), the copula status of is would
seem unambiguous. Ignoring what is at present a moot issue in
regard to the pidgin, I tentatively impute to the Cape Dutch
Creole an equative construction with 'be' joining a subject and a
complement. The complement could be filled by adjectival
constituents, of course, and also by a presumably unrestricted
set of noun phrases.
(46) a. 'n Kleurling is darie lang hare (GA, 2.266)
b. Interviewer: En u is in die metodiste kerk?
Informant: Nee meneer, ek is ee Griekwa
Independente kerk. Di[t] is wat ek is (GA, 2.1047)
The mesolectal forms of Cape Dutch that gave rise to Afrikaans
lexicalized individual patterns (ek is jammer, ek is honger, ek
is dors, ek is spyt, ek is lus; see Donaldson 1993:188-89) but
did not adopt this usage wholesale.
In the course of the eighteenth century creolizing generations
drew on the resources of a fully developed Cape Dutch acrolect as
well as those of a stable Cape Dutch Pidgin. Creolization did
not proceed according to the textbook scenario described above
(§2), whereby a pidgin provides the input for L1 acquisition,
with minimal influence from (if not actual withdrawal of) the
superstrate language. Rather, sociolinguistic conditions at the
Cape were such that the input consisted of the superstrate in
close proximity of a coterritorial pidgin variety. In this way,
the resulting Cape Dutch Creole could show both strong metropoli-
tan characteristics alongside significant hybridization.
The interaction between social factors and glottogenesis are
much more complex than philological, variationist, interlect-
alist, and some creolist positions would have us believe. And
there seems to lurk the danger in renascent models of semicreoli-
zation that simplistic hypotheses are replaced by even simpler
and empirically less robust ones. If I may close by again
quoting Valkhoff (1966:231): 'It is not always either one thing
or another in the evolution of such a delicate social phenomenon
as speech or language'. In the history of Afrikaans it was not
always Dutch or substratum grammar, but three linguistic tradi-
tions--European, African (Khoikhoi), and Asian--that have met and
converged with one another to produce a new whole that is truly
more than the sum of its parts.
1. Only a handful of German and French loanwords in
Afrikaans are directly attributable to language contact at the
Cape during the seventeenth and eighteenth centuries (Raidt
1983:66-69). Lexical borrowing from Bantu languages and English
2. It is not my intention to resuscitate old controversies
here. I would refer the reader interested in historiography to
Reinecke 1937:563-81 and Nienaber 1949:96-141 for excellent
discussions of the early literature, and to the annotated biblio-
graphy of Reinecke et al. (1975:322-77). Valkhoff (1971:462-70),
Scholtz (1980:29-34), Raidt (1983:41-46, 1991:232-36), and
Makhudu (1984:11-25) all survey previous literature, albeit
through the filter of each author's particular point of view.
Happily, the old enthnocentrisms have largely subsided, and few
will mourn their passing; on the ideological dimension of our
subject see Roberge 1990.
The anglophone reader will find critical discussions of the
philological approach in Roberge 1986, Den Besten 1987a.
3. A few years later, Nienaber himself (1955) would at-
tribute the Afrikaans double negation to a 'foreigner' hybridi-
zation on the part of Khoikhoi speakers of Cape Dutch. Though
received sympathetically by Combrink (1978:83-85), Nienaber's
hypothesis has consistently failed to win acceptance by the
philologists (cf. Raidt 1983:189-90) because it is beyond direct
empirical verification. Den Besten (1978:40-42, 1985:32-35,
1986:210-24) has greatly revised and elaborated on the notion of
a Khoikhoi substratum origin for the Afrikaans double negation,
an idea that has always found favor among creolists (Valkhoff
1966:17; Holm 1988:174; 1989:343, 346; 1991).
4. One could naturally expand this list to include lin-
guists either influenced by Scholtz and/or Raidt or who have
conducted their diachronic investiations within a comparable
framework (e.g., Loubser 1961, Smuts 1969, Pheiffer 1980,
5. The variationist research program has resulted in the
compilation of an invaluable data base for Afrikaans of the
Griquas and the Richtersveld (Van Rensburg, ed., 1984, 1987). On
the Afrikaans of the Cape Malay, see Kotzé 1984 and Davids 1991.
6. I am aware of only one attempt at the latter, viz. Pone-
lis's (1991) history of Afrikaans phonology: 'Die uitgangspunt
vir die sosiostilistiese kontekstualisierung is dat Afrikaans uit
'n Hollandse koine stam (Scholtz 1980): 'n vorm van versorgde
(Amsterdamse) Hollands wat afgewyk het van sowel vernekulêre
Hollands as formele Vroeë Nuwe Nederlands' (p. 1).
7. The anglophone reader will find a précis of Hesseling's
views in Markey and Roberge (eds.) 1979.
8. In the first edition of his book Hesseling (1899:54-55)
alluded to the possibility of European children being exposed to
creolized Dutch through their aias (nursemaids). Years later
others would continue this thread; viz. Franken (1953:36-38),
Valkhoff (1966:176-77), and Van Marle (1978:61-63).
9. Makhudu (1984:3) absolves Markey of the charge of Euro-
centrism, and it is clear that the inadequacies of Markey's paper
are in part due to its programmatic nature and in part to a
superficial knowledge of Afrikaans. Compare the egregiously
misleading chapter on Afrikaans in Hutterer 1975:278-87, which
makes no mention of the 'Coloured' community and claims that
Afrikaans evolved solely from the nonstandard varieties of Dutch
imported to the Cape.
10. Den Besten (1989:228) is unconcerned with the specific
label (i.e., 'semicreole', 'creoloid', etc. versus his 'fort
creole') as it is 'nonsensical to occupy oneself with such
nitpicking discussions in the absence of a theoretically sound
typology of "new languages"'.
11. Similarly, Hesseling 1923:118-19, Le Roux 1923:88-98,
Rademeyer 1938:66-67, Valkhoff 1966:227-29, Links 1989:83. See
further Ponelis 1992 and 1993:225-47.
12. Ponelis (1993:230) reports that 'Western varieties of
Afrikaans have preserved the remnants of what must have been a
full set of periphrastic pronominal possessives', including ek
se, my se, jou se, sy se 'his/her', hom se. With the exception
of sy se 'her', these forms are not recorded in the descriptions
of either Rademeyer or Links (loc. cit.).
13. Ponelis (1993:32) reads Husing as the subject, jou as
14. I see the utterance in (11b) as a case of shallow
embedding and thus structurally different from jouw siecken hond
ghij die brood tecken (slave, 1671, cited from Franken 1953:47),
in which the first clause is comparable to Afrikaans jou ver-
brande skurk! (HAT, 494).
15. According to Den Besten (1987b:17-18), the temporal
adverbs strack (Dutch straks/strakjes 'presently, just now') and
soon function as future markers in the following sentence from
Ten Rhyne: Icke strack nae onse grote Kapiteyn toe, die man my
soon witte Boeba geme (1673 [1933:140]); consider also ik ja
strakjes voort lopum zoo (Kolbe 1727:1.121-22). This interpreta-
tion is supported by Franken's (1953:47) observation that logo
'soon' could similarly indicate the future in South African
Creole Portuguese. Whether the Cape Dutch Creole continued this
usage is another matter. If it did, I am inclined to think at
this point that it distinguished proximate (strack) from remote
Armstrong, James C. and Nigel A. Worden. 1989. 'The slaves,
1652-1834', in Elphick and Giliomee (eds.), 109-83.
Bender, M. Lionel. 1987. 'Some possible African creoles: A
pilot study', in Gilbert (ed.), 37-60.
Bickerton, Derek. 1979. 'Beginnings', in Hill (ed.), 1-22.
------. 1981. Roots of Language. Ann Arbor: Karoma.
------. 1984. 'The language bioprogram hypothesis', The
Behavioral and Brain Sciences 7: 173-88.
------. 1989. 'The lexical learning hypothesis and the pidgin-
creole cycle', in Pütz and Dirven (eds.), 11-31.
Boshoff, S. P. E. 1921. Volk en taal van Suid-Afrika.
Pretoria: De Bussy.
------. 1959. 'Die wonder van die ontstaan van Afrikaans', in
M. S. du Buisson (ed.), Die wonder van Afrikaans. Johannes-
burg: Voortrekkerpers, 25-35.
Bosman, D. B. 1923. Oor die ontstaan van Afrikaans. Amsterdam:
Swets & Zeitlinger.
------. 1947. 'Die ontstaan van Afrikaans', in C. M. van den
Heever and P. de Villiers Pienaar (eds.), Kultuurgeskiedenis
van die Afrikaner. Cape Town, Bloemfontein, and Port
Elizabeth: Nasionale Pers, 2: 1-22.
Bruyn, Adrienne and Tonjes Veenstra. 1993. 'The creolization of
Dutch', Journal of Pidgin and Creole Languages, 8: 29-80.
Buccini, Anthony F. 1992. 'The colonial Dutch dialects: New
evidence for the origin of Afrikaans and the development of
Standard Dutch in seventeenth-century Holland'. Paper
presented to the Berkeley-Michigan Germanic Linguistics
Roundtable, University of California at Berkeley, 10-11
Combrink, J. G. H. 1978. 'Afrikaans: Its origin and develop-
ment', in L. W. Lanham and K. P. Prinsloo (eds.), Language
and Communication Studies in South Africa. Oxford and Cape
Town: Oxford University Press, 69-95.
Conradie, C. J. 1981-82. 'Die ontwikkeling van die Afrikaanse
voltooide deelwoord'. Tydskrif vir Geesteswetenskappe,
21: 275-84, 22: 96-126.
Coseriù, Eugenio. 1978. Zur Entstehung der romanischen Spra
chen. Darmstadt: Wissenschaftliche Buchgesellschaft.
Davids, Achmat. 1991. 'The Afrikaans of the Cape Muslims from
1815 to 1915'. M.A. thesis, University of Natal, Durban.
------. 1994. 'Afrikaans--die produk van akkulturasie', in
Olivier and Coetzee (eds.), 110-19.
den Besten, Hans. 1978. 'Cases of possible syntactic
interference in the development of Afrikaans', in Muysken
------. 1985. 'Die doppelte Negation im Afrikaans und ihre
Herkunft', in Norbert Boretzky, Werner Enninger, and Thomas
Stolz (eds.), Akten des 1. Essener Kolloquiums über 'Kreol-
sprachen und Sprachkontakte' vom 26.1.1985 an der Universi-
tät Essen. Bochum: Studienverlag Dr. N. Brockmeyer, 9-42.
------. 1986. 'Double negation and the genesis of Afrikaans', in
Pieter Muysken and Norval Smith (eds.) Substrata versus
Universals in Creole Genesis. (Creole Language Library,
1.) Amsterdam: John Benjamins, 185-230.
------. 1987a. Review article of Raidt 1983. Journal of Pidgin
and Creole Languages 2: 67-92.
------. 1987b. 'Die niederländischen Pidgins der alten Kap-
kolonie', in Norbert Boretzky, Werner Enninger, and Thomas
Stolz (eds.), Beiträge zum 3. Essener Kolloquium über
Sprachwandel und seine bestimmenden Faktoren. Bochum:
Studienverlag Dr. N. Brockmeyer, 9-40.
------. 1988. 'Universal-Grammatik und/oder Zweitsprachen-
erwerb: Der Fall Afrikaans', in Norbert Boretzky, Werner
Enninger, and Thomas Stolz (eds.), Beiträge zum 4. Essener
Kolloquium über 'Sprachkontakt, Sprachwandel, Sprachwechsel,
Sprachtod'. Bochum: Studienverlag Dr. N. Brockmeyer,
------. 1989. 'From Khoekhoe foreignertalk via Hottentot Dutch to
Afrikaans: The creation of a novel grammar', in Pütz and
Dirven (eds.), 207-54.
------. 1993. 'The Khoekhoe basis for the Afrikaans associative
construction'. Paper presented to the Society for Pidgin
and Creole Linguistics Conference, University of Amsterdam,
10-12 June 1993.
Donaldson, Bruce C. 1993. A Grammar of Afrikaans. (Mouton
Grammar Library, 8.) Berlin and New York: Mouton de
du Plessis, H. G. W. 1984. 'Sintaktiese variasie in die Afri-
kaans van die Griekwas', in Van Rensburg (ed.), 1: 37-195.
------. 1994. 'Taalkontakvariasie in Afrikaans', in Olivier and
Coetzee (eds.), 120-29.
du Toit, André and Hermann Giliomee. 1983. Afrikaner Political
Thought: Analysis and Documents. Vol. 1. (Perspectives in
Southern Africa, 22.) Berkeley, Los Angeles, and London:
University of California Press.
du Toit, P. J. 1905. Afrikaansche studies. (Diss. Ghent.)
Ghent: A. Siffer.
Elphick, Richard. 1977. Kraal and Castle. Khoikhoi and the
Founding of White South Africa. (Yale Historical Publica-
tions, Miscellany, 116.) New Haven and London: Yale
Elphick, Richard and Hermann Giliomee. 1989. 'The origins and
entrenchment of European dominance at the Cape, 1652--c.
1840', in Elphick and Giliomee (eds.), 521-66.
Elphick, Richard and V. C. Malherbe. 1989. 'The Khoisan to
1828', in Elphick and Giliomee (eds.), 3-65.
Elphick, Richard and Robert Shell. 1989. 'Intergroup relations:
Khoikhoi, settlers, slaves and free blacks, 1652-1795', in
Elphick and Giliomee (eds.), 184-239.
Elphick, Richard and Hermann Giliomee (eds.). 1989. The Shaping
of South African Society, 1652-1840. 2nd. ed. Middletown,
CT: Wesleyan University Press.
Fishman, Joshua A. 1987. 'Post-exilic Jewish languages and
pidgins/creoles: Two mutually clarifying perspectives',
Multilingua 6: 7-24.
Franken, J. L. M. 1953. Taalhistoriese bydraes. Amsterdam and
Cape Town: Balkema.
Franken, J. L. M. (ed.). 1938. Duminy-dagboeke. (Van Riebeeck
Society, 19.) Cape Town: Van Riebeeck Society.
Gilbert, Glenn G. 1993a. 'Popular Brazilian Portuguese: A
convergence creole, derived from a dual source'. Paper
presented to the Society for Pidgin and Creole Linguistics
Conference, Los Angeles, 10-12 January 1993.
------. 1993b. 'Adverbial placement in acrolectal convergence
creoles'. Paper presented to the Society for Pidgin and
Creole Linguistics Conference, University of Amsterdam,
10-12 June 1993.
Gilbert, Glenn G. (ed.) 1987. Pidgin and Creole Languages:
Essays in Memory of John E. Reinecke. Honolulu: University
of Hawaii Press.
Grüner, Rolf. 1982. 'Die invloed van Duits op vroeë Afrikaans--
'n sosiohistoriese perspektief', Acta Academica 15: 33-48.
Hagman, Roy S. 1977. Nama Hottentot Grammar. (Indiana Univer-
sity Publications, Language Science Monographs, 15.)
Bloomington: Indiana University.
Hancock, Ian F. 1971. 'A survey of the pidgins and creoles of the
world', in Hymes (ed.), 509-23.
Haugen, Einar. 1976. The Scandinavian Languages. Cambridge,
MA: Harvard University Press.
Hesseling, D. C. 1899. Het Afrikaansch. Leiden: E. J. Brill.
------. 1923. Het Afrikaans. 2nd ed. Leiden: E. J. Brill.
Hill, Kenneth C. (ed.). 1979. The Genesis of Language. Ann
Hoijer, Harry. 1948. 'Linguistic and cultural change', Language
Holm, John. 1988-89. Pidgins and Creoles. 2 vols. Cambridge:
Cambridge University Press.
------. 1991. 'American Black English and Afrikaans: Two
Germanic semi-creoles'. Paper presented to the Society for
Pidgin and Creole Linguistics Conference, Chicago, January
Hutterer, Claus Jürgen. 1975. Die germanischen Sprachen.
Budapest: Adadémiai Kiadó.
Hymes, Dell (ed.). 1971. Pidginization and Creolization of
Languages. Cambridge: Cambridge University Press.
Kainz, Friedrich. 1943. Psychologie der Sprache. Vol. 2.
Stuttgart: Ferdinand Enke Verlag.
Katzen, M. F. 1982. 'White settlers and the origin of a new
society, 1652-1778', in Monica Wilson and Leonard Thompson
(eds.), A History of South Africa to 1870. Cape Town:
David Philip, 70-93.
Kloeke, G. G. 1950. Herkomst en groei van het Afrikaans.
Leiden: Universitaire Pers.
Kloss, Heinz. 1978. Die Entwicklung neuer germanischer
Kultursprachen seit 1800. 2nd ed. (Sprache der Gegenwart,
37.) Düsseldorf: Pädagogischer Verlag Schwann.
Kolbe, Peter. 1727. Peter Kolbe. Naaukeurige en uitvoerige
beschryving van de Kaap de Goede Hoop. Amsterdam.
Kotzé, E. F. 1984. 'Afrikaans in die Maleierbuurt: 'n diachro
niese perspektief', Tydskrif vir Geesteswetenskappe 24:
------. 1989. 'How creoloid can you be? Aspects of Malay
Afrikaans', in Pütz and Dirven (eds.), 251-64.
------. 1991. 'Afrikaans in the context of creole linguistics--
dismantling linguistic apartheid'. Paper presented to the
Society for Pidgin and Creole Linguistics Conference,
Chicago, January 1991.
------. 1994. 'Afsonderlik ontwikkel deur verbastering?--
Afrikaans en Middelnederduits in die kreoolse konteks', in
Olivier and Coetzee (eds.), 130-38.
Kroenlein, J. G. 1889. Wortschatz der Khoi-khoin (Namaqua-
Hottentotten). Berlin: Deutsche Kolonialgesellschaft.
Kruisinga, E. 1906. 'De oorsprong van het Afrikaans', Taal en
Letteren 16: 417-39.
Lass, Roger. 1987. The Shape of English. London and
Melbourne: J. M. Dent & Sons.
------. 1990a. 'How to do things with junk: Exaptation in
language evolution', Journal of Linguistics 26: 79-102.
------. 1990b. 'Where do extraterritorial Englishes come from?',
in S. Adamson et al. (eds.), Papers from the Fifth Inter-
national Conference on English Historical Linguistics.
Amsterdam: John Benjamins, 245-80.
Le Page, Robert. 1987. 'The need for a multidimensional model',
in Gilbert (ed.), 113-29.
Le Roux, J. J. 1923. Oor die Afrikaanse sintaksis. (Diss.
Utrecht.) Amsterdam: Swets & Zeitlinger.
Le Roux, T. H. 1910. Beschrijvende klankleer van het Afrikaans.
(Diss. Leiden.) Leiden: A. W. Sijthoff.
Lichtenstein, Hinrich. 1812-15 [1928-30]. Travels in Southern
Africa in the Years 1803, 1804, 1805 and 1806. Translated
by Anne Plumptre. (Van Riebeeck Society, 10-11.) Cape
Town: Van Riebeeck Society.
Links, Tony. 1989. So praat ons Namakwalanders. Cape Town:
Lockwood, W. B. 1965. An Informal History of the German Lan
guage, with Chapters on Dutch and Afrikaans, Frisian and
Yiddish. Cambridge: W. Heffer.
Loubser, J. E. 1961. Die saamgestelde verbale vorm van Neder-
lands na Afrikaans. Groningen: J. B. Wolters.
Louden, Mark L. 1993. 'Is Yiddish a creole language?' Paper
presented to the Society for Pidgin and Creole Linguistics
Conference, University of Amsterdam, 10-12 June 1993.
Louw, S. A. 1948. Dialekvermenging en taalontwikkeling: Proewe
van Afrikaanse taalgeografie. Amsterdam: North Holland.
Lubbe, H. J. 1974. 'Valkhoff en die ontstaan van Afrikaans,
veral n.a.v. sy New Light on Afrikaans and "Malayo-Por-
tuguese"', Tydskrif vir Geesteswetenskappe 14: 89-99.
Makhudu, Dennis P. 1984. 'Is Afrikaans a creole language?' MA
thesis, Southern Illinois University. Ann Arbor: Uni-
versity Microfilms International.
Mansvelt, N. 1884. Proeve van een Kaapsch-Hollandsch idioticon.
Utrecht: A. J. van Huffel. Rpt. in H. J. J. M. van der
Merwe (ed.) 1971. Vroeë Afrikaanse woordelyste. (Patriot-
vereniging vir Afrikaanse teksuitgawes, n.s. 5.) Pretoria:
J. L. van Schaik.
Markey, T. L. 1981. 'Diffusion, fusion and creolization: A
field guide to developmental linguistics', Papiere zur Lin-
guistik 24: 3-37.
------. 1982. 'Afrikaans: Creole or non-creole?', Zeitschrift
für Dialektologie und Linguistik 49: 169-207.
Markey, T. L. and Paul T. Roberge (eds.). 1979. On the Origin
and Formation of Creoles: A Miscellany of Articles by Dirk
Christiaan Hesseling. Introduction by Pieter Muysken and
Guus Meijer. (Linguistica Extranea. Studia, 4.) Ann
Mentzel, O. F. 1785-87 [1921-44]. A Geographical and Topographi-
cal Description of the Cape of Good Hope. Ed. by H. J. Man-
delbrote. (Van Riebeeck Society, 4, 6, 25.) Cape Town:
Van Riebeeck Society.
Meshtrie, Rajend. 1992. English in Language Shift: The History,
Structure, and Sociolinguistics of South African Indian
English. Cape Town: Cape Town University Press.
Mossop, E. E. (ed.). 1935. The Journals of Wikar, Coetsé and Van
Reenen. (Van Riebeeck Society, 15.) Cape Town: Van
Mühlhäusler, Peter. 1974. Pidginization and Simplification of
Language. (Pacific Linguistics, B/26.). Canberra:
Australian National University.
------. 1986. Pidgin and Creole Linguistics. (Language in
Society, 11.) Oxford and New York: Basil Blackwell.
Muysken, Pieter (ed.) 1978. Amsterdam Creole Studies II.
(Publikaties van het Instituut voor Algemene Taalwetenschap,
20.) Amsterdam: University of Amsterdam.
Nienaber, G. S. 1934. Oor die Afrikaanse taal. Amsterdam:
Swets & Zeitlinger.
------. 1949-53. Oor Afrikaans. 2 vols. Johannesburg:
------. 1950. Taalkundige belangstelling in Afrikaans tot 1900.
Johannesburg: Afrikaanse Pers-Boekhandel.
------. 1955. 'Iets naders oor die ontkenning in Afrikaans',
Hertzog-Annale 2: 29-45.
------. 1963. Hottentots. Pretoria: J. L. van Schaik.
------. 1994. 'Die ontstaan van Khoekhoe Afrikaans', in Olivier
and Coetzee (eds.), 139-52.
Olivier, Gerrit and Anna Coetzee (eds.). 1994. Nuwe perspektiewe
op die geskiedenis van Afrikaans opgedra aan Edith H. Raidt.
Johannesburg: Southern Books.
Paardekooper, P. C. 1990. 'Een voegwoordraadsel: Lat "dat"'.
Suid-Afrikaanse Tydskrif vir Taalkunde 8: 174-77.
Pauwels, J. L. H. 1958. Het dialect van Aarschot en omstreken.
Brussels: Belgisch Interuniversitair Centrum voor
Pheiffer, R. H. 1980. Die gebroke Nederlands van Franssprekendes
aan die Kaap in die eerste helfte van die agtiende eeu.
Cape Town, Pretoria, and Johannesburg: Academica.
Platt, J. T. 1975. 'The Singapore English speech continuum and
its basilect "Singlesh" as a "creoloid"', Anthropological
Linguistics 17: 363-74.
Polomé, Edgar. 1983. 'Creolization and language change', in
Woolford and Washabaugh (eds.), 126-36.
Ponelis, F. A. 1987. 'Uit die geskiedenis van die Afrikaanse
betreklike konstruksie'. (Suid-Afrikaanse Tydskrif vir
Taalkunde, geleentheidsuitgawe, 5.) Linguistic Society of
------. 1988. 'Afrikaans en taalversteuring', in Tydskrif vir
Geesteswetenskappe 28: 119-49.
------. 1991. Historiese klankleer van Afrikaans. (Universiteit
van Stellenbosch. Annale, 1990/1.) Stellenbosch: Univer-
sity of Stellenbosch.
------. 1992. 'Die ontwikkeling van die possessief in
Afrikaans', in H. Bennis and J. W. de Vries (eds.), De
binnenbouw van het Nederlands. Een bundel artikelen voor
Piet Paardekooper. Dordrecht: ICG Publications, 273-90.
------. 1993. The Development of Afrikaans. (Duisburger Arbeiten
zur Sprach- und Kulturwissenschaft, 18.) Frankfurt a.M.:
Pütz, Martin and René Dirven (eds.). 1989. Wheels within
Wheels: Papers of the Duisburg Symposium on Pidgin and
Creole Languages. (Duisburger Arbeiten zur Sprach- und
Kulturwissenschaft, 5.) Frankfurt a.M.: Peter Lang.
Rademeyer, J. H. 1938. Kleurling-Afrikaans: Die taal van die
Griekwas en Rehoboth-Basters. Amsterdam: Swets & Zeit-
Raidt, Edith H. 1974. 'Nederlandse en Kaapse spreektaal in die
17de en 18de eeu', in F. F. Odendal (ed.), Taalkunde--'n
lewe. Studies opgedra aan W. Kempen. Cape Town: Tafelberg,
------. 1975. 'Nuwe aktualiteit van 'n ou polemiek', in
Afrikaans--Dit is ons erns. Pretoria: Suid-Afrikaanse
Akademie vir Wetenskap en Kuns, 39-54.
------. 1976a. 'Die herkoms van objekskonstruksies met vir', in
1875-1975; Studies oor die Afrikaanse taal. Johannesburg:
------. 1976b. 'Linguistische und soziologische Faktoren im
Sprachwandel des Afrikaans im 18. Jahrhundert', in Leonard
Forster and Hans-Gert Roloff (eds.), Akten des V. Interna-
tionalen Germanistenkongresses. (= Jahrbuch für interna-
tionale Germanistik, A/2.) Bern: Herbert Lang, 2.2: 155-64.
------. 1977. 'Afrikaans and "Malayo-Portuguese": Light and
Shadow', African Studies 36: 70-78.
------. 1978. 'Enkele aspekte van taalverandering', in H. Snyman
(ed.), Uit vier windstreke. Studies opgedra aan prof. dr.
Meyer de Villiers. Cape Town: Nasou, 104-19.
------. 1980. 'Die Reduplikation im Afrikaans', in Heinz Rupp
and Hans-Gert Roloff (eds.), Akten des VI. Internationalen
Germanistenkongresses Basel 1980. (= Jahrbuch für interna-
tionale Germanistik, A/8.) Bern, Frankfurt a.M., and Las
Vegas: Peter Lang, 2: 494-501.
------. 1981. 'Oor die herkoms van die Afrikaanse reduplikasie',
in A. J. Coetzee (ed.), Hulsels van Kristal: Bundel aan-
gebied aan Ernst van Heerden. Cape Town: Tafelberg, 178-89.
------. 1983. Einführung in Geschichte und Struktur des Afri
kaans. Darmstadt: Wissenschaftliche Buchgesellschaft.
------. 1984a. 'Interne ontwikkeling van Afrikaans', in T. J. R.
Botha (ed.), Inleiding tot die Afrikaanse taalkunde.
Pretoria and Cape Town: Academica, 52-82.
------. 1984b. 'Vrouetaal en taalverandering', Tydskrif vir
Geesteswetenskappe 24: 256-86.
------. 1991. Afrikaans en sy Europese verlede. 3rd ed. Cape
Reinecke, John E. 1937. 'Marginal languages'. Ph.D. disserta
tion, Yale University. Ann Arbor: University Microfilms
Reinecke, John E., Stanley M. Tsuzaki, David DeCamp, Ian F. Han-
cock, and Richard E. Wood. 1975. A Bibliography of Pidgin
and Creole Languages (Oceanic Linguistics Special Publica-
tions, 14.) Honolulu: University of Hawaii Press.
Roberge, Paul T. 1985. 'Grammatical prerequisites to
phonological change?', Zeitschrift für Dialektologie und
Linguistik 52: 188-217.
------. 1986. 'What's happening in Afrikaans historical lin-
guistics'? [review article of Raidt 1983], Zeitschrift für
Dialektologie und Linguistik 53: 194-202.
------. 1987. 'Phonomorphological indeterminacy', in Wolfgang
U. Dressler, Hans C. Luschützky, Oskar E. Pfeiffer and John
Rennison (eds.), Phonologica 1984. Cambridge: Cambridge
University Press, 267-72.
------. 1990. 'The ideological profile of Afrikaans historical
linguistics', in John E. Joseph and Talbot J. Taylor (eds.),
Ideologies of Language. London and New York: Routledge,
------. 1991. 'Etymological opacity, hybridization, and the
Afrikaans double negation'. Paper presented to the Tenth
International Conference on Historical Linguistics, Free
University of Amsterdam, 12-16 August 1991.
------. 1992. On the origin of the Afrikaans demonstrative
pronouns hierdie and daardie. Paper presented to the
Berkeley-Michigan Germanic Linguistics Roundtable, Univer-
sity of California at Berkeley, 10-11 April 1992.
------. 1993. On the origins of the Afrikaans verbal hendiadys.
Paper presented to the Society for Pidgin and Creole
Linguistics Conference, University of Amsterdam, 10-12 June
------. 1994. 'On detecting a prior linguistic continuum in Cape
Dutch', in Olivier and Coetzee (eds.), 153-65.
Romaine, Suzanne. 1988. Pidgin and Creole Languages. London and
New York: Longman.
Ross, Robert. 1976. Adam Kok's Griquas: A Study in the Develop-
ment of Stratification in South Africa. (African Studies
Series, 21.) Cambridge: Cambridge Unviersity Press.
Sankoff, Gillian. 1979. 'The genesis of a language', in Hill
Scholtz. J. du Plessis. 1963. Taalhistoriese opstelle.
Pretoria: J. L. van Schaik.
------. 1965. Afrikaans uit die vroeë tyd. Pretoria: J. L. van
------. 1972. Afrikaans-Hollands in die agtiende eeu. Cape
------. 1980. Wording en ontwikkeling van Afrikaans. Cape
Schonken, F. T. 1914. De oorsprong der Kaapsch-Hollandsche
volksoverleveringen. Amsterdam: Swets & Zeitlinger.
Siegel, Jeff. 1985. 'Koines and koineization', Language and
Society 14: 357-78.
Slomanson, Peter. 1993. 'On a problem in the grammar of Orange
River Afrikaans'. Paper presented to the Society for Pidgin
and Creole Linguistics Conference, University of Amsterdam,
10-12 June 1993.
Smith J. J. 1927. 'The evolution and recognition of the Afri-
kaans language', in Official Year Book of the Union 8
------. 1952. 'Theories about the origin of Afrikaans'.
Johannesburg: Witwatersrand University Press.
Smuts, J. 1969. Vroeë Afrikaans. (Afrikaanse taalmonografieë,
4). Cape Town: Nasou.
Teenstra, M. D. 1830 . De vruchten mijner werkzaamheden.
Ed. by F. C. L. Bosman. (Van Riebeeck Society, 24.) Cape
Town: Van Riebeeck Society.
ten Rhyne, Willem. 1673 . Schediasma de Promontorio Bonae
Spei. in I. Schapera and B. Farrington (eds.), The Early
Cape Hottentots. (Van Riebeeck Society, 14.) Cape Town:
Van Riebeeck Society.
Thomason, Sarah Grey and Terrence Kaufman. 1988. Language
Contact, Creolization, and Genetic Linguistics. Berkeley,
Los Angeles, London: University of California Press.
Trudgill, Peter. 1978. 'Creolization in reverse', Transactions
of the Philological Society, 1976-77: 32-50.
Valkhoff, Marius F. 1966. Studies in Portuguese and Creole. With
Special Reference to South Africa. Johannesburg: Wit-
watersrand University Press.
------. 1971. 'A descriptive bibliography of the linguistics of
Afrikaans', in Thomas A. Sebeok (ed.), Current Trends in
Linguistics. The Hague and Paris: Mouton, 7: 455-500.
------. 1972. New Light on Afrikaans and "Malayo-Portuguese."
Louvain: Éditions Peeters Imprimerie Orientaliste.
van der Merwe, Elda. 1993. 'Orange River Afrikaans: Post-creole
or not?' Paper presented to the Society for Pidgin and
Creole Linguistics Conference, University of Amsterdam,
10-12 June 1993.
van der Merwe, H. J. J. M. 1963. 'Die ontstaan van Afrikaans',
in H. J. J. M. van der Merwe (ed.), Inleiding tot die
taalkunde. Pretoria: J. L. van Schaik, 67-88.
------. 1964. 'Ontstaansteorieë oor Afrikaans', in H. J. J. M.
van der Merwe (ed.), Studierigtings in die taalkunde.
Pretoria: J. L. van Schaik, 135-55.
------. 1966. 'Afrikaans 'n basterkind van Hollands-Kreools-
Portugese fornikasie aan die Kaap!?', Tydskrif vir Geestes-
wetenskappe 6: 377-94.
------. 1968. Die ontstaan van Afrikaans. In H. J. J. M. van
der Merwe (ed.) Afrikaans--sy aard en ontwikkeling.
Pretoria: J. L. van Schaik, 15-66.
------. 1970. Herkoms, en ontwikkeling van Afrikaans. Johannes
burg: Afrikaanse Pers.
van Ginneken, J. J. A. 1913. Handboek der Nederlandsche taal.
van Marle, Jaap. 1978. 'A case of affix generalization in
Afrikaans', in Muysken (ed.), 57-67.
van Rensburg, M. C. J. 1983. 'Nie-standaardvorme, variasiepa-
trone en Afrikaans uit die vorige eeu', in G. N. Claassen
and M. C. J. van Rensburg (eds.), Taalverskeidenheid. 'n
Blik op die spektrum van taalvariasie in Afrikaans,
Pretoria, Cape Town, and Johannesburg: Academica, 134-61.
-------. 1984. 'Oranjerivier-Afrikaans', in Van Rensburg
(ed.) 1 (2): 513-18.
-----. 1985. 'Taalverskeidenheid, taalversteuring en Afrikaans',
in H. J. Lubbe (ed.), Fokus op die taalkunde. (= Acta
Academica, Nuwe Reeks B.) Bloemfontein: University of the
Orange Free State, 123-66.
------. 1989. 'Orange River Afrikaans: A stage in the pidgin/
creole cycle', in Pütz and Dirven (eds.), 135-51.
------. 1994. 'Die ontstaan van Afrikaans in 'n intertaal
konteks', in Olivier and Coetzee (eds.), 166-79.
van Rensburg, M. C. J. (project director). 1984. 'Die Afrikaans
van die Griekwas van die tagtigerjare'. Bloemfontein: Uni-
versity of the Orange Free State.
van Rensburg, M. C. J. (project director). 1987. 'Finale verslag
van die groter navorsingsprojek: Gesproke Afrikaans: Die
Afrikaans van die Richtersveld'. Bloemfontein: University
of the Orange Free State.
Washabaugh, William and Sidney M. Greenfield. 1983. 'The
development of Atlantic creole languages', in Woolford and
Washabaugh (eds.), 106-19.
Webb, V. N. 1993. 'Die herkoms van Oranjerivierafrikaans', in
Linguistica. Festschrift E. B. van Wyk. Pretoria:
J. L. van Schaik [in press].
Whinnom, Keith. 1971. 'Linguistic hybridization and the "special
case" of pidgins and creoles', in Hymes (ed.), 91-115.
Wittmann, Johannes. 1928. 'Die sprachgeschichtliche Entwicklung
des Deflexionstypus in Afrikaans'. (Diss. Bonn.) Bonn:
Woolford, Ellen. 1983. 'The social context of creolization',
in Woolford and Washabaugh (eds.), 1-9.
Woolford, Ellen and William Washabaugh (eds.). 1983. The Social
Context of Creolization. Ann Arbor: Karoma Publishers.
Zimmer, Stefan. 1992. 'On language genesis: The case of
Afrikaans', in Perspectives on Indo-European Language,
Culture and Religion. Studies in Honor of Edgar C. Polomé.
(Journal of Indo-European Studies, Monograph no. 9.)
McLean, VA: Society of Man, 2: 347-59.