Atlas of Pidgin and Creole Language Structures Online

1. Introduction ⇫ ¶

Fanakalo (also spelled Fanagalo) is a southern African pidgin that continues to be used two centuries after its inception. It originated in South Africa, and is also spoken in Zimbabwe, and Zambia (where it is usually known as Chilapalapa), Mozambique, Malawi, and Namibia. It was carried to these countries by migrant workers to the South African mines and by white South African farmers who immigrated to these countries in the early part of the 20th century. Fanakalo is a “crystallized” pidgin in terms of its fairly stable structure and circumscribed contexts of use. It is a contact language used prototypically in work situations: on farms, on the mines of the Witwatersrand, which draws a multilingual workforce from all over southern Africa, in other urban labour and trade situations, and in domestic employment (between employers and maids, cooks, gardeners). Within South Africa it occurs mainly in the provinces of KwaZulu-Natal and Gauteng (the mining area). Currently one can hear it in situations of sustained labour contacts, as well as in “transactional” communications in petrol stations, shops, markets, rural trade stores etc., in interactions in which one of the interlocutors is black and the other is Indian or white. It is less well known in other provinces, where a rival urban lingua franca in the domain of labour is Afrikaans. In rural areas population demographics often dictate that white farmers learn the local Bantu language (especially Tswana in the Free State, Xhosa in the Eastern Cape, and Zulu in KwaZulu-Natal). In former times Fanakalo was also used in non-labour contexts. Mayne (1947: ii) reports that it was used sporadically “by white men amongst themselves when no other means of communication are available“, and Mesthrie (1989: 215–216) reports the same when North Indians had no other means of communication with South Indians. Fanakalo was also used in education contexts when Indians in rural areas were being introduced to an English education by missionaries (Mesthrie 1989: 5).

2. Sociohistorical background ⇫ ¶

The first contacts between Europeans and indigenous peoples in South Africa took place in the Western Cape (1652 to 1800, which roughly marks the Dutch period), where the lingua franca Afrikaans arose out of the experience of colonization and slavery. As Afrikaners moved into the Eastern Cape from c. 1770 onwards, Afrikaans was no longer a viable means of communication with the Xhosa people, and several strategies of communication arose. They involved communication by signs, by simplified (or “broken” versions of) Xhosa, simplified Afrikaans, or a mixture of these (Mesthrie 1998). In addition, communication was sometimes effected through interpreters. English would be added to the frontier in 1820 with the arrival of the first batch of settlers from England. Fanakalo seems to have come into existence in the Eastern Cape from the early 1800s. Mesthrie (1998: 13) gives the earliest recorded sentence as Wena tandaza O Taay ‘You (must) worship God’ (uttered by the missionary John Read, at Kat River in 1816, who thought he was speaking Xhosa). Fanakalo does not appear to have been widespread in this period: It is but one of several strategies that appear in the archival and travel literature of the times, and judging from the sources was not very frequently used. Only one fragment was encountered in the period 1800–1850 in a detailed survey undertaken by Mesthrie for the Eastern Cape.

Of the many diffuse strategies of communication on the Eastern Cape frontier the one that won out later in the new colony of Natal (established 1843) further north along the coast was the Fanakalo option. Initially known as Kitchen Kaffir, the Pidgin was likely to have been brought over by people with the experience of the frontier (Afrikaners moving away from the British in the Cape Colony, their ‘coloured’ servants, English adventurers, and possibly some officials). No concrete evidence of this link exists however, and the accounts and examples of Fanakalo give a picture of a pidgin being invented anew out of the contacts between British settlers and the Zulu population that outnumbered them. Two major crystallizing events for Fanakalo took place in this period: (a) the arrival of indentured Indians in large numbers to the coastal province of Natal (starting in 1860); and (b) the discovery of diamonds and gold in the interior (starting in 1867).

Indians acquired Fanakalo rapidly as a means of communicating with English employers and local Zulus, and occasionally amongst themselves when there was no other lingua franca between North and South Indians. They were probably the ones to stabilize the Pidgin. The gold and diamond rush to the interior resulted in a new Babel containing a variety of European and African languages and some Chinese (Indians were banned from venturing into the interior from Natal). Fanakalo became an important means for the mining bosses to communicate with – and control – the labour force.

3. Sociolinguistic situation ⇫ ¶

For much of the 20th century the Pidgin was closely connected with labour in the mining industry. Since the 1990s there have been moves afoot to phase out Fanakalo because of its negative stereotypes and replace it with English. It is not clear whether these efforts have been successful. Because the mining industry attracted workers from other regions within the continent, Fanakalo has been taken to many parts of Africa, where it often carries positive overtones – representing the sophistication of those young men who have worked abroad in the cities. In South Africa it is frequently denigrated as a language of exploitation and cheap labour. This contrasts with earlier optimistic – if ignorant – descriptions hailing it as a future lingua franca of southern Africa (e.g. Hopkin-Jenkins 1947).

Fanakalo is receding slightly insofar as English is spreading as a lingua franca. In our experience it is facing competition from English amongst younger people, even on farms. There are some non-transactional situations in which it is used. These include a kind of expatriate solidarity or nostalgia for South African things by some emigrants who use occasional expressions from Fanakalo corresponding with family in South Africa (Adendorff 1995a). Fanakalo is sometimes used as a secret language for Zimbabwean or Natal tourists abroad (for those who do not command a Bantu language or Afrikaans anyway). It can be used by Zulu speakers as a form of code divergence that playfully signifies harsher relations with interlocutors than is possible using Zulu (Adendorff 1995a).

Adendorff (1995a) drew an important conceptual distinction between “Mine” and “Garden” Fanakalo. He showed that whereas the mining variety is overwhelmingly Nguni (chiefly Zulu) in its lexis, examples of Garden Fanakalo incorporate a great deal of English. Drawing on Adendorff’s distinction, we would like to describe Fanakalo in terms of three focal types: (a) Farm Fanakalo (typically white or Indian employers and Zulu employees) (b) Garden/Domestic Fanakalo (urban environment where employers are less familiar with Zulu and have learned Zulu with the specific intention to speak to their servants) (c) Mine Fanakalo (planned variety (codified by the mining industry) that arose from Farm Fanakalo). Running through these is a common core that few analysts would disagree with.

4. Phonology ⇫ ¶

Fanakalo utilizes a basic five-vowel system, as shown in Table 1.

This five-vowel pattern is typical of Nguni, where length is not phonemic, and diphthongs are non-existent (vowel combinations either via morphological processes or from loanwords lead to sandhi or glide insertion). Fanakalo does have sporadic diphthongs resulting from fast speech forms: ai ~ hayi ‘no’; aikhona ~ hayikhona ‘no’: daisa ~ dayisa ‘sell’; gwai ~ gwayi ‘tobacco’ etc. Penultimate vowels – as in Nguni – are typically lengthened: thus [bu:ga] ‘see’ versus [bugi:le] ‘saw’. Tonal distinctions are not made, apart from the deictics: lò ‘this’ versus ló ‘that’ (versus lo ‘the’); làpha ‘here’, lápha ‘there’ (versus lapha ‘in’). Final voiceless vowels of Zulu are usually dropped in Fanakalo: thus az ‘to know’ rather than -azi; hash ‘horse’ rather than ihashi.

The basic consonant system is given in Table 2.

This schema is subject to much variation: e.g. aspiration, as in Zulu, is used extensively (e.g. phelile ‘completed’, thatha ‘take’, khaya ‘house’). However, it is unlikely that aspiration is a distinctive feature in Fanakalo. Likewise, one can hear an ejective quality on certain consonants, and the interaction between ejectives and non-ejectives is deserving of further study. Clicks are not used in Fanakalo; they are replaced by velar consonants (usually /k/). The brackets in the consonant chart indicate variation, along ethnic lines. Zulu speakers produce /ɬ/ and /ɮ/ as alveolateral fricatives; other speakers substitute /ɬ/ and /ɮ/ with an initial alveopalatal fricative followed by a lateral [ʃl] and [ʒl]. In rural areas Zulu speakers frequently use /l/ in place of /r/ which occurs in non-Zulu words. Delayed voicing of medial consonants may be retained by Zulu speakers but medial consonants are usually turned into voiced consonants by others (thus [fa:ka] ‘put’ versus [fa:ga]); this results in the variance in spelling between Fanakalo (correct in Zulu) and Fanagalo (correct for many Fanakalo speakers). Prenasalized consonants may be retained by Zulu speakers but dropped by others: thus ndlebe ‘ear’ versus dlebe; nkuni ‘firewood’ versus kuni etc.

As for syllable structure, Fanakalo mostly follows the Zulu preference for CVCV syllables or NCV (where N denotes a homorganic nasal, though as already stated, the nasal is dropped in initial position for non-Zulu speakers). However, the neat Zulu pattern is broken up in words from English and Afrikaans; hence Fanakalo admits clusters with /st/, /sk/, /sp/. The clusters /gw/ and /kl/ are possible in gwai ‘tobacco’ and klina ‘to clean’ (and possibly other words).

5. Noun phrase ⇫ ¶

Nouns do not generally have inflections, in contrast to the range of noun class prefixes of Zulu. However, Mine Fanakalo has ma- as plural marker (usually for [+human] nouns), though some handbooks also give zi- as a plural prefix for animals and -imi for inanimates. In Farm Fanakalo, somewhat surprisingly, no regular noun plural exists. If plurality needs to be clarified or emphasized, speakers use the periphrastic zonke ‘all’. Jugmohan’s (1990) “light” subjects produce occasional plurals in (prefix) ma-. In addition they show an innovation not recorded in any other database – the use of plural suffix -s from English. This is an intriguing anglification of the morphology by urban Indian speakers. (We have no evidence of it, however, in our rural database.)

The demonstratives in Fanakalo are the following:

(1) lò ‘this’ vs. ló ‘that’

làpha ‘here’ vs. lápha ‘there’

It is the preponderance of lo as a definite article and demonstrative pronoun that gives Fanakalo one of its many disparaging names (Isilolo ‘the lo-lo language’). Lo also occurs in the name Fanakalo (‘like this’ i.e. pertaining to instructions of employer to employee). In ordinary usage as well as in handbook translations lo occurs as both definite and indefinite article. Thus Mayne (1947) translates ‘a’ in sentences like ‘Use a dry cloth’ (p. 25) and ‘Dig a trench here’ by lo (not munye ‘one’). Example (2) is from Jugmohan (1990: 58), with his translation (as indefinite).

Indeed, as Cole (1953: 6) remarked, “speakers of Fanakalo acquired the habit of putting lo before every noun, and a ‘personal pronoun’ before every verb, even if the subject be expressed”. For lo this is certainly a matter for closer future investigation, using categories like ‘specific’, ‘generic’, ‘known to hearer’ etc. Other determiners with a quantifier function in Fanakalo are zonke ‘all’, maningi ‘many’, ayi-maningi ‘few’ (literally ‘not many’).

The pronouns of Farm Fanakalo are taken from Zulu:

mina ‘I’

wena ‘you’

yena ‘he/she/it’

thina ‘we’

nina ‘you’ (pl)

bona ‘they’

These are free morphemes which in Zulu are absolute pronouns, despite the regularity of the final na syllable which occurs in Nguni for reasons of stress. In Nguni, these free pronoun forms are used only for emphasis or contrast; in other contexts only the subject prefix is used. Handbooks of Mine Fanakalo do not give the forms nina (‘you pl.)’ and bona (‘they’), substituting the periphrastic forms wena zonke ‘you all’ and yena zonke ‘he/she/it all’. These are innovations of Fanakalo and are not possible in Zulu. In Farm Fanakalo the forms zonke wena and zonke yena (with reversal of the order) are lesser used alternatives (and not characteristic of deep speakers).

For possession, periphrasis occurs: ga X denotes ‘of X’ (where ga < ka, one of the possessive particles of Zulu). With regard to the word order, Fanakalo follows the Nguni patterns:

Thus, the possessor follows the possessed in Zulu and Fanakalo, unlike the unmarked Afrikaans and English order (though English also has the “of”-construction, which resembles the Zulu and Fanakalo pattern).

6. Adjectives ⇫ ¶

Concerning the comparison of adjectives, Jugmohan’s (1990: 75) speakers use the adjective in its invariant form followed by ga ‘than’ (literally ‘of’):

The fusion of the forms ga and lo seems premature to us: More likely the morphology is ga ‘of, pertaining to’ plus lo (art). In other words, since with pronouns the form is ga and not galo, it is clear that lo is a free form that precedes the full noun.

7. Verb phrase ⇫ ¶

Verbs are marked by the -a ending (as in Zulu). This form is the default tense and aspect option signifying present tense, imperative or infinitive. The past tense in Fanakalo is formed by the suffix -ile, which is taken from the perfective in Zulu. This suffix -ile covers the range ‘perfective – simple past’ in Fanakalo as well. The future in Fanakalo in our database is marked by zo (discussed above). Bold (1977: 10) gives a combination of zo + -ile as future perfect: Mina zo idl-ile ‘I will have eaten’. Although this is rare in our conversational database, it is a possible combination given the right discourse context. Examples are (6) and (7).

A tense category that has gone unnoticed by previous writers in Fanakalo is the anterior marker gate. Mesthrie (2007) notes that in Farm Fanakalo gate is frequently used to mark past habitual or past perfect. With non-stative verbs gate collocates with present as well as past verb forms as follows:

With stative verbs like azi ‘to know’, the usual past in -ile does not exist; past tense is expressed by gate + verb:

Inherited from Zulu is the present meaning of some adjectives which are formally perfective verbs. Thus:

Compare these two with the translations of Mina idl-ile ‘I ate’ and Mina gate idl-ile ‘I had eaten’.

Although the Fanakalo verb system is no way as complex as that of Zulu (where there are short and long perfects, (remote) past and narrative pasts, situatives (or continuous pasts)) etc., it is surprisingly systematic. This highly structured nature of tense in Fanakalo has not been appreciated by previous commentators, nor its reliance on the stative/non-stative aspectual distinction.

Other verb inflections, which are all taken from Zulu, are:

-isa (causative) theng-a ‘buy’; theng-isa ‘cause to buy, sell’;

-wa (prs passive) phek-a ‘cook’; phek-wa ‘is cooked’;

-we (past passive) phek-a ‘cook’; phek-iwe ‘was cooked’;

-ela (benefactive) theng-a ‘buy’; theng-ela ‘buy for’ (e.g. ‘Buy (a shirt) for (me)’.)

The preverbal particle (h)ayi is used as verb phrase negator.

In combination with other particles (e.g. zo for future) the negator remains just before the main verb, in our database at least:

Hopkins-Jenkins (1947: 16) gives the reverse order (neg + fut with no alternatives):

This order also occurs in Mayne (1947: 14); whether we are dealing with a genuine dialect difference needs to be investigated further.

Copula absence is normal with predication:

However, with location, copular khona is mandatory:

Some examples of serial verbs occur in Fanakalo, rather like the simple serial order of colloquial English:

8. Simple sentences ⇫ ¶

Fanakalo is SVO in structure in main and subordinate clauses. However, it is not rigidly SVO insofar as it permits topic – comment order as well.

In terms of word order it is noteworthy that Fanakalo employs the adjective-noun order:

Subjects are always overt in Fanakalo:

In fact, Fanakalo prefers a topicalized syntax with every full subject NP followed by an appositional pronoun. Thus ‘John – he went’ rather than ‘John went’ (cf. 13–14).

Apart from Mayne (1947) there is no discussion of reflexives in any previous descriptions of Fanakalo. Mayne records zi as the reflexive (prefixed to verbs), based on similar usage in Zulu (meaning ‘by oneself’). This appears to us inauthentic (Mayne’s intention being to bring about a synthesis/compromise between complexified Fanakalo and simplified Zulu). In our oral database, the hybrid English-Zulu innovative form self ga pro occurs:

We suspect that this is a recent innovation. Other reflexive meanings are handled by emphasis and intonation. It is not possible to form reciprocals in Fanakalo in any grammaticalized way.

9. Complex sentences ⇫ ¶

Relative clauses involve embedding either with a zero relative marker or with lo. Lo has no tonal marking in this instance, unless a distinction between ‘this one who’ (lò) versus ‘that one who’ (ló) is specifically intended. Sentence (25) is from Adendorff (1995b: 12).

Noko is an important subordinator for clauses with ‘if’ and ‘whether’:

Apart from the presence of complementizer noko, no other changes occur to main or subordinate clause.

10. Lexicon ⇫ ¶

Cole (1953) calculated, on the basis of the dictionaries and handbooks of the time, that Fanakalo was comprised of about 70% Nguni (chiefly Zulu) lexis; 24% English and 6% Afrikaans (see table 3 for some examples).

A more explicit count was done by Adendorff (1995a) for Mine Fanakalo. He counted lexical items in terms of types (not tokens) from 3 instructors on the mines whose home languages were Tsonga, Xhosa and Zulu respectively. His results, which we adapt into percentages, were very similar to Cole’s (see Table 4).

based on Adendorff (1995a:81)

Adendorff (1995a: 181) remarks on the high type-token ratio in his sample (i.e. occurrences of separate lexical items compared to the total number of lexemes). He notes further (1995a: 181–182) that lexical richness prevails over “processes that entail reconstituting lexical roots in order to yield new items. Such a principle underlies compounding, reduplication (an instance of compounding) and circumlocution. In the mine variety Fanakalo evidently possesses sufficient lexical primes without having to resort to these means”.

Neither Cole nor Adendorff report words specific to Xhosa or Sotho in their sample; even though (a) Xhosa was the dominant language of the eastern Cape where Fanakalo probably originated and (b) Sotho is the majority indigenous language in the mining region. Accordingly regional variation in Fanakalo’s lexicon is slight. Even in Zimbabwe only a few words specific to the region occur – see Ferraz (1980) for specific examples. In Farm Fanakalo of KwaZulu-Natal some words from Indian languages pertaining to food and vegetables are used.

It is worthy of note that almost all Fanakalo verbs come from Zulu (in Trapp’s (1908) 60 sentences this is true of all verbs; Cole’s (1953) examples have only 6 verbs derived from Afrikaans or English).¹