SHORT NOTE
ON THE (SENSATIONAL) SURVIVAL OF KIKONGO
IN 20TH-CENTURY CUBA
Armin Schwegler
University of California, Irvine
As is well known, Kmtyw ritual languages (e.g., Yoruba, Ewe, Efik)
survive to this day in several countries of Latin America. Together with
Brazil, Cuba is perhaps best known for the preservation of such languages
and for the religious groups that use them in their ceremonies. In Cuba, one
of these ritual languages — the so-called lengua congo — has always been
characterized as a hybrid speech, supposedly the product of a convergence
(syncretism) of multiple Bantu languages on Cuban soil. For this and other
reasons, the lengua congo has, at times, also been called (habla) bant´u (i.e.,
'Bantu tongue').1
The purpose of this Short Note is to draw attention to a forthcoming bipartite
article (written in Spanish; see Schwegler, 1998 and in press a) which
demonstrates that the traditional typologization of the habla congo as essentially
a mixture of multiple Bantu languages is fundamentally flawed. The
Kmtyw component of the habla congo is shown to consist of straightforward
Kikongo — a language long known to have played an important role in the
transatlantic slave trade. Evidence for the claim comes from Lydia Cabrera's
Vocabulario Congo (henceforth VC), a written source published in 1984 that
lists approximately 3000 lexical items of Kmtyw origin. (In several instances,
1 This synonymous term indicates well the wide meaning "Congo" has in Cuba and
elsewhere in Latin America. "Congo" was, for instance, often used generically in combination
with tribal or regional names to denote ethnic origins (cf. CONGO Mombasa = 'TRIBE [called]
"Mombasa"'), even if the ethnic group in question came from well outside the former Congo
kingdom proper. For examples, see Cabrera (1984, p. 52).
160 SHORT NOTE
the lexicon features extended Spanish/"Congo" phrases or sentences.)2 Originally
collected by Cabrera around 1935–1940, the same data are used in my
two-part study to advance the claim that in Cuba, Kikongo must have been a
fluently-spoken language well into the 20th century. This conclusion in turn
raises the possibility that Kikongo may survive on the island to this day.
The study also examines why scholars consistently misanalyzed the true
nature and origin of the habla congo. It is shown, for instance, that investigators'
complete unfamiliarity with Kmtyw languages, false word segmentations,
vague word definitions, and a blind — not to say "naive" — trust in
informants, as well as several other factors, all contributed to the false conclusion
that the habla congo resulted from the convergence of multiple Kmtyw
languages. Moreover, the fact that the Bakongo (speakers of Kikongo) were
among the earliest slaves to arrive in Cuba and other parts of Latin America
further fortified the false notion that their language and religious practices
(santer´ıa) "naturally" underwent significant hybridization, thereby triggering
a fairly rapid abandonment of Kikongo among Cuban Bozal speakers and
their descendants.
Due to limitations of space, my investigation examines only a relatively
small portion of Cabrera's rather extensive corpus (150 words listed in the
sections "A–C" [pp. 15–58] of the VC are etymologized). Although these
sections constitute only approximately 20% of Cabrera's dictionary, a cursory
analysis of the remainder of the book suggests that the general conclusions
drawn from this limited data set can be extended to the entire corpus with a
high degree of confidence.
In carrying out the etymological analysis, it was found that the VC
essentially contains three types of data:
(1) "Congo" words whose Kikongo etymology is entirely transparent
(approximately 40% of the corpus);
(2) "Congo" words whose Kikongo etymology is not entirely transparent
(approximately 25% of the corpus);
(3) Words without any apparent connection to Kikongo sources (approximately
35% of the corpus). These are yet to be etymologized successfully.
2 For example, "¿En qu´e pueblo naci´o? / ¿Saputa ngeui kinani kunanbansa saputa?"
(pp. 124–125, see "Preguntas")
SHORT NOTE 161
Table 1. "Congo" Expressions Whose Kikongo Etymology Is Entirely Transparent3
Vocabulario Congo Kikongo sources
lango 'agua' ('water') < `nlangu 'water, that which is aquatic'
lango kamatoto 'agua de pozo'
('well/ground water')
< `nlangu k´ama `nt´oto lit. 'water extracted
from the earth'
yamasa 'agua' ('water') < ya masa 'place where there is water;
aquatic place'
mamba 'agua' ('water') < m´amba 'water, liquid, juice, inundation,
etc.'
toalalango 'agua' ('water') < to¯ala `nlangu 'bring/transport water'
d´undu 'albino' ('idem') < nd`undu 'albino, white, blond, eureurasiaan'
mulundu 'altura, loma' ('crest, hilltop') < mu ` nl `unda 'hill, crest, hilltop'
lolo 'amarillo' ('yellow') < l`o (reduplicated) 'yellow'
Table 2. "Congo" Expressions Whose Kikongo Etymology Is Not Entirely Transparent
Vocabulario Congo Kikongo sources
m´ungwa Lango 'agua bendita'
('holy water')
< m`ungw´a `nlangu lit. 'salt water'
lango faso 'agua caliente'
('hot water')
< `nlangu fw`a so lit. 'water "ch-ch-ch"
fall(s) (from the boiling pot)' = 'boiling
water that splashes and sizzles'
timba 'amar, amarse' ('to love, to make
love')
< t´ımba 'to be erect, hard, rigid; to get hard
(vulgar, with sexual connotations)'
nu´a mua 'boca' ('mouth') < nnwa ` mvwˆa 'mouth + onomat. for the
sound made when eating'
Illustrative examples of the first and second group are given in Tables 1
and 2 above. Readers will note that in the first group the Kikongo origins
are so transparent phonetically and semantically that even a cursory look at
the data suffices to call into question the supposedly heavily-mixed hybrid
nature of the "Congo" lexicon.
One of the most startling aspects of the VC is the high degree of phonetic
preservation of its data, much of which must originally have reached
Cuba over 400 years ago. Such an observation naturally invites this question:
3 Here as elsewhere in this article, the Spanish translations are those given by Cabrera.
References to the Kikongo source materials are given in Schwegler (in press a).
162 SHORT NOTE
If the data source is characterized by the virtual absence of phonetic erosion,
why then does a significant portion of this supposedly all Kikongo-derived
vocabulary still resist etymologization? In answer to that question, the forthcoming
study shows that a series of methodological deficiencies (non-critical
recollection of vocabulary, false translations, imprecise transcriptions, etc.) in
Cabrera's work — a work considered by many to be truly authentic (see, for
instance, Anhalt, 1996, p. 94) — have conspired to yield a relatively large
subset of data that are anything but a true testimony of the so-called habla
congo. In some instances, these deficiencies are patently obvious, so much so
that convincing etymologization is nonetheless possible. This is the case, for
instance, with "Congo" mungua (VC, p. 25). Cabrera first falsely translates it
as 'sugar', but then unknowingly corrects herself on page 143 by including
mungua as one of the entries for sal 'salt' (cf. Kik. m`ungwa 'salt', Laman,
1964, p. 612).
Cabrera's uncritical acceptance of translations offered by her informants
crucially contributes to obscuring the true nature and origin(s) of the habla
congo. Many of those translations are what we may call "associative" or
"contextual," rather than literal.4 The examples in Table 3 below are illustrative
of the problem (a brief examination and explanation of the associative
steps follows the examples).
Nyuka must have been translated as 'high' instead of 'to feel dizzy'
because it is "high up" where one typically gets dizzy; kizumba is 'dance,
party' because it is in traditional Afro-American dances that one 'jumps,
hops, skips'. Finally, the associative step carried out by the informant in
timba 'to make love' ! 'to be erect; to be sexually aroused (male)' is so
transparent that it requires no further explanation.
Having examined briefly some of the ways in which Cabrera erred in her
data collection, we are now in a better position to understand why the admittedly
fairly large portion of the not-yet successfully etymologized "Congo"
words (VC, letters A–C) does not automatically invalidate the monogenetic
hypothesis (habla congo < Kikongo). In carrying out my etymological research,
I found that many "Congo" items do, in fact, neatly match up phonetically
with potential Kikongo sources, but that they resist successful ety-
4 Explored in Part 1 of the study, the reasons for the informants' behavior are complex
and socioculturally rooted.
SHORT NOTE 163
Table 3. Associative Steps in Translations of Habla Congo Survivals
Vocabulario Congo Likely literal meaning Kikongo etymology
nyuka 'alto' (p. 20)
('high')
'to feel dizzy' < ny`uka 'to feel dizzy'5
kizumba 'baile, fiesta'
(p. 20) ('dance, party')
'to jump, hop, skip (in a
dance)'
< ki `nsumba 'jump, hop,
fastbody movement (e.g.,
in a dance)'6
timba 'amar, amarse'
(p. 20) ('to love, to make
love')
'to be aroused sexually'
(male)
t´ımba 'to be erect,
hard, rigid; to get hard
(vulgar, with sexual
connotations)'7
mologization because the semantic arguments for a direct genetic link cannot
be made in an airtight fashion. One cannot help but suspect that Cabrera's
propensity to record associative rather than literal meanings has unwillingly
(and artificially) created such etymological cruxes. For this reason, and in the
absence of evidence to the contrary, "Congo" words without any apparent
(semantic) connection to Kikongo sources do not disprove the monogenetic
hypothesis.
The recognition that the Cuban habla congo is "simply" straightforward
Kikongo has considerable theoretical consequences.8 For one thing,
it suggests that the Bakongo may have played a far more important role
in the formation of Afro-American language and society than scholars ever
imagined before. This impression is fortified further by recent findings from
other speech areas, notably El Palenque de San Basilio (Colombia), whose
200 or so surviving Afro-Palenquerisms have similarly strong Kikongo roots
5 Ny`uka 'avoir le cauchemar' (Laman, 1964, p. 818).
6 Ki-nsumba 'saut, bond'; compare also ki-nsumba kyam`akinu 'saute de danse' (Laman,
1964, p. 281).
7 Timba 'ˆe. Raide, rigide; se raidir a diverses significations; dans les dialectes, surtout au
sens obsc`ene' (Laman, 1964, p. 973).
8 As stated explicitly in Part 1 of my study, I am not claiming that Kikongo was necessarily
the only Kmtyw source of the habla congo. Certain lexemes with a phonetically similar root
have such a wide geolinguistic distribution within the Bantu territory that the probability of
multiple origins (convergence) is indeed a high one for a number of "Congo" words. My study
of the VC suggests, however, that such items are also traceable to the main donor language,
that is, Kikongo.
164 SHORT NOTE
(Schwegler, 1996, 1999, and in press b). The finding also raises doubts about
the supposed early and abrupt abandonment of Kmtyw languages among
certain slave groups. As the Cuban case demonstrates, Kikongo survived for
centuries as a private in-group language, and as such it may have had a
significant impact on the evolution of other types of local speech, including
popular American Spanish. It remains to be determined just exactly what the
long-term linguistic consequences of this impact were. But in light of what
we know now about the true nature of the habla congo, there is good reason
to examine this and other related questions from afresh.
References
Anhalt, N. G. (1996). Cunyaye para Lydia Cabrera. Am´erica Negra, 12, 91–
103.
Cabrera, L. (1984). Vocabulario congo. El bant´u que se habla en Cuba.
Miami: Colecci´on del Chichereku.
Laman, K. E. (1964). Dictionnaire kikongo-franc¸ais. 2 vols. Ridgewood, NJ:
The Gregg Press.
Schwegler, A. (1996). Chi ma nkongo: Lengua y rito ancestrales en El
Palenque de San Basilio (Colombia). 2 vols. Frankfurt/Madrid: Vervuert
Verlag.
Schwegler, A. (1998). El vocabulario (ritual) bant´u de Cuba. Parte I: Acerca
de la matriz Kmtywa de la "lengua congo" en El Monte y Vocabulario
Congo de Lydia Cabrera. Am´erica Negra, 15, 137–185.
Schwegler, A. (1999). El vocabulario Kmtywo de Palenque (Colombia).
Segunda Parte: compendio de palabras (con etimolog´ıas). In L. Ortiz
(Ed.), El Caribe hisp´anico: perspectivas ling¨u´ısticas actuales (Homenaje
a Manuel ´ Alvarez Nazario) (pp. 171–253). Frankfurt/ Madrid: Vervuert
Verlag.
Schwegler, A. (in press a). El vocabulario (ritual) bant´u de Cuba. Parte II:
Ap´endices 1–2. Am´erica Negra, 16 (1998).
Schwegler, A. (in press b). The Kmtyw vocabulary of Palenque (Colombia).
Part 1: Introduction and corpus of previously undocumented Afro-
Palenquerisms. Journal of Pidgin and Creole Languages, 15(2).