Cosper, Ronald. Barawa Lexicon : a wordlist of eight South ... - Llacan

Cosper, Ronald. Barawa Lexicon : a wordlist of eight South Bauchi (West Chadic) languages : 

Boghom, Buli, Dott, Geji, Jimi, Polci, Sayanci and Zul, (vol 39). Munich : LINCOM EUROPA, 1999, 

158 p. LINCOM Studies in African Linguistics. 

The publication of new data on little documented Chadic languages is rare enough an event for any 

linguist to be enthusiastic when they hear the news. So was I when Ronald Cosper’s Barawa Lexicon 

was announced and published by Lincom Europa. Before this, the main reference on the topic was 

(Shimizu, 1978). 

Shimizu’s work, written after a survey carried out in 1974 and 1975, comprises wordlists covering 108 

items on 32 languages, map, subclassification, cognate percentages, and sound correspondences. 

‘Barawa’ is a term chosen by Shimizu to name the speakers of these 32 languages, which he considers 

to be a dialect continuum. He adds that « the use of the term ‘Barawa’ to denote this group was 

confirmed both in the Dass district generally and by the Chief of Boot » (op.cit., p. 7). In the various 

trips I did to the Dass area, Boot, Sigidi and during my fieldwork in the Sayanci area over the past 

decade, the term Barawa was accepted by all speakers except those of Sigidi and Zaar who don’t know 

the word. 

Cosper’s Barawa Lexicon covers 915 lexical items for 8 languages : Jimi, Zul, Geji, Polci, Dott, 

Sayanci, Buli and Boghom. In the introduction, he gives the chart of phonemes for the eight languages. 

The main interest of this work lies then in the increase in the number of lexical items documented. 

Unfortunately, from what I can judge, the quality of the data is far from satisfactory. I will give a few 

examples where I will compare Cosper’s data with the one I collected on two of these languages : Zaar 

(Cosper’s Sayanci) on which I have been working regularly since 1991, and Zoéi (Cosper’s Dott) on 

which I did 2 weeks’ fieldwork in 1999. 

Zaar 

Cosper marks two tones for Zaar, although he says « Some of these languages, e.g., Sayanci, may in 

fact have three phonetic levels of tone, but as Schneeberg points out, they are phonologically 

analyzable as two-level systems. » Sayanci does have a three-level phonetic tone system. The 

phonological two-level system can only be the result of a phonological analysis involving a very 

elaborate system of rules (cf. Schneeberg 74). Failing that, we are left, in the case of Cosper’s work, 

with a two-level phonetic transcription, which in the case of Saya, is simply wrong. 1 

We will just mention a few cases of plain mistakes in marking the tones : 

number gloss Cosper Own data 

2 woman, female gËt gÈt 

124 compound dËn dÈn 

146 grindstone vúun vuun 

148 hut víin vììn 

448 war máàndŒ maandŒ 

664 fight máàndÉ maandŒ 

294 insect zhíßar ??? 

304 worm zhißŒr ©ìbÈr 

Some cases of strange translations or notations : 


5 girl ÑaayŒt (sic !) Ñaa-gÈt 

7 daughter nyíi, ÑáagŒt ñíí 

688 walk, wander nyalán ÑgeláÑ [ÑgyeláÑ] 

453 conversation ÑgáàghŒnlaa ??? 2 

508 short afúbm àà fúpm h¢ 3 

1 When quoting my own data, I will use a three-level (phonetic) transcription (H= acute ; L = grave ; M = 

unmarked ; long vowels will be marked with double letter, each one bearing a tone). 

2 I have no translation for the word conversation. Ñgáá‹Ën laa means to rest, lit. to take work (??!!!).

Items 477, 495, 595 and 596 concern the ‘indefinite’ [cf. Hausa wani (sing.) /wasu (pl.)]. As a 

determiner, it is postposed to the noun and appears as wón / gón [gwón]. As a pronoun, it has two 

forms : gón (sing.) gyaa gón (pl.) resp. [gwón] and [gyaa gwón]. 


477 most gyaagwón ??? 

495 some (wasu) gyaagwón gyaa gón 

595 one (wani) gwán gón 

596 some (wasu) gyaagwán gyaa gón 

A cursory look at the items concerning demonstratives is sufficient to wonder whether the author has 

taken the care of reading his manuscript before sending it to the printer. The same translation (gíi) is 

given for (601) this and (602) that , which should be resp. gín/giní and gíí, whereas (597) this is given 

as gíin and (598) that as gii ! 

Generally, what strikes the reader when having a cursory look at Cosper’s lexicon is the absence of a 

minimum analysis of the data, which leads the author to mistaking syntagms for lexical items, or being 

inconsistent in his use of verbs or verbal nouns in the translations. 

Let’s take the following items : 


734 fart tuurgunyoghŒn túúr‹Ën ñô‹Ñ 4 

748 smell 5 •uughŒnzhóghÑ wúmgËn ©ò‹Ñ 6 

75 mouth vii vìì 

704 shout dláarvi ›áár‹Ën vìì 7 

706 speak wúlvi wúl‹Ën vìì 8 

Likewise, (30), saghátkŒn (sa‹átkËn) is the verbal noun of sa‹át to be intelligent ; intelligent person 

should be mur ‹É sa‹ákËn 

The mistakes are too numerous for me to go on listing them. To conclude this passage concerning Zaar, 

I will just give the list of independant pronouns : 

Zo•i 


608 I myân myááni / myáàn 

609 thou kyan kyááni / kyáàn 

610 he, she, it yân yááni / yáàn 

611 we myâni myààní / myàán 

612 you kyáàni kyààní / kyàán 

613 they yaashi yààshí 

The same problems appear in the work on zo•i. The general impression is that of a sloppy work due to 

lack of analysis. 

If one looks at the list of verbs given by Cosper, it is obvious that some suffixes have not been properly 

identified. 

3 

Lit. : it is not far/long/tall (àà = 3s. Perf.); the negation (h¢) is missing in Cosper’s translation. The adjectives 

ndóngat or ndúmi are a good translation. So is the verb •at, to be short. The adjective ndúúri is used to 

qualify something that has been shortened. 

4 

Lit : to cook shit ; [ñôghÑ] is in fact /ñôkn/. 

5 

From the Hausa sansana, one guesses the English verb is supposed to be transitive. 

6 

Lit : hear/feel a smell. •úghËn is the noun verb of the verb •u, to beat. The expression •u zhòghÑ may mean 

« to have a strong smell » 

7 

lit. : cut mouth ; NB : riiÑ = a shout. ÿi‹á (3s. Continuous) riiÑ = he is shouting 

8 lit. : say mouth.

The suffix -ti is used in zo•i to form verbal nouns. On the forms given by Cosper, the suffix appears 

haphazardly : 

number gloss Cosper Own data 9 

715 peep at yéli yeli, yeliti 

716 see yeli yeli, yeliti 

717 watch yeliti yeli, yeliti 

762 blow (of wind) bakti fwat; fwatti 

763 burn (of fire) bak bàk; bàktì 

730 desire ndŒkti ndÉkŒ; ndÉkŒti 

739 hear, feel wom wóm, womti 

Likewise, a -tiya suffix (secondary verbal noun ?) appears in his data, which has not been properly 

analysed : 


675 return doÑtíya dòÑ, dòÑtì 

774 break (in pieces) tlaartíya ¨ar, ¨arti 

885 be suitable ( ?) taatita, taatiyá taa, taati 10 

The perfect/accomplished in zo•i is formed with -ak suffixed to the verb : 


788 fill zuwak dzù, dzùtì 

724 be tired kŒmmak kÉm, kŒmti 

691 agree to kappák kap, kapti 

As a conclusion, I can only say that this work is seriously questionable, and should never have been 

accepted for a publication. It is not a reliable source, and I would not advise anyone to use it for 

comparative work. 

As matters stand, (Shimizu, 1978) and (Kraft, 1981) remain the only reliable sources. 

Bernard CARON 

Inalco 

UMR 7594 

Kraft, Charles H. Chadic Wordlists. Berlin : Dietrich Reimer, 1981. Coll. Marburger Studien zur 

Afrika und Asienkunde ; n° 23-25. 

Schneeberg, Nan. Sayanci Phonology. Ph.D. : Indiana University. 1974. 

Shimizu, Kiyoshi. The Southern Bauchi Group of Chadic Languages. A survey report. Africana 

Marburgensia, 1978, Special Issue 2, p. 50. 

9 My own data comes directly from my field notes, and needs to be further analysed, specially concerning tone : I 

have heard and marked three tones, which may eventually be reduced to two. For each verb, I give the lexical 

form, followed by the verbal noun. 

10 For this verb, I have been given the meaning to look like, to resemble.

Cosper, Ronald. Barawa Lexicon : a wordlist of eight South ... - Llacan

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?