Today’s post is another post in the (unofficial) series “Refuting wrong (or unproven) but widespread claims about the Baskeet language.” Baskeet is an Omotic language spoken in the Basketo Special Woreda and in parts of the Melokoza Woreda of the Gamo-Gofa Zone in Southern Ethiopia. One widespread claim in the literature holds that the Western branch of the Ometo languages is made up of Baskeet and its “dialects” Dookka and Doolla.
Spelling variants of Dookka are: Doco, Doko, Doqo, Dokko, Dokka
Spelling variants of Doolla are: Dolo, Dola, Dollo, Daula, Dahoula
Here are some quotes from the linguistic literature (emphasis mine):
- Fleming (1976): “West Ometo dialect cluster: Basketo, Dokka, Balta, et al.”
- Alemayehu (2002) writes: “Bender, in one of the first notes on this language, uses the names ‘Basketo’ and ‘Mesketo’, and he gives the names of two dialects: ‘Doko’ (Doqo, Dokko, Dokka), and ‘Dollo’ (1975: 236).”
- Bender (2003): “Basketo (Bask.) or Mesketo (basket-noona = “Basket-mouth”), Doko (Dk.)-Dolo” (p. 2). Bender (2003) speaks consistently of Basketo and Doko-Dolo (see e.g. p. 51).
- The “Basketo” entry of the Encyclopaedia Aethiopica (2003) starts as follows: “Together with the closely related Dokko-Dollo, the B[asketo] language forms the western branch of the Ometo cluster […].” (I: 498)
- In the most recent overview of Omotic, Azeb Amha (2012) lists “Basketto-Dokko-Dolo” in her list of Omotic languages and (their) dialects.
In the course of my documentation project I have so far not come across any dialectal differences in Baskeet (but, I admit, I haven’t had the possibility yet to visit the Baskeet-speaking areas in the Melokoza Woreda of the Gamo-Gofa Zone). So I wanted to find out on which language data these statements about the alleged dialectal differentiation were based.
1. Dookka data
In fact, the Dookka-/Doolla-dialect “thing” goes back to one single source, Conti Rossini (1927), or rather an analysis of Conti Rossini’s data in the works of M.L. Bender (see below)! Carlo Conti Rossini included in his 1927 publication a “Doco” [= Dookka] vocabulary list, which listed the data he himself had collected as well as data from the notebooks of the explorers Antoine T. d’Abbadie and Vittorio Bottego & Maurizio Sacchi. Altogether, the “Doco” list is made up of a list of numerals and about 100 other lexical entries. (Note that Conti Rossini makes no mention of Baskeet!)
I have taken a closer look at the “Doco” data presented in Conti Rossini (1927). Almost all entries in his word list can be easily identified as Baskeet – even though the transcription might not be perfect and the translation might sometimes be not quite right. Let us take, for instance, the first entries on p. 249:
Compare this with my Baskeet data (in a simplified Latin-based orthography):
- atina’ ‘boy, male person, son’
- acci ‘teeth, tooth’
- awa ‘sun’
- aila ‘slave’
- haikk’ire ‘die’ (actually: ‘he will die’)
- aissh ‘meat’
The identification of some terms in Conti Rossini’s list is obscured by the translation, see e.g. ulle, translated by him as ‘bambino’ [= ‘child’]; ulle is, however, an intimate term of address (‘my dear, my love’) for people who are younger than or of the same age as the speaker. Only very few entries (approx. 6) in Conti Rossini’s list could not be identified, among them bellebir ‘cantare’ [= ‘sing’]. Some terms could be identified but they are not Baskeet, e.g. gāya ‘pipa’ [= ‘pipe’], which is likely to be a Gofa word (which borrowed it from Cushitic); the Baskeet term would be dampi goyss ‘pipe’. As in most early word lists there are some classical mistakes, e.g. the translation of tira [‘chest’] as ‘I’ [in Italian ‘io’], probably because one interlocutor pointed at his chest and the other one interpreted this gesture has a translation of ‘I’. :-)
To conclude, there is no doubt that the “Doco” data collected by Conti Rossini is from the same variety as the Baskeet data that I am collecting in my project (and the same variety as the Baskeet data presented by other researchers in other publications)!
2. Doolla data
The only language data of the alleged Doolla “dialect” is from the notebooks of Antoine T. d’Abbadie and published in Conti Rossini (1927) under the heading “Dollo”. “Dollo” is described by him as a “part of the Doqo in the West close to Dime” (frazione dei Doqo confinante a W coi Dime). The language data that he presents is just a list of numerals (click to enlarge):
The numerals up to 20 are easily identifiable as Baskeet numerals (cf. pettan 1, nam’i 2, hayzzi 3, oiddi 4, isshin 5, lehi 6, tabza 7, lamakai 8, saakali 9, tabb’a 10 and lamtam 20). The numerals for the higher tenths could also be Baskeet (at least, they are made up of Baskeet lexemes) but they seem to represent a vigesimal (base 20) system; see lāmi gābā ’40’, lit. “two gābā”, lāmi gābā tā’bā ’50’, lit. “two gābā (and) ten”.* I have never come across such a base 20 system in Baskeet, but I can, of course, not exclude that such a system existed in the language at the end of the 19th century when Abbadie visited the area.**
The Doolla data in Conti Rossini (1927) is too scarce to provide any evidence for a Doolla “dialect”, i.e. a regional variety of Baskeet.
3. M.L. Bender’s interpretation of the Dookka data
Conti Rossini’s Dookka and Doolla data were the input of M.L. Bender‘s works on the classification of Omotic languages. I have only taken a closer look at one of his latest publications, Bender (2003): Omotic Lexicon and Phonology. Here Bender claims that Baskeet and its alleged Dookka “dialect” share (only) 77% of their basic vocabulary (2003: 51). Bender’s only source for Dookka is Conti Rossini (1927). As Conti Rossini (1927) only includes some 100+ Dookka entries, the most common entry in Bender’s list of correspondences of the Northwest Ometo Family (p. 54-64) is “n.d.” [= no data] in the Dookka column. Furthermore, the Dookka and Baskeet lexemes often just seem different because the translations are wrong (or not quite right) and thus non-cognate lexemes are compared and seen not to match (which decreases the rate of shared vocabulary). Here are some examples:
- p. 56, no. 33: [Dookka]: oka ‘cow’ is compared to Baskeet miz-, miiz–
–> But: oka is translated wrong; cf. Baskeet okki ‘heifer’ / miiz ‘cow’
- p. 57, no. 48: [Dookka]: yeyde ‘fear’ is compared to Baskeet iic’
–> But: yeyde is translated wrong; cf. Baskeet yeide ‘come’ / iicc’ide ‘fear’
- p. 62, no. 120: [Dookka]: guts’e ‘small’ is compared to Baskeet gilla
–> But: There is also a (cognate) word guutts ‘small’ in Baskeet.
The lack of data and the faulty comparisons lead to a skewed picture. Bender claims that there are major differences between a Dookka and a Baskeet variety (the distance between Doolla and Baskeet or Dookka is not calculated). As Bender’s works are often quoted, the claim that West Ometo is made up of Baskeet and its “dialects” Dookka and Doolla is repeated over and over again in the literature. However, it does not become “righter” by being copied repeatedly! :-(
4. If the language in Dookka and Doolla was Baskeet than why was it not called that way?
It would require a historical and anthropological study to answer this question and I can, at present, only speculate about the reasons. Dookka, Doolla and Baskeet were possibly considered to be different dialects because they represented different political and/or geographical units and their inhabitants were thus assumed to speak different dialects. Or early researchers and explorers just called the language variety according to the place where they worked or according to the origin of the speaker they worked with. As different researchers worked in different places and made different transcription and translation errors, the picture of a dialect clusters arose.
It is interesting (but unfortunately also very complex!) to find out in reference to which geographical and/or political entities the terms Baskeet, Dookka and Doolla were used in the early geographical and anthropological literature. A. T. d’Abbadie (1890: Géographie de l’Éthiopie, ce que j’ai entendu, faisant suite à ce que j’ai vu), J. Borelli (1890: Éthiopie méridionale. Journal de mon voyage aux pays Amhara, Oromo, et Sidama (septembre 1885 à novembre 1888)) and E. Haberland (1959: Die Basketto und verwandte Stämme) divide the area where Baskeet is spoken today (i.e. the Basketo Special Woreda and the western part of the Melokoza Woreda) into different “kingdoms”. The number and names of “kingdoms” vary from author to author.***
The German anthropologist Haberland, for instance, speaks of five “tribes”, Basketo [= Baskeet], Doko [= Dookka], Balt’a [= Balts’a], Dola-Mašira [= Doolla-Mashira], Dafa [= D’apa] and Laha [= Laaha], which are said to form, from a cultural and ethnic point of a view, a “complete whole“. They are said to be clearly distinct, for instance, from the neighbouring Melo (p. 190f) – see his map below. According to Haberland, the traditional leader (kaati) of Baskeet was the head of the Goyrina’-clan. The traditional leaders of Balts’a and Dookka were members of the Goshina’-clan. The traditional leaders of Doolla-Mashira and D’apa were Murki, those of Laaha were “Salakatna” [correct transcription not known to me]. These areas seemed to form a cultural but not a political unit.
While Haberland uses the cover term “Basketto [= Baskeet] group” in reference to the five above-mentioned “tribes”/”kingdoms”, Abbadie and Borelli use “Dokko” and “Doko” [= Dookka] as a cover term! In Haberland’s work “Basketto” has thus two usages:
- in its wide sense it refers to the entire (ethnically) Baskeet area, i.e. an area that was divided into independent political units,
- in its narrow sense (as indicated on the map below) it refers to one of these political units of the ethnic Baskeet area, i.e. the area around the town Laska and the home of the Baskeet kaat (“Baskeet king”).
The term “Doko” [= Dookka] is used by Haberland in reference to another political/ geographical unit of the ethnic Baskeet area (cf. his map).
In contrast, in the works of Abbadie and Borelli, “Dokko”/”Doko” is used as a cover term for the entire (ethnic) Baskeet area. So in fact, Abbadie’s and Borelli’s “Dokko”/”Doko” and Haberlands “Basketto group” are synonymous!
Doolla is used in the early anthropological and geographical literature for an area in the Northwestern lowlands of the ethnic Baskeet area; see “Daula” on Haberland’s map above.
5. Today’s usage of the term(s) Dookka (and Doolla)
In today’s usage, the term “Dookka” (i) occurs in the names of two administrative units, (ii) it can refer to the geographical unit that is outlined on Haberland’s map (1959) (see above) and (iii) it be used as synonym of “Baskeet”.
5.1. Usage as an administrative term
There are two small administrative units (neighbourhoods: k’äbäles) in the Basketo Special Woreda which include the term Dookka, i.e. Dookka-C’ari-Subba and Dookka-Ayma (in Amharic: ዶኮ ጨሬ ሱባ and ዶኮ አይማ). The two neighbourhoods are located some kilometres north of Laska in the direction of Balts’a. I cannot say whether Doolla is also used as a name for a ward (k’äbäle) in the Melokoza Woreda – I haven’t visited the area yet.
5.2. Usage as a geographical term
“Dookka” continues to be used as a geographical term for the area where the head of the Goshina’-clan, i.e. the Dookki kaat, and most of his clan members live (cf. Haberland’s map). As such, Dookka encompasses roughly the wards of Dookki-Ayma, Dookki-C’ari-Subba, Wad’-Bal’aattsa (and possibly also Boola) of today’s administration. The centre of Dookka is about here:
The old geographical/political term Dookka has no relevance for today’s administration.
5.3. Dookka and Baskeet as synonymes
I had mentioned above that Haberland’s “Basketto group” is synonymous to Abbadie’s and Borelli’s “Dokko”/”Doko”. The partly synonymous use of “Dookka” and “Baskeet” can also be observed today.
- Instead of saying that someone speaks “Baskeet”, I have heard people saying “s/he speaks Dookka”.
- I have heard that people used the verbs baskeetittsabe ‘Speak Baskeet!’ and dookkittsabe ‘Speak Dookka!’ as synonymes.
- And if Baskeet speakers want to contrast the thick light-coloured local Baskeet beer to other beer types, they may speak of Dookki parsa ‘Dookka (i.e. Baskeet) beer’ – e.g. in contrast to the dark-coloured Amari parsa ‘Amhara beer’ (i.e. t’älla ጠላ)’. See a picture of Dookki parsa (Amharic translation: borde) below:
At the end of this lengthy (and hopefully not too confusing) post let me summarise my main point: We have no evidence that Dookka and Doolla are dialectal variants of Baskeet (or, for that matter, that there is any other dialectal variation). Thus we have no evidence for a “West Ometo dialect cluster”. Instead we have evidence that the Dookka and Doolla data published in Conti Rossini (1927) represent the same Baskeet variety that I am documenting in my project. Furthermore, we have evidence that the terms Dookka and Baskeet (in their wide sense) are used as synonyms.