Home / Archive / 2011 / Joachim Crass & Ronny Meyer (eds.) 2009. Language Contact and Language Change in Ethiopia.
Document Actions



The present book is one of the many outcomes of research carried out at the University of Mainz on language contact phenomena in the Ethiopian Linguistic Area. While a previous book edited by Joachim Crass and Ronny Meyer concentrated on copulas, focus morphemes and deictic elements (Crass & Meyer 2007), the present work does not have a particular thematic focus; rather it deals with various contact-induced and internally motivated language change phenomena in Omotic, Cushitic and Ethiosemitic languages. Apart from the introduction by the editors, the book contains six articles. In the following sections, the articles will be summarised and assessed individually. The review will end with a general evaluation of the book.


Binyam Sisay’s article “Copula and/or focus: The morpheme -(k)ko in two East Ometo languages” (pp.8-15) is a contribution to the on-going debate on whether certain morphemes attached to predicates in non-verbal clauses in Ethiopian languages should best be analysed as copulas or as focus morphemes (see other contributions dealing with this question in Crass & Meyer 2007). The author takes a clear stand, refutes an earlier description of Koorete (Hayward 1982) and argues that the ‑ko-morpheme is a focus morpheme (and not a copula) and that the copula is zero in affirmative non-tensed, non-dependent clauses (a feature shared with Yemsa and Benchnon, see Rapold & Zaugg-Coretti’s contribution in the book). As the predicates of non-verbal clauses are usually in focus, ‑ko has been mistakenly interpreted as a copula in earlier works. In verbal clauses, the morpheme -ko can mark any constituent as being in focus. The author acknowledges that focus marking in Koorete could have originated in cleft-constructions and that ‑ko could have been a copula in an earlier stage of the language but he shows that, synchronically, there is little reason to analyse - ko as anything other than a focus morpheme. Based on data from Hirut (2004), he proposes a similar (re-)analysis of the -kko-morpheme in the closely related language Haro.


The arguments presented in favour of an analysis of - ko as a focus marker are convincing at first sight. However, the article is very short and a more elaborate account of the Koorete focus system would have been very welcome. I would have liked, for instance, to learn more about the interdependency of focus marking and verbal agreement marking (which is only exemplified with a single example on p.11) and to get a brief overview of the relativisation strategies in Koorete in order to recognise the traces of relative structures in main clauses with focussed subjects (p.12). Furthermore, non-verbal clauses with predicates other than common nouns or adjectives and with more complex predicates (e.g. nominal predicates that are modified) should have been presented to strengthen the argument.


A statement in the conclusion of the article comes as a surprise: Koorete is categorised as a language with only a “weakly” grammaticalised focus marking system based on the fact that its historical origins are still fairly transparent. If focus marking is syntactically determined in a language and if speakers have to make a choice in each sentence which constituent to mark as focussed (as the author himself shows in his article), I wouldn’t choose to characterise the language as having only a “weakly” grammaticalised focus system.


Ronny Meyer’s article “The quotative verb in Ethiosemitic languages and in Oromo” (p.17-42) is a detail-rich and enlightening contribution dealing with the morphosyntactic, semantic and pragmatic properties of the verb ‘say’. The article concentrates on the languages Amharic and Muher, however, comparative data from various other Ethiosemitic languages is provided, too. (Contrary to the expectations raised by the title of the contribution, Oromo is actually not treated in any depth.)


After the introduction, section 2 of the article is dedicated to an in-depth discussion of the grammatical features of the verbs 'say’, 'tell' and 'speak'. The etymology and the synchronic distribution of these utterance verb lexemes across Ethiosemitic language are discussed and it is shown that only cognates of the 'say' verb are attested throughout Ethiosemitic (while more than one non-cognate ‘tell’ and more than one non-cognate 'speak' lexemes are found). Meyer then examines the argument structure of the verbs 'say', 'tell' and 'speak' and points out important differences between them. The verbs differ with respect to which arguments are expressed obligatorily, how theme arguments are encoded morphologically, which arguments can be passivised, whether theme arguments can be marked for definiteness, and how object suffixes on the verbs are interpreted.


The section is full of interesting details and I learned many things that I was not aware of before. I would have also liked to know how non-pronominal “addressees” are encoded (only examples with pronominal “addresses” are given). Unfortunately, there are also some terminological inaccuracies. I find, for instance, the author’s use of the terms “direct object” and “addressee” problematic. He writes e.g. on p.21: “'say' and 'tell' are “trivalent verbs with a subject […], a direct object […] and an addressee”. The term “direct object” is used by the author for the constituent that expresses what is said or told and “addressee” is used for the constituent that expresses who receives/is addressed by what is said or told. There are obviously significant differences in the argument structure of 'say' and 'tell' (as argued in detail by the author himself) which would warrant not labelling the constituent which expresses what is 'said' and the constituent which expresses what is 'told' indiscriminately as “direct objects”. While the theme argument of 'tell' does display direct object characteristics, no such characteristics can be associated with the theme argument of 'say' (it can neither be passivised nor marked as definite and accusative). [1] I also find the use of the term “addressee” in coordination with “subject” and “object” (cf. quote above) inapt, as it denotes a semantic role while “subject” and “object” are labels for grammatical functions. It is not clear to me why the term “indirect object” is not used instead.


Section 3 of Meyer’s contribution gives an overview of the wide range of constructions in which the verb 'say' is used in Muher, not all of which can be discussed here. Apart from animate speakers, the Muher 'say' verb also allows, in certain contexts, nouns with inanimate referents to be used in subject function; like in many Ethiopian languages, failure to open a door, for instance, is literally translated as a fictive speech act, namely “The door said «I don't open»”, meaning '(s.o.) couldn't open the door'. The verb 'say' is used as a light verb to integrate ideophones into clauses i.e. to make them inflectable. Furthermore, the 'say' verb is a common translation equivalent of 'think'; it can introduce beneficiaries (lit. “saying for you” meaning 'for the benefit of you’) and, most astonishingly, it is used to express hit (lit. “I will say to you” meaning 'I will hit you')  [2] .


Section 4 starts off with a review of Cohen et al.’s (2002) article on the grammaticalisation of 'say' in East Africa, or more precisely, with a review of their typology of phrasal verb constructions. In the literature on Ethiopian languages, a “phrasal verb” is a combination of an (invariant) ideophone with an inflecting light verb (usually 'say' or 'do'), e.g. Amharic qučč’ alä 'sit down'. Meyer shows that Cohen et al.’s (2002) typology of phrasal verb constructions is of limited use for Ethiosemitic languages (and, unfortunately, also based on out-dated sources); he reduces their typology to two language types: Type I which uses 'say' to integrate ideophones into intransitive as well as transitive sentences and Type II which uses 'say' to integrate ideophones into intransitive sentences and 'do'’ into transitive sentences. The first language type is common in North Ethiosemitic; the second language type is prevalent in South Ethiosemitic. Finally, Meyer discusses some noteworthy examples from Amharic and Muher which demonstrate that 'say' can also integrate ideophones into transitive sentences, a fact that has so far remained unaccounted for in the quite extensive literature on ‘say’ verbs in the Horn of Africa.


At the end of my evaluation of Meyer’s long and enlightening contribution, I have to admit that I did not understand various details of his arguments and that I found the wording lacking in clarity in some paragraphs. Here are some randomly selected aspects that would have required clarification: 1. The verbal root ngr is repeatedly translated as 'report as a fact ' (cf. e.g. p.22 above) but it does not become clear why the translation has to include ‘fact’ and the reader wonders if the same verb form could not also be used when lies, assumptions etc. are reported. 2. How can the subject of the main clause and the quoted clause be different in ‘Hei told me: “Ii/j will come tomorrow”’ (p.22), i.e. how is interpretation Ij possible? 3. I can’t follow the argument that the non-prototypical “objecthood” of the theme argument of 'say' could be explained by its “semantic incorporation” into the verb (p.25). 4. In the examples given, I can’t see which subjects of phrasal verbs with 'say' “initiate and/or control the event but [are] at the same time also affected by it” (p.37f). 5. Why is the subject of the 'say' verb in ex. (16) (given on p.27) labelled “addressee” on p.29 and not “(fictive) speaker”?


Ongaye Oda’s paper “The spread of punctual derivation in Dullay and Oromoid languages” (pp.43-57) is a re-examination of one defining feature, the punctual derivation, of the Sagan Linguistic Area, a sub-area of the Ethiopian sprachbund, which was proposed by Sasse (1986). The linguistic area named after the Sagan River in Southwest Ethiopia is assumed to consist of languages from two (sub‑)branches of Lowland East Cushitic (Werizoid and Konsoid) as well as the Highland East Cushitic language Burji and the unclassified language Ongota. The author summarises and evaluates earlier works on this linguistic area in general and the punctual derivation in particular and then sets out to provide more data on his native language Konso. In most languages of the area, the punctual (or “singulative”) derivation is marked by a geminated root-final consonant and usually expresses that an action is performed only once; see Konso le ɓ- 'kick' vs. leɓ ɓ- ‘kick once’ (p.46).


The most valuable part of the paper is section 8, in which the author discusses the semantics of the punctual derivation in Konso in more detail. He shows, among other things, that imperative forms of punctual verbs express that orders are to be fulfilled immediately (not, as expected, fulfilled once) (ex. Konso leɓ-ɓ-i 'kick immediately!', p.53). He also discusses the interesting interaction between the punctual derivation ('doing something once') and the frequentative derivation ('doing something several times') as well as the interplay between verbal derivation and number marking on objects.


It is the new Konso data that makes the article worth reading; unfortunately, however, the article does not do justice to the aims as they are stated in the title and the introduction, because the “spread of the punctual derivation” is mentioned but not argued for. The far-reaching historical conclusions drawn in the last section are not backed up by the data provided earlier; see, for instance, the quote from the final section 9: “Since this [i.e. the punctual] derivation is not known outside the Dullay and Konsoid languages and, given the fact, that Dullay and Konsoid are different Cushitic subgroups spoken in adjacent areas, and that the feature is more productive in Dullay, we further conclude that the occurrence of the punctual derivation in the Konsoid subgroup is due to language contact (emphasis mine).” The author has not discussed varying degrees of productivity of the derivation in the different languages and dialects anywhere earlier in the article. Neither is any evidence provided that the punctual derivation spread is from Dullay to Konsoid (and not, for instance, vice versa). The conclusions as they are stated in section 9 are rather quotes of conclusions that were drawn by other authors in earlier publications and not the logical conclusions of the data provided by the author.


The joint article “Exploring the periphery of the Central Ethiopian Linguistic Area: Data from Yemsa and Benchnon” (pp.59-81) by Christian Rapold & Silvia Zaugg-Coretti examines the extent to which twelve allegedly contact-induced grammatical similarities that were proposed by Crass & Meyer (2008) for languages of the Highland East Cushitic (HEC)-Gurage contact zone are attested in two Omotic languages spoken outside this subarea of the Ethiopian sprachbund. While Yemsa is spoken in the neighbourhood of the HEC-Gurage contact zone, Benchnon is geographically further removed and (apart from Amharic) without direct contact to the languages Crass & Meyer investigated.


After an overview of earlier works on the Ethiopian sprachbund, and introductions into the sociolinguistic situations of Benchnon and Yemsa, Rapold & Zaugg-Coretti discuss Crass & Meyer’s features (e.g. the use of the ablative morpheme as a marker of the standard in comparative constructions, the use of the similative morpheme 'like' as a marker of complement and purpose clauses, the use of a periphrasis with 'know' to express an experiential perfect 'have ever V-ed') one by one and check whether they hold true for the two Omotic languages they have most expertise in. Apart from merely checking Benchnon and Yemsa for the (non‑)existence of these features, the authors also evaluate the diagnostic strength of the proposed features for the definition of the Ethiopian sprachbund as a whole or for subareas.


The results of the tests are as follows: In Yemsa, 7 out of 12 features are attested; Benchnon shares only 5 out of 12 features. Four features are shared by all languages in Crass & Meyer’s and Rapold & Zaugg-Coretti’s sample. Possibly only three of these four features are diagnostic for language contact between Omotic, Cushitic and Ethiosemitic: experiential perfect with 'know', benefactive focus with 'say' and the use of 'like' as a marker of complement and purpose clauses.  [3]


The article is a clear and easy read. It shows that too much of the work on the Ethiopian Linguistic Area has so far concentrated on languages in the centre of the country, largely ignoring languages from the periphery. Rapold & Zaugg-Coretti’s article is a welcome “reality check” and we need these kinds of tests to realise which proposed features are spread across which languages and language groups and/or how far the Ethiopian sprachbund actually extends. Finally, I’d like to point out two important problems the article alludes to and that future research on language contact has to take into account, namely (1) that some features proposed for the Ethiopian sprachbund are quite common cross-linguistically, or extend far beyond Ethiopia, and are therefore not good candidates for the establishment of the sprachbund, and (2) that some shared features are (almost) only attested in data elicited through the meta-language Amharic and could therefore be spontaneous translation effects of educated speakers but not necessarily common in natural data.


Sascha Völlmin’s short contribution “Some dialectal differences between Gumer and Chaha (Gurage)” (pp.83-95) takes stock of the phonological and phonetic details that differentiate the two closely related Western Gurage varieties Gumer and Chaha. There are slight pronunciation differences in the realisation of 3rd person copulas and 3rd person object suffixes between the two varieties. In the longest section of the article, the author examines in detail the (non-)gemination of second verb radicals, an analytical problem that has already attracted considerable attention in the study of closely related Gurage varieties (see e.g. Rose 2006). While gemination is almost entirely lost in the verb forms of Chaha, Völlmin shows that Gumer has retained geminated radicals in the same environments as Endegeñ, another Western Gurage variety, and that the degemination in Gumer is thus less advanced than expected.


Given that the dissimilarities between Gumer and Chaha that the author’s meticulous work revealed are very small, I doubt that it is justified to call Gumer and Chaha separate “dialects”, as the author does in his conclusion.


The last contribution, Silvia Zaugg-Coretti’s article “The morpheme =tu as a focus marker in Yemsa (Omotic) and Oromo (Cushitic)” (pp.97-120), deals with a case of direct morpheme borrowing between two neighbouring but unrelated languages. After giving a detailed overview of the focus-marking functions of Yemsa =tu, the author evaluates competing hypotheses about the diachronic source of the morpheme. She rules out the hypothesis that Yemsa inherited the focus marker from earlier stages of Omotic and she finds no evidence that it was grammaticalised in Yemsa from an earlier copula morpheme. She also considers it unlikely that Yemsa borrowed the morpheme from a Highland East Cushitic or Ethiosemitic language. Although few other structural influences of Oromo on Yemsa can yet be pinned down (but see p.115), she argues convincingly that the Oromo focus marker -tu is the most likely source of the Yemsa morpheme; on the one hand, due to the formal identity of the morphemes in the two languages and, on the other hand, due to the striking functional similarities of the morphemes, as a comparison of her Yemsa data with Dabala and Meyer’s (2003) Wellega Oromo data shows. The direct borrowing of the morpheme was facilitated by the intense language contact between Yemsa and Oromo.


The article is rich in data, clear and informative. The research questions that other linguists working on focus in Ethiopian languages could follow up on (p.117) are greatly appreciated. Finally, I have only one terminological remark: It is not clear to me why the author speaks of “medial verbs” in the Oromo section (e.g. page 108) but of “converbs” in the Yemsa section, given that the respective verb forms in both languages seem to be functionally similar.


The book is very well-edited; it contains only very few typos and errors in formatting. Only the following are worth noting: p.10, paragraph in the middle, “(7b)” should be “(7a)”; p.21, ex. (2), “näggä-rä-ɲɲ” should be “näggär-ä-ɲɲ”; p.56, first paragraph of section 9; “typological distribution” should be “geographical distribution”; p.91, sentence before table 7, “speaker L and B” should be “speaker L and H”.


The book is not as coherent as its predecessor (Crass & Meyer 2007) and the lengths of the contributions vary greatly. Not all articles address the topic of the book, language contact and language change, in any depth (only Rapold & Zaugg-Coretti’s and Zaugg-Coretti’s contributions actually centre on this topic). However, all articles make an original and significant contribution to a better understanding of the Ethiopian languages they deal with and thus provide us with important data and analyses for the discussion of contact-induced phenomena in the area. We need more contributions like the ones presented here, i.e. tests of the features proposed for the Ethiopian Language Area (Rapold & Zaugg-Coretti’s contribution), new assessments of data collected and hypotheses brought forward at the time when Ethiopian languages were first investigated (Binyam Sisay’s and Ongaye Oda’s contributions), as well as fine-grained and empirically well-grounded studies of individual grammatical phenomena (Meyer’s, Völlmin’s and Zaugg-Coretti’s contributions).


Last but not least, a compliment to the editors and/or publisher: The cover of the book is a beautiful, cliché-free photo of every-day life in Addis Ababa.


Cohen, David, Marie-Claude Simeone-Senelle and Martine Vanhove 2002

'The grammaticalization of ‘say’ and ‘do’: An areal phenomenon in East Africa.' In: Güldemann, Tom and Manfred von Roncador (eds.). Reported discourse: A meeting ground for different linguistic domains. Amsterdam, Philadelphia: Benjamins. Pp.227-251

Crass, Joachim and Ronny Meyer 2007

'Deictics, copula and focus in the Ethiopian convergence area.' (Afrikanistische Forschungen, 15.) Cologne: Köppe

Crass, Joachim and Ronny Meyer 2008

'Ethiopia.' In: Heine, Bernd and Derek Nurse (eds.). A linguistic geography of Africa. Cambridge: Cambridge University Press. Pp.228-249

Dabala Goshu and Ronny Meyer 2003

'Focus phenomena in Wellega Oromo.' Afrika und Übersee 86:161-197

Hayward, Richard J. 1982

'Notes on the Koyra language.' Afrika und übersee 65,2:211-268

Hirut Woldemariam 2004

A grammar of Haro with comparative notes on the Ometo linguistic group. PhD dissertation, Addis Ababa University

Rose, Sharon 2006

'Durational conditions on Endegeñ gemination.' In: Uhlig, Siegbert (ed.). Proceedings of the XVth International Conference of Ethiopian Studies. Wiesbaden: Harrassowitz. Pp.843-850

Sasse, Hans-Jürgen 1986

'A southwest Ethiopian language area and its cultural background.' In: Fishman, Joshua A., André Tabouret-Keller, Michael Clyne, Bh. Krishnamurti and Abdulaziz Mohamed (eds.). The Fergusonian impact: In honor of Charles A. Ferguson on the occasion of his 65th birthday. Volume 1: From phonology to society. Berlin: Mouton de Gruyter. Pp.327-342

[1] Note that the author points out on p. 25 that the theme argument of ‘say’ is not a “prototypical direct object constituent” – but the terminology is not reconsidered.

[2] One could hypothesise that the ‘say’ verb acquired this meaning when the actual ‘hit’ ideophone was omitted to leave a threat unsaid (but not misunderstood).

[3] The fourth feature, the use of different copulas in main and subordinate clauses, is also quite common beyond the Ethiopian sprachbund.




There are no comments yet.

Do you have any additions or comments?

Submit comment.

  1. Deutsch
  2. English
  3. Français