1Liliane Hodieb’s monograph is the first comprehensive grammatical description of Wushi (also known as Babessi), a South Ring Grassfields Bantu language of North-West Cameroon. The monograph is organized in five chapters preceded by an introduction and followed by a conclusion, and an appendix. In the following paragraphs, I summarize and comment on each chapter specifically in the light of some comparative issues related to Ring languages in particular, and to the larger Grassfields Bantu group in general. I then provide some general comments at the end.
2The unnumbered introduction opens with a discussion of the Cameroon “Anglophone crisis” (p. 17-18), which made fieldwork impossible, obliging the author to collect data from a single consultant via social media platforms. The author observes that the crisis contributes to the endangerment of Wushi since many of its speakers have been forcefully displaced to other, French-speaking parts of Cameroon and over to neighboring Nigeria, justifying the need for a grammatical description of the language while it is still possible to gather some data. In §4, geographical, linguistic, and political information about Cameroon is provided, followed by a presentation of the Grassfields region and its languages in general and Wushi in particular. In §5, the existing literature on the distinction between Bantu and Grassfields Bantu languages as well as justification for the classification of the latter is presented. Hodieb is of the opinion that to reach a definitive understanding and classification of Grassfields Bantu languages, detailed descriptions of individual languages, such as Wushi are needed. Included in this section is also a brief presentation of research questions as well as the theoretical frameworks employed, i.e., Structuralism and Cognitivism (p. 46-47). Unfortunately, not all major theories employed in the study are presented here. This is especially the case of Autosegmental Phonology, which is utilized in Chapter 2 for tonal analyses. It is also worth pointing out that the relevance of some of the research questions is not clear, e.g., the question whether Wushi has dialectal variants is not addressed later and, indeed, comes as a surprise because Hodieb herself already declared that “aucun dialecte de la langue n’est connu à l’heure actuelle” (p. 29).
3Hodieb discusses the phonology of Wushi in two separate chapters, beginning with segmental phonology in Chapter 1 (p. 49-87) and tonology in Chapter 2 (p. 89-134). In Chapter 1, the author presents a rich and complex phoneme system of 28 consonants, including two labial-velar stops, /kp/ and /gb/, and 12 vowels, which all have long phonemic counterparts. Secondary articulation, involving palatalization, labialization, prenasalization, spirantization, and aspiration is also frequent and described in detail. In addition to /l/, Hodieb unexpectedly describes /j/ as a lateral consonant (p. 52, 57) rather than an approximant, as in other Grassfields Bantu languages. Since she also identifies a labial-velar approximant /ɥ/ in a few words, the phonemic status and distribution of /j/ and /ɥ/ could have been clarified. The high number of 12 phonemic vowels is unusual for a Grassfields language expected to have a maximum of 10 vowels (Watters 2003: 234) and, therefore, would have deserved some additional commentary and justification. In this regard, appropriate acoustic evidence may have been helpful in the identification and distinction of the vowels. Likewise, the high number of diphthongs sometimes involving the same vowels that either devocalize or palatalize when followed by other vowels further complicates the analysis. The complexity of the segmental system is attributed partly to spirantization, which is shown to be in process (Janson 2007) and more frequent in Wushi than in other Ring languages. Like other Ring languages, Wushi mostly has monosyllabic syllables, and disyllabic words are predominantly due to compounding or derivation (p. 49-50). Syllables can be open or have a glottal coda. The most common syllable type is CV(ʔ), while NCV(ʔ), and CVV(ʔ) are also possible for monosyllabic roots and CVCV(ʔ) for disyllabic roots.
4Chapter 2 focuses on tone in both the nominal and verbal systems. Four underlying tones: H, L, HL, and LH are proposed for Wushi, departing from other studies of Grassfields and Bantu languages, which have generally postulated only two underlying tones: H and L (Watters 2003; Odden 2014: 201). However, the author points out that contour tones are rare in the language (p. 98, 109), and evidence for the phonologization of contour tones is insubstantial. The study would have benefitted from more convincing arguments for underlying contour tones, especially because it appears that several contour tones in the language result from the combination of level tones, mainly when a syllable is lost but its tone survives (p. 105). Tonal processes, i.e., downstep, tone spread, tone assimilation and dissimilation that result in H, L, M, HL, LH, HM, ꜜM and ꜜH surface tones are presented. A more detailed discussion of the phonetic distinction between M vs. ꜜH and L vs. ꜜM tones would have been welcome particularly because Ring languages, except Babanki (Akumbu 2019), tend to have either a M or ꜜH tone, but not both (Hyman 1979: 176-177). The author demonstrates that tone spread occurs from noun roots to underlying toneless noun class suffixes. Tonal processes targeting verb roots are common in context and the H tone undergoes more changes than the L tone. The reverse is true in the noun system, where the L tone is affected more than the H tone.
5Chapter 3 (p. 135-173) covers nominal morphology. Wushi has a fully operational noun class system of the Grassfields Bantu type. Nine noun classes are identified, six of which are marked by suffixes, one by a prefix, while the remaining two do not have any markers. Among Ring languages, Wushi appears to have gone furthest in losing its nominal prefixes and taking up suffixes, except in class 2 which still has a prefix. The class 2 prefix is most likely a retention, reflecting the old nominal prefix system of the language. A closer examination of Wushi’s suffixes is, nonetheless, required. In Ring languages where nouns regularly drop their noun class prefixes under modification, “suffixes” are analyzed as determiner enclitics attached under conditions of “de-focalisation” and representing the final bracket of an NP, e.g., in Kung (Kießling 2019). Concord marking in noun phrases is described and subcategorized based on how concordial agreement is implemented on possessives and demonstratives vs. how it is marked on adjectives and associative constructions (summarized on p. 160). In addition, the author argues that classes 1 and 3 are in the process of merging, with only a single difference in their concord system seen on the 1sg possessive marker (p. 143-145). Furthermore, it is observed that locatives (p. 146-147) do not fall into any of the Wushi noun classes. Instead, they are marked by independent prepositions that resemble Proto-Bantu forms but do not show agreement with the accompanying nouns. The chapter closes with a comparative study of a few Ring languages, which reveals among others that while some noun class markers in Ring languages can be V, Wushi lacks V all alone as a noun class marker. At the same time, Proto-Ring vowels in noun class markers have all been reduced to schwa in Wushi.
6Chapter 4 on verbal morphology (p. 175-232) first describes the structure of verb forms before turning to aspect and mood distinctions. Contrary to Robinson (2021), Hodieb argues that Wushi is an aspectual language, where, unlike other Ring, Grassfields, and Bantu languages, tense is not marked morphologically. The aspectual distinctions recognized and described include the perfective marked by a HL tone associated with the verb or predicate, the imperfective marked by nɔ̌, the anterior expressed by nà(ʔ), the potential marked by wɔ̀(ʔ), and the dissociative or distant marked by kə̀. Regarding mood, the infinitive, imperative, conditional, and indicative are identified and presented. The verbal extensions found to occur in Wushi are -tə̀ ‘pluractional and repetitive’, -sə̀ ‘causative’, -kə̀ ‘repetitive and iterative’ -nə ‘qualitative and expression of states’, and -mə̀ considered to be marginal and disappearing in the language. Hodieb discusses various criteria for identifying serial verb constructions (SVCs) in a language and concludes that “les séries verbales ne semblent pas être attestées en wushi” (p. 218) and that a study of stories and a variety of texts may be required to identify SVCs. However, earlier on, Robinson (2021) had identified and described both symmetrical and asymmetrical SVCs in the language, raising doubts about the current analysis, which does not consider Robinson’s study at all. This is probably an indication that very few revisions were made between 2020 when the thesis was defended and 2023 when the monograph was published.
7Chapter 5 on syntax (p. 233-293) provides a description of the syntactic structure of noun phrases and sentences. Canonical word order is SVO (p. 275), although it can be slightly modified to VSVO by repeating the verb when the subject is focalized. In the noun phrase, modifiers follow the head noun, with the exception of numerals and some rare stative verbs such as ŋkáʔ ‘be big’ and vɛ́ ‘be small’ that precede the head noun. Among Grassfields languages, Wushi stands out as a language whose numerals precede rather than follow the head noun. Due to the lack of agreement marking between the noun and the preceding numeral, Wushi numerals are described as having verbal or predicative function rather than acting as nominal determinants (p. 247). It is not clear whether numeral modifiers could be analyzed as relative clauses or not. Ordinal numbers show a mixed ordering, with ‘first’ and ‘second’ preceding and other ordinal numbers following the head noun. Hodieb demonstrates that there is no word class of adjectives in Wushi and that stative verbs are used to express properties in the language. To do so, the verbs receive an i- prefix, which allows them to function attributively and predicatively (p. 238, 270-274). It is also demonstrated that nouns lose their affixes to stative verbs that express qualities. However, some modifiers, e.g., kʉ̀ː ‘weak’ (1), take suffixes other than those of the head noun, while others, e.g., fòː ‘white’ (2) occur without suffixes while the head noun retains its suffix.
(1)
|
a.
|
ɣá: kʉ̀:-nə̀ mɛ̂
|
|
|
ɣà:-ø
|
kʉ̀:-nə̀
|
mɛ̂
|
|
|
body-cl3
|
weak-cl4
|
det
|
|
|
‘weak body’
|
|
b.
|
ɣà:sə́ kʉ̀:-nə̀ mɛ̂
|
|
|
ɣà:-sə́
|
kʉ̀:-nə̀
|
mɛ̂
|
|
|
body-cl10
|
weak-cl4
|
det
|
|
|
‘weak bodies’
|
(2)
|
glâsə́ fòː mɛ̂
|
|
glâs-sə́
|
fòː
|
mɛ̂
|
|
glass-cl10
|
white
|
de
|
|
‘white glasses’
|
8These examples suggest that the head noun or modifier relate with their affixes in specific ways that need to be further investigated and properly characterized. It is probably the case that depending on the noun class involved, nouns (and modifiers) may drop or retain their affixes under modification, as in some other Ring languages, e.g., Aghem (Hyman 2010) and Isu (Kießling 2010). This observation can be extended to associative constructions where affixes show similar complex relations depending on whether N1 and N2 have a prefix, a suffix or neither.
9The conclusion, which like the introduction is not numbered, summarizes the main points discussed in each chapter and points to what the author calls “curiosities” or “particularities” of Wushi. The first phenomenon highlighted is polysemy, although I do not find this to be unique to Wushi. Hodieb points out that several words and morphemes may have up to three distinct meanings in the language, e.g., tʃʊ́ə̀ ‘acid; wrist; sharp’ and that only the context of use can help to disambiguate such items. The second particularity is reduplication, which is generally used for intensification in the language, although there are also a few lexical items whose base forms are reduplicated. The third point, which the author recognizes as a common strategy in Ring languages, is the use of the verb ljɔ̌ ‘pass’ to express comparatives and superlatives in Wushi. The fourth noteworthy feature is that fô ‘white’, fê ‘black’, and bà ‘red’ are the three basic colour terms of Wushi and that the remaining colours are expressed by English loans, e.g., blû ‘blue’. I think that the three basic terms are shared by Grassfields languages, but it seems that, unlike Wushi, most of the languages also have terms for a number of other colours. Hence, borrowing is likely a result of contact and the overall influence of English. Finally, Hodieb reviews the migratory history of the Grassfields, characteristics of languages of the region, and the language contact situation. She discusses common phonological features of Bantoid languages and observes that Tikar, Mambiloid, and Grassfields languages are classified as Bantoid because they show traces of correspondences with Proto-Bantu but also have irregular and unusual features that set them apart from Bantu.
10The appendix (p. 327-349) includes a Wushi-English lexicon. Although the thesis is written in French, the appendix comes in Wushi and English as a first step to meeting one of Hodieb’s objectives, which is to produce pedagogical material for Wushi, a language spoken in Anglophone Cameroon.
11I have identified some issues, particularly with the analysis of vowels, tones and the verb system that would require more data and further analysis in the future, a future which one hopes will allow for in-the-field research in the North-West Region of Cameroon. One thing that should have been adjusted in the monograph is the numbering of chapters, sections, and sub-sections. As it stands, the numbering does not reflect the internal organization of the respective chapters and is quite confusing, e.g., all chapters have a section identified as 2, and it is not immediately evident whether one is dealing with 1.2, 2.2, 3.2, 4.2, or 5.2. For the same reason, it would have been helpful to number the introductory and concluding chapters as well.
12Liliane Hodieb must be commended for carrying out work on her thesis under the stringent conditions imposed by the ongoing Anglophone crisis in Cameroon that began in 2016 in the region where Wushi is spoken, preventing her from doing fieldwork, and by the COVID-19 pandemic which made the finishing phases of library research impossible because libraries in Paris were closed. Despite the odds, she was able to present a detailed description of the phonology, morphology, and syntax of Wushi. Her monograph is a very welcome contribution to descriptive grammars of Grassfields Bantu languages for a number of reasons. First, the quantity and quality of data collected from a single speaker through social media is impressive and the depth of the analyses presented is laudable. This is seen, for example, in the detailed discussion of the causative, which also builds on evidence from several Grassfields Bantu languages. Secondly, the book contributes to the ongoing debate on Niger-Congo noun class systems by demonstrating that Wushi has lost noun class prefixes and gone furthest among Western Grassfields languages in taking up noun class suffixes. This study has significantly increased our knowledge of Grassfields Bantu languages, and it has the potential of serving as a model for other grammars of languages of the region and beyond.