TAM/NEG MARKING IN THE BANTU LANGUAGES

The inflectional morphology of the Bantu verb has been the subject of many previous comparative and descriptive studies. However, by comparing morphological templates from a wide variety of Bantu languages, which has not been done systematically before, it is possible to make intriguing generalizations which have hitherto gone largely unnoticed.

The structural data that is of interest here comprise inflected verb forms, in fact any finite verb form marked for tense, aspect, mood, focus, negation, whatever, be it by inflectional morphemes, auxiliaries, or both. The morphemes in question are those which do not occupy the SM (subject marker), OM (object marker) or STEM (verb stem, root) slots in a so-called slot system (morphological template). These typically include various TAM (tense-aspect-mood) and NEG (negation) markers, and occasionally others, e.g. FOC (focus) markers. Hence the SM, OM and STEM slots can, for present purposes, be seen as morphosyntactic pivots. The relevant parts are what appears before the SM (the initial markers), between the SM and OM (the medial markers) and after the STEM (the final markers).

Let us look closer at the inflected one-word forms nilikisoma 'I read it' from Swahili, tutáaluile 'we did not fight' from Lungu, and the periphrastic constructions nicaali kugura 'I have not bought' from Ruri and naáré kóha ' I was giving' from Kuria. These four verb forms can be morphologically segmented as follows:

The above examples involve several sets of markers that are cognates. One such set comprises Swahili li (inflection), Ruri li, and Kuria ré (auxiliaries), all deriving from PB *di. Another cognate set is made up of the negation markers ca in Ruri and tá in Lungu. A third set is composed of the past tense markers a in Lungu, Ruri, and Kuria.

The remarkable thing with the examples above is that the linear ordering of the various markers are comparable across the data. What is even more remarkable is that by applying the same process to an increasing amount of languages and data, we eventually arrive at a slot system applicable to the Bantu languages in general. I dub the resulting mega-slot system simply as the Pan-Bantu Slot System, or PBSS for short.

In its current form, PBSS recognizes 16 medial slots (labelled M1 through M16), seven final slots (F1-F7), and at least four initial slots (I1-I4); see tables immediately below. However, note that these are based on generalisations, and some/much work still needs to be done to improve details.

The slots in the above tables include various items, These items consist of information given on two lines. For instance, the first item in slot F1 (see table immediately above) is [ i(C)e PERF,PAST ], the first item in F4 is [ a *a ]. The first piece of information (in bold-print) gives the most commonly found form (V = any vowel, C = any consonant, N = any nasal, ... = various forms), the second gives either its etymological origin (e.g. *a) or its most wide-spread grammatical function (given with standard linguistic abbreviations; though ... = assorted functions).

The easiest way to understand the PBSS is to see it as a predictive tool. If one were to pick, from any random Bantu language, a random finite verb form involving two or more, say, medial markers, then the PBSS will predict the linear ordering of the markers in question. In fact, the PBSS performs remarkably well. Currently it generalizes over data from some 120-ish Bantu languages, totalling roughly 3,200 morphological templates. Appr. 95% of the data display expected morphosyntactic behaviour. Or conversely, only appr. 5% of the data directly flout the generalizations captured in the PBSS.

(More to be added.)

Jouni F. Maho. 1999. A (tentative) verb slot system for Shona. Unpublished report for the ALLEX (African Language Lexicon) Project. Dept of Oriental and African Languages, Göteborg University. Pp 15+2.

Jouni F. Maho. 2005. Book notice for 'A glossary of terms for Bantu verbal categories' by Rose, Beaudoin-Lietz & Nurse. Language, v. 81, p. 786-787.

Jouni F. Maho. 2007. The linear ordering of TAM/NEG markers in the Bantu languages. SOAS working papers in linguistics, v. 15, p. 213-225.
  [ complete text PDF ]