Grammaticalised movement expressions

Steve & Alison Nicolle steve-alison_nicolle at SIL.ORG
Sat May 10 08:35:41 UTC 2003

There seems to be a resurgence of interest in the expression and use of
space and motion cross-linguistically (e.g. the recent Paris conference),
and in relation to this I would like to test a few hypotheses that I made in
a book chapter that's just been published (Steve Nicolle: "Distal aspects in
Bantu languages" in KM Jaszczolt & K Turner (eds) Meaning Through language
Contrast vol 2, pp. 2-22).

What I am interested in is movement expressions (typically verbs of movement
and position such as 'go', 'come', 'arrive', 'pass by' and 'remain') which
have undergone grammaticalisation. There has been a lot written about
movement verbs (especially 'come' and 'go') becoming tenses through
grammaticalisation, but what I have found in some East African Bantu
languages is movement verbs that behave morpho-syntactically just like
tense/aspects but which have not shifted their meaning from physical
progression to temporal progression (i.e. 'go' still means 'move away from
the deictic centre' not 'future' but is no longer a verb syntactically).

In many Bantu languages the typical verb phrase consists of a verb with
prefixes for subject agreement and tense/aspect and an optional incorporated
argument (usually an object) agreement marker, and suffixes relating to mood
and valency change (passives etc) etc. Tense/aspect markers are typically
monosyllabic prefixes sandwiched between the subject agreement marker and
the verb stem: A-li-soma (3s-PAST-study) "S/he has studied" etc.
Movement verbs can occur in constructions like the following:
A-me-enda ku-soma (3s-PERF-go INF-study) "S/he has gone (in order) to

What often happens is that the movement verb gets attached to the following
verb, which loses its infinitive prefix, and a change in meaning occurs,
e.g. in Digo: A-nda-soma (3s-FUT-study) "S/he will study", where 'nda' is a
monosyllabic future tense marker derived from 'enda' (go). However, there
are also cases where the same phonological reduction and morphological
changes take place but the resulting 'tense marker' continues to encode a
meaning to do with physical movement (rather than metaphorical movement
through time). Digo has a few 'movement grams' which fit into the
'tense/aspect slot' in the verb phrase, and a few which are
morphosyntactically half way between verb and tense/aspect in that they are
disyllabic and co-occur with a tense marker, but are prefixed to a main verb
(without an infinitive prefix), e.g.
a-ch-enda-sgala (3s-NARR-GO-stay) "s/he went and stayed (at a certain
place)" (where 'ch' is a shortened form of the narrative tense 'chi', the
final vowel having been absorbed by the initial vowel of 'enda').

There are some hypotheses that I would like to test in relation to this kind
of phenomenon:
1) There is an implicational scale, such that if a language has movement
grams (the grammaticalised movement expressions described above) derived
from verbs such as 'arrive', 'pass by', 'remain' then it will also have
movement grams derived from the more 'basic' movement verbs 'go' and 'come'.
(At the moment I am assuming that 'go' and 'come' are basic at a purely
intuitive level.)

2) Languages which use a larger set of movement verbs as sources for
movement grams will allow more of these to co-occur in a single verb phrase
than languages which derive their movement grams from a more limited set of

3) Where a language exhibits a large number of movement grams these will
exhibit fewer of the formal features of grammaticalisation (e.g.
phonological reduction, affixation) than movement grams in a language with
fewer movement grams.

At a more general level, I would be interested to see how movement grams are
distributed among languages of different typological types.

If anyone has any data or insights into this  kind of phenomenon, I would be
very interested to hear from you.

Thanks, Steve

Dr Steve Nicolle,
Linguistics Consultant,
BTL (East Africa) and
Digo Language and Literacy Project

P.O. Box 12,
80403 Kwale,

(+254) 722 878 259

More information about the Lingtyp mailing list