194
6.3 Inversive and Inversive Stative
The inversive -or and inversive stative -ok are a pair of back-vowel suffixes which are common in many Bantu languages. The inversive is sometimes also called the reversive,
since it usually signifies an undoing or reversing of an action. Some of the clearest examples of this semantic relationship are shown in 159 below.
159 Basic and inversive verbs
7 8 F
. 8 F
. -7
8 1
9 2 8
1 9 2
.7 9
. 9
2 In each pair of words, the addition of the inversive suffix -o adds the meaning
component of “undoing” or “reversing” the basic action. The stative counterpart -ok also signifies a reversal of the action, but it is intransitive. Note that throughout this thesis
I sometimes refer to both the inversive and inversive-stative suffixes generically as the “inversive” since their harmony patterns are identical.
The majority of commonly-used inversive and inversive-stative verbs have been lexicalized, taking on a slightly different or more specific meaning than the combination
of their component parts. Oftentimes the original verb root no longer occurs by itself in the language. Sometimes a verb will have both inversive and inversive-stative forms, but
other times only one or the other is deemed grammatical by native speakers. Because the inversive extensions are not nearly as productive as the applicative, causative or passive
extensions, there is a much smaller set of data available to explore vowel harmony in these extensions. Nonetheless, the following generalizations can be made. There are three
195 alternations of the suffix, and the alternations exhibit frontback asymmetry, which is not
uncommon in Bantu verbal extensions see especially Hyman 1999. The pattern is presented in 160 below.
160 Inversive vowel harmony
-o -ok following i, e, , o root vowel:
e ? -u -uk
following u, a - - k
following As the pattern shows, all front-vowel verb roots have the extension [-o ], but back-vowel
roots cause height harmony in the suffix. The extension always agrees with u o in both height and [ATR]. Another interesting and important point is that there is only one
potential example of the inversive suffix following in the root, but it is difficult to be sure about the basic verb root from which it was derived. This issue is dealt with in detail
further below. Examples of the inversive in combination with each different root vowel are
shown below. Because the inversive and inversive-stative exhibit the same harmony behavior, and because of the limited number of examples of each, I include them all
together in the following data sets. I include the majority of examples of inversive verbs in my data in order to show the well-attested patterns and also to point out the patterns for
which I have fewer examples. For the sake of clarity, I indicate the morpheme break between the root and the extension in all examples, but recall that many of these forms
are lexicalized, and that this morpheme division is perhaps only historical, not synchronic.
The examples in 161 show verbs with the high-vowel realization [-u ] or [-uk], all of which have either u or a in the root.
196 161
[-u ] following u a N K
N K N K
N K ST
2 ST
2 ST
2 ST
2 N
N N
N 8 F
. 8 F
. 8
2 8
8 8
; 2U :
8 8 - -
8, - 2 8
8 8
89
+
8
8 -
U 8 1
. .
Where possible, I list probable related forms. Some have a clearer semantic connection than others. For example, it is possible to imagine that ‘move’ and ‘boil’ could be related,
but this connection is not quite as clear as that between ‘cover’ and ‘uncover.’ The [-o ] and [-ok] realizations of the extensions occur following i e o root
vowels. Examples involving each of these root vowels are in 162 below. Note that roots might have this suffix alternation as well, as I discuss further below. Fewer
examples are given for e and o roots simply because there are fewer examples in the data.
75
The [i-] prefix before the root in [kw-i-naa -a] is the reflexive prefix, which is used to derive new verb stems. Like the inversive, there are a number of examples in my data of verbs with the reflexive prefix.
Most of these words seem to have been lexicalized.
197 162
[-o ] following i e o N K
N K N K
N K ST
2 ST
2 ST
2 ST
2 N
N N
N 8
1 9 2
8 1
9 2 9
2 9
. - 8
- 8
. 9
Q 9
- 8 2
8 8
- - 8
8 .
2 The third form of the suffix, [- ], follows roots only. The few examples
available in the data are in 163 below. 163
[- ] following N K
N K N K
N K ST
2 ST
2 ST
2 ST
2 8
- 8
. 8
8 It is possible that the first example above, [ u-
- k-a], is actually composed of two inversive suffixes, since it is not uncommon for Bantu verbs to have multiple derivational
extensions. Despite the fact that there are not clearly related forms at least in my data for
most of the words above, it is probable that most if not all of these forms did indeed develop historically by the addition of the inversive extension. My confidence is based on
198 the fact that there are no exceptional forms in my data, that is, no forms which look like
they might contain an inversive but which have a different harmony pattern than that which I describe here. For example, there are no verbs with a front vowel in the first
syllable of the stem and what appears to be the high-vowel inversive variant [-u ] or [- uk]. Instead, all front vowels in the first syllable are followed by [-o ] or [-ok].
Finally, there is one root vowel for which I have only a single interesting example. As mentioned above, there is only one potential example of the inversive suffix
in combination with an underlying root, shown in 164 below. 164
Root vowel with inversive ST
2 ST
2 ST
2 ST
2 U
2 The origin of the form [ o-te -o -a] is difficult to verify for a number of reasons, most
notably because of both prefix and root vowel difference between the two forms. There does seem to be a reasonable connection between the meaning ‘to trap’ and ‘to remove,’
but not all Ikoma speakers agree that these words are related. And the fact that the stem vowel changes from to [e] makes it difficult to confirm that these forms have the same
historical root, especially since there are no other root examples which confirm that mid-vowel raising actually occurs with the inversive.
The possibility that an [ATR] alternation occurs here is a tempting analysis, since we have seen similar patterns involving another [+ATR] mid-vowel suffix i.e. the
subjunctive -e which affect the same [-ATR] front vowel . However, even though these parallel patterns are very interesting, we must note that there is a significant
difference between the two as well. In the subjunctive examples, as with all other root-
199 vowel alternations, the root vowel change does not affect the height of the prefix see
§4.2 and §5.1.1 above and §6.4 below. The general pattern elsewhere in Ikoma is that even though the root alternates to [+ATR], the prefix still behaves as if the root is
[-ATR]. However, in the case of [ o-te -o -a], the prefix is mid, not high. It is possible that this form has long been lexicalized and now conforms to more typical prefix
patterns. Or, it is possible that [ u-t -a] and [ o-te -o -a] are not actually related.
Because of these uncertainties, we really need more data before positing this type of root vowel alternation.
Finally, note that the inversive can be combined with the causative suffix, which is the topic of §6.6 below. There we see that the presence of the causative affects the
inversive’s normal harmony patterns. Specifically, the [-u -a] alternation becomes [-o -i] in the inversive-causative, but only following [a] root vowels, not following [u] root
vowels.
6.4 Causative, Passive, Perfective and Agentive