Subjunctive Clauses. the probehood of lower functional heads Next we add to the picture data from agreement in what we refer to as subjunctive clauses. Although there is no special subjunctive morpheme in Ibibio, the clauses in question function as the complements of verbs like yem ‘want’ and have subjects that are disjoint from the subject of the matrix clause (rather than controlled PRO subjects, as in (42c)). Some examples are: ▇. ▇▇▇▇ a-▇▇▇ ▇▇▇▇ a-si-nam. ▇▇▇▇ 3sS-want Emem 3sS-IMPF-do ‘▇▇▇▇ wants Emem to be doing it.’ ▇. ▇▇▇▇ a-yem (ɲɲin) i-di. ▇▇▇▇ 3sS-want we 1pS-come ‘▇▇▇▇ wants us to come.’ c. Ami n-yem afɨt owo e-kpa. I 1sS-want all person 3pS-die ‘I want everyone to die.’ We call these subjunctive clauses by comparison with Romance languages, in which the complement clause in sentences like these would be in the subjunctive mood. The first structural property of subjunctive clauses to notice is that the verb displays normal phi-feature agreement with the subject, as is already evident in (48). In this respect, these clauses are like finite indicative clauses in Ibibio, rather than like infinitival clauses. The second notable property of these clauses is that they have no overt Tense head. The embedded verbs in (48) consist of a verb root, an agreement prefix, and in one case an aspect head ((48a)), but there is no visible tense morpheme. The examples in (49) show that it is impossible to have an overt T morpheme in the complement of a verb like ‘want’.
Appears in 2 contracts
Sources: Research Paper, Research Paper
Subjunctive Clauses. the probehood of lower functional heads Next we add to the picture data from agreement in what we refer to as subjunctive clauses. Although there is no special subjunctive morpheme in Ibibio, the clauses in question function as the complements of verbs like yem ‘want’ and have subjects that are disjoint from the subject of the matrix clause (rather than controlled PRO subjects, as in (42c)). Some examples are:
▇. ▇▇▇▇ a-▇▇▇ ▇▇▇▇ a-si-nam. ▇▇▇▇ 3sS-want Emem 3sS-IMPF-do ‘▇▇▇▇ wants Emem to be doing it.’
▇. ▇▇▇▇ a-yem (ɲɲin) i-di. ▇▇▇▇ 3sS-want we 1pS-come ‘▇▇▇▇ wants us to come.’
c. Ami n-yem afɨt owo e-kpa. I 1sS-want all person 3pS-die ‘I want everyone to die.’ We call these subjunctive clauses by comparison with Romance languages, in which the complement clause in sentences like these would be in the subjunctive mood. The first structural property of subjunctive clauses to notice is that the verb displays normal phi-feature agreement with the subject, as is already evident in (48). In this respect, these clauses are like finite indicative clauses in Ibibio, rather than like infinitival clauses. The second notable property of these clauses is that they have no overt Tense head. The embedded verbs in (48) consist of a verb root, an agreement prefix, and in one case an aspect head ((48a)), but there is no visible tense morpheme. The examples in (49) show that it is impossible to have an overt T morpheme in the complement of a verb like ‘want’.
a. ▇▇▇▇ a-▇▇▇ ▇▇▇▇ a-ke-nam. ▇▇▇▇ 3sS-want Emem 3sS-PAST2-do ‘▇▇▇▇ wants Emem to have done it.’
b. ▇▇▇▇ a-yem Emem a-di-nam. ▇▇▇▇ 3sS-want Emem 3sS-FUT2-do ‘▇▇▇▇ wants Emem to do it (in the future).’ One conceivable interpretation of this observation is that verbs like ‘want’ select a complement that is in the simple present tense. That would be superficially consistent with (48)-(49), because the simple present tense is realized phonologically as Ø, as seen (again) in (50). ▇▇▇▇ a-▇▇▇ ▇▇▇▇. (also: i-yem … , e-yem… , etc.) ▇▇▇▇ 3sS-seek goat. (1pS-seek, 3pS-seek) ‘▇▇▇▇ is looking for a goat.’ (‘We are looking…’ ‘They are looking…’) However, including negation in the clause reveals a structural difference. In the simple present tense, negation shows up as a suffix on the verb, as in all other finite indicative clauses in Ibibio: ▇▇▇▇ i-▇▇▇-me ▇▇▇▇ ▇▇▇. ▇▇▇▇ I-look-NEG goat the ‘▇▇▇▇ is not looking for the goat.’ We take this to be an indication that the verb moves to T in Ibibio, and thus surfaces to the left of negation, rather than to its right, much as in the famous case of French (▇▇▇▇▇▇▇ 1989). The null present tense T triggers this sort of verb movement in just the same way as overt past and future Ts do (compare i-k-i-yem-me ‘s/he did not look for it’; i-▇▇-▇▇▇-me ‘s/he will not look for it’, with tense markers ki ‘past2’ and di ‘future2’). There is, however, no such movement of the verb past negation in subjunctive clauses in Ibibio. Rather, the negative particle ke shows up as an independent word to the left of the verb in (only) this sort of clause.16
▇. ▇▇▇▇ a-▇▇▇ ▇▇ Emem a-si-nam. ▇▇▇▇ 3sS-want NEG Emem 3sS-IMPF-do ‘▇▇▇▇ wants Emem not to be doing it.’
▇. ▇▇▇▇ a-ke-bo ke ɔmmɔ e-dep ebot. ▇▇▇▇ 3sS-PAST-say NEG they 3pS-buy goat ‘▇▇▇▇ said that they should not buy a goat.’ Subjunctive clauses thus have somewhat different structures from main clauses in Ibibio. The best analysis of these facts, we claim, is that the T node is simply absent in subjunctive clauses in Ibibio. Verbs like yem ‘want’ and bo ‘say to’ select for an AspP or other lower functional projection, rather than for a TP. This expresses the impossibility of there being an overt tense morpheme in examples like (49) in a straightforward way. It also accounts for the absence of verb movement past negation in (52), because the trigger for verb raising is missing in these clauses. Finally, it explains the position of the subject relative to negation in (52). The normal order of morphemes in Ibibio is Subject – T+Verb – Neg – Other, the subject coming before both negation and the finite verb (see (51)). The position of the subject before the finite verb is presumably due to an EPP feature on T that triggers raising of the subject from Spec, vP to Spec, TP. As a side effect of this raising, the subject ends up before the negative particle as well. The sentences in (52) are thus unusual in that the subject comes after negation. This also follows from saying that there is no T in subjunctive clauses in Ibibio: there is thus no functional head higher than negation in these clauses—nothing that could bear an EPP feature that would cause the subject to move past negation. Saying that subjunctive clauses in Ibibio have no T thus captures a cluster of properties that distinguish these clauses from others in the language.17 Now recall that, despite the absence of T in subjunctive clauses in Ibibio, there is still ordinary-looking phi-feature agreement with the subject in (48). This must be a manifestation of agreement on one of the lower functional heads in Ibibio—for example, on the Aspect head in (48a). In this situation, agreement on the lower functional heads clearly does not depend on there being agreement on T. On our favored view, an example like (48a) can be given a representation like the one in (53). [ ▇▇▇▇ [VP want [ Neg [AspP Emem Asp [vP <Emem> [VP do it ]]]]]
move 1 Agree 2 This analysis uses the additional assumption in (54), which we have tacitly assumed throughout. The highest verbal functional head in an Ibibio clause has an EPP feature.
Appears in 1 contract
Sources: Research Paper