Coordinate Subjects, Expletives, and the EPP in Early Irish

Abstract

This paper examines subject-verb agreement in Early-Irish sentences with coordinate subjects. We claim that Early Irish (Old and Middle Irish) is a 'variable agreement' language, which exhibits both singular and plural agreement with coordinate subjects. The type of agreement depends on adjacency between subject and verb and the valency of the verb. In particular, unaccusative and passive verbs exhibit both singular and plural agreement more frequently than transitive verbs. We argue that this is due to the availability of a default third person singular null locative expletive item, which controls singular agreement. Moreover, unaccusative and passive verbs also allow locative inversion with other PPs, leading to the same singular agreement. Furthermore, we suggest that, in contrast to Modern Irish, which lacks such an expletive, Early Irish could license its presence in intransitive/passive sentences because that stage of the language exhibited EPP-effects.

Transitivity and Subject Positions in Old Irish

“…Properly identifying FP is beyond the task of this paper. In brief, it may be a modal projection as suggested in Lash and Griffith (2018) and Lash (), or one of the various subject‐hosting projections discussed in Cardinaletti (, ) and Haeberli () (see also Biberauer & Kemenade for an overview). PolP = polarity phrase and CP is the complementizer phrase which hosts the verb after verb movement.…”

“…These five examples have both an apparent subject ní ‘something’ and an apparent object: Ailill, Medb, Fergus, Ailill, and Fergus respectively. One example of a passive verb with an infixed indirect object pronominal was also omitted, namely, 108.3560 ( dommárfas ‘was shown to me’) (for the construction, see Thurneysen : 50–1, Lash & Griffith : n.15 and n.16). In Ml., there are no such examples.…”

“…(31b) in Lash (: 288). Lash & Griffith (: 122–6) argue that the subject is in fact in its base position within the eventive core in these kinds of examples.…”

“…It also showed that there are thirty‐six examples of absolute clause‐final subjects where the verb is non‐transitive. Clause‐final subjects that follows multiple intervening elements are probably involve ‘extra‐clausal’ right‐dislocation because there only so many possible intraclausal positions available for the intervening elements and for the subject itself there is at most three: Pos1, Pos2, and its base position within the eventive core (for further arguments in favour of a right‐dislocation analysis of this type of sentence, see Lash & Griffith ). Examples of this kind are therefore excluded from the table below and are not discussed here, although it is an open question whether the omitted examples would conform to the analyses adopted here; I leave this for future research .…”

“…For an overview of some of the characteristics of intransitive as opposed to transitives clauses, see Levin & Hovav (). Indeed, Lash & Griffith () have shown that unaccusative intransitives and passives in Old and Middle Irish are distinct from transitive verbs in occasionally allowing their subjects to remain in their base position within the eventive core, while the usual subject position in the space between the initial verb and the eventive core is, in such cases, occupied by a null expletive. In light of such a difference based on the transitivity of the clause, the generalization offered in (14) would not be surprising.…”