Walker 1997
Walker, James. (1997). A lack of agreement: Celtic syntax meets the (revised) Minimalist Program. Cahiers Languisitiques d’Ottawa. 25
1. Introduction
• Main question: How does the MP handle Celtic?
• Concentrates on finite clauses, since their features are shared by most Celtic languages
o Most examples from SG
2. Issues in Celtic Syntax • While it is usual to label Celtic finite clauses as VSO, it is better to call them VSO(X), where X represents any number of constituents, such as NP complements, PPs, AdvPs, and so on
(1) a. Shuidh mi sat I ‘I sat.’
b. Sgrìobh an duine an litir. wrote DEF man DEF letter ‘The man wrote the letter.’
c. Is mise do dhuine. is(copula) I-reflexive your man ‘I’m your man.’
d. Chluich a’chlann anns an achadh. played DEF-children in DEF field. ‘The children played in the field.’
• A few things can precede the verb: NEG, COMP in embedded clauses
o Its been said that NEG and COMP occupy the same head
• If we assume that in VSO, the verb has moved to a functional head higher than V, then we have a few questions to answer:
o Q: What position does the finite verb occupy?
• A: (i) the verb goes to T or (ii) to C
• Evidence against V-T-C movement
o (1) No complementary distribution between V heads and COMPs in VSO
o (2) we may have adverbials preverbal in VSO embedded clauses
o (3) it is possible for negative polarity items to occur before the verb (cf. 2)
• Evidence for V-T-C movement
o It is impossible to place any constituent between the complementizer and the finite verb; the complementizer, the verb-stem, and the inflectional elements form a tightly bound complex, leading some to posit that they form a single prosodic unit.
• Phonological fusion could be used as an argument that the verb moves to COMP, with concomitant incorporation.
o The SOV word order of some non-finite clauses suggests that all the relevant constituents are raised out of the VP
(2) Aon deor amhàin níor bhain sé as an chorn. one drop one NEG-PAST took he out-of the cup ‘Not one drop did he take from the cup.’
o Q: What spec-position does the subject occupy? • A: Chomsky and McCloskey say that Subj remains in spec-VP at least until LF • Rouveret claims that while VSO and SVO both have V-T-C movement, VSO doesn’t have A-movement from spec-VP to spec-TP • Bobaljik and Carnie say that subjects must at least move to spec-TP to check for Case o Q: What mechanism differentiates VSO languages from SVO languages?
3. The MP and Celtic Syntax 3.1. Early Minimalism • McCloskey’s account of Celtic word order and agreement:
In (10) McCloskey argues that F1 corresponds to T and F2 corresponds to Agr.
3.2. Revised Minimalism • Revisions which have an impact on the previous analysis of Celtic o Agr is eliminated • We can compensate by using a light verb instead
o LF-movement is now movement of formal features only • The verb raises V-v-T before Spell-Out, and possibly T-C at PF, since it is prosodically driven • Morphological Merger as per Marantz 1989:99 may explain this PF process 4. Conclusion • The above analysis requires the MP to accommodate the notion that not all movement is motivated by the need to check features. It seems that some types of movement are dictated by prosodic requirements. • We need to examine other clause types to find if the basic assumptions here hold as well • A take home message: a change in theory should not invalidate all the observations of preceding theories.
