Abstracts for Welsh Syntax Seminar, Plas Gregynog, 2001

Emyr Davies 

I'd like to look at one specific type of coordination in Welsh involving the coordination of heads, particularly within PPs and VPs. It seems that prepositional heads and possibly verb heads cannot be conjoined if they share the same complement strings. This contrasts with examples in English like:

He walked to and from Llanelli
He wrapped and sent the parcel

Welsh has a restriction where the complement has to be adjacent to the first conjunct, forcing a pronominal complement on the second conjunct. I will then consider how to account for this restriction and what it suggests about coordinate structures generally.

 MORPHOLOGICAL DEVELOPMENT IN WELSH

Janig Stephens (UWIC, Cardiff)

 The early language acquisition patterns of Welsh speaking children indicate an initial stage without any finite verb (Aldridge, Borsley and Clack 1995) followed by an optional infinitive stage when children use both finite and non-finite verb forms (Borsley and Jones 2001) before they become regular in their use of inflected verbs.

 The Welsh copula 'bod' has a number of varied morphological realisations, determined by the structure of the clause. The earliest finite verb found in children's Welsh is the third person singular present tense "mae" or "ma". It remains the most frequent form, although other realisations emerge relatively early too. Two forms, 'sydd' and 'bod', are associated with subordination; 'sydd' is also found in interrogatives and topicalised structures as well as in relative clauses. They signal the development of subordination in the children's data.

 This study is based on a corpus consisting of 50 children's files, between 2 and 5 years of age, recorded in Welsh medium nursery schools in South Wales. The transcripts have been set in the CHILDES system and  CHAT formatted. They have been analysed using some of the CLAN programmes to identify the children's progress in the acquisition of the morphological realisations of 'bod'.

References

Aldridge M. , Borsley R.D and Clack S. 1995 The acquisition of of Welsh clause structure in Proceedings of the 19^th Annual Boston University Conference on Language Development. 2 Vols, Ed by D. MacLaughin & S McEwen, 37-47. Somerville, Mass: Cascadilla Press.

 Borsley R.D. & Jones B.M. 2001 The development of finiteness in early Welsh clauses (in Press) Journal of Celtic and Teaching

MacWhinney B. 1995 The CHILDES Project: Computational tools for Analyzing Talk.  Hillsdale New Jersey: LEA

PERFECT AND PROGRESSIVE CONSTRUCTIONS IN WELSH AND ENGLISH

Robert D. Borsley )University of Essex)

The perfect and progressive constructions of Welsh are superficially rather different from their English counterparts. Both involve the copula bod followed by what can be called an aspectual particle. (1) and (2) illustrate:

(1)   Mae Gwyn wedi   darllen y    llyfr.

is      Gwyn PERF read     the book

‘Gwyn has read the book.’

(2) Mae Gwyn yn       darllen y    llyfr.

is     Gwyn PROG read     the book

‘Gwyn is reading the book.’

However, more complex examples show important similarities to English. The data in (3) is parallel to that in (4).

(3)a. Mae Gwyn wedi   bod yn       darllen y    llyfr.

   is     Gwyn PERF be   PROG read     the book

   ‘Gwyn has been reading the book.’

     b. * Mae Gwyn wedi bod wedi darllen y llyfr.

     c. * Mae Gwyn yn bod wedi darllen y llyfr.

     d. * Mae Gwyn yn bod yn darllen y llyfr.

(4)a. Kim has been reading the book.

     b. * Kim has had read the book.

     c. * Kim is having read the book.

     d. * Kim is being reading the book.

This suggests that a unified account is desirable. Within Principles and Parameters theory one might attribute the facts to the order of functional categories PERF and PROG. However, the idea that be/bod in a progressive construction is a realization of a PROG category is problematic given evidence that there is a single be/bod taking a range of complements. The Welsh data poses further problems since it is not clear how to fit both bod and the aspectual particles into such an analysis. A very different approach to the English facts has been developed within Head-driven Phrase Structure Grammar (HPSG) by Anthony Warner, but this approach cannot be extended to Welsh. It is possible, however, to develop an HPSG approach based on Warner’s account of examples like (4d) which is applicable to both Welsh and English. If one assumes that the aspectual particles are heads taking a phrasal complement, giving structures like [wedi [bod [yn [darllen y llyfr]]]], the facts appear to require non-local selection. However, this can be avoided if one assumes: (a) that Welsh non-finite verbs can be ordinary infinitive forms, or perfective forms, or progressive forms, (b) that while most contexts require an infinitive form, wedi requires a perfect form and yn a progressive form, and (c) that wedi is a perfect element and yn a progressive element. Given these assumptions, the facts can be attributed to three constraints: one ruling out a perfective complement with a perfective copula, a second ruling out a perfective complement with a progessive copula, and a third ruling out a progressive complement with a progessive copula. Very similar constraints can be proposed for English given the assumption that perfective have is a copula verb. For Welsh, it would be possible to combine either the first and second constraints or the second and third. It seems, however, that neither move is possible in English.

HWYRACH 'PERHAPS' AND CLAUSE-NESS IN WELSH

Bob Morris Jones (Universitry of Wales Aberystwyth)

This paper examines examples such as the following:

and attempts to establish its syntactic characteristics on a descriptive level, and the extent to which it exhibits clausal characteristics in comparison with clauses such as:

This study compares the description of the hwyrach clause with contemporary standard analyses of Welsh, and identifies possible contentious issues as to what is a finite clause in Welsh.