Abstracts for the Ninth Welsh Syntax Seminar, Plas Gregynog, 2002

On the Nature of Welsh VSO Clauses

 Robert D. Borsley , University of Essex

It is widely assumed that Welsh VSO clauses such as (1) have a basic SVO structure.

(1) Gwelodd Ieuan ddafad.

      saw         Ieuan sheep

      ‘Ieuan saw a sheep.’

Thus, Roberts (forthcoming) notes that ‘the general consensus of work on Welsh’ is ‘that VSO clauses involve an operation which moves the verb out of VP to the left over the subject’. The basic idea is not limited to Principles and Parameters/Minimalist work, but is also found in LFG in Bresnan (2001). It could also be incorporated into HPSG through the distinction between constituent structure and order domains. Thus, the idea is available within a variety of frameworks. Is it a good idea? One strand of argument for an SVO analysis has highlighted the existence of VP-like constituents in non-finite clauses, such as the bracketed sequence in (2), and clauses involving the auxiliary gwneud ‘do’, such as (3).

(2) Disgwyliodd Rhian [i  Ieuan weld dafad]

      expected       Rhian  to Ieuan see   sheep

      ‘Rhian expected Ieuan to see a sheep.’

(3) Gwnaeth Ieuan weld dafad.

      did          Ieuan see    sheep

      ‘Ieuan saw sheep.’

Such examples might support an SVO analysis of VSO clauses if one assumed that all forms of a word must be associated with the same kind of structure, but there are reasons for scepticism about this assumption. An SVO analysis of VSO clauses might be plausible if gwneud clauses and non-finite clauses were like simple VSO clauses in all respects except word order. However, they also differ in the realization of agreement and negation. Thus, it is doubtful whether examples like (2) and (3) provide any support for an SVO analysis. Another type of argument for an SVO analysis involves contrasts between subjects and objects, especially the fact that a reflexive can appear in object position with an antecedent in subject position but not vice versa. This contrast might support an SVO analysis if binding theory needed to refer to constituent structures.  However, there is evidence in Welsh as in English that binding theory should refer to argument structure. Certain other facts might be seen as providing support for an SVO analysis, but on closer inspection they appear not to. Thus, there seems to be no real evidence for an SVO analysis. Certain facts involving negation and mutation appear to provide some evidence against such an analysis. It seems, then, that a simple VSO analysis, as in e.g. Borsley (1989), is preferable.

REFERENCES

 Borsley, R. D. (1989), ‘An HPSG approach to Welsh’, Journal of Linguistics 25, 333-354.

Bresnan, J. (2001), Lexical-Functional Syntax, Oxford: Blackwell.

How Many Aspects Are There In Welsh?

Bob Morris Jones, University of Wales Aberystwyth, bmj@aber.ac.uk

The data for this talk comes from Thomas (1996: 8190-97), who talks about yr ategion berfol ‘the verbal supports’. The ‘verbal supports’ include recognized aspect markers:

 There are nine other items and expressions which appear to occupy the same place as wedi and yn. For example:

 On this basis, there are eleven aspects in Welsh. Can this be so? This study considers whether this question can be answered.

 REFERENCE

Thomas, Peter Wyn (1996). Gramadeg y gymraeg, Caerdydd: Gwasg Prifysgol Cymru.

The Welsh 'Genitive of Respect' Revisited

Bob Morris Jones, University of Wales Aberystwyth, bmj@aber.ac.uk

This study examines a distinctive syntactic pattern in Welsh, and aims to determine whether it can be accounted for by productive syntactic rules or whether it is, at least in part, formulaic. This pattern is labelled by Morris-Jones (1931: 165) as the ‘genitive of respect’, a label which will be used here for convenience of exposition. Welsh does not have an overtly-marked case system, and Morris-Jones is referring to a sequence of categories, namely adjective + pronominal clitic + noun:

The study evaluates five possible explanations, namely that the 'genitive of respect' is:

o      a determiner phrase;

o      an adjective phrase;

o      the product of subject raising;

o      the remnant of a clause;

o      a partially formulaic construction.

REFERENCES

Morris-Jones, John, (1931). Welsh Syntax, Cardiff: The University of Wales Press Board.