Gjesteforelesninger: David Adger
Forskergruppe i teoretisk lingvistikk og Tekstlaboratoriet inviterer til gjesteforelesninger med
David Adger, professor i lingvistikk,
Queen Mary, University of London 17. og 18. oktober 2007.
1) Intra-personal Variability and Agreement
Tid: Onsdag 17. oktober, 14.15-16.00.
Sted: Seminarrom 132, Harriet Holters hus.
I argue that our theoretical models of grammar should, in fact, be configured so as to capture intra-personal variation, and show that it is impressively systematic and controlled by grammatical factors. I develop the argument on the basis of agreement facts from Buckie, a dialect of Scottish English.
2) Possession in the Gaelic Clause and DP
Tid: Torsdag 18. oktober, 11.15-13.00
Sted: Seminarrom 2, P.A.Munchs hus. [ BL07 102 ]
This paper looks at the expression of possession in Gaelic. I show that clausal possession is not simply a locative predication within a small clause. Rather it involves an inherently case marked possessive subject with raising of a lower nominal into the higher subject position.
David Adger er kjent som en en svært kunnskapsrik lingvist og en glimrende foreleser. Han har skrevet mange bøker,
vi nevner her Core Syntax (OUP), samt flere titalls artikler i tidsskrifter som bl.a. Language, Linguistic Inquiry,
Lingua, Natural Language and Linguistic Theory . Han har først og fremst arbeidet med gælisk, kiowa og engelsk
A topic little tackled in formal grammar is individual level variability (intra-personal variation): where one person has a set
of options in their grammar. Apparent cases of this phenomenon are often dismissed as 'performance errors' or as register
or dialect variation. In this paper I argue that our theoretical models of grammar should, in fact, be configured so as to
capture this phenomenon, and show that intra-personal variation is impressively systematic and controlled by grammatical
I develop the argument on the basis of agreement facts from Buckie, a dialect of Scottish English. In previous work, I
developed an approach to intra-personal variablity in Buckie focussing on varation in past tense forms of the verb 'be'
(Adger 2006). In this paper, I extend the account to what has been termed in the literature 'singular concord' (Henry 1995).
In Buckie, plural noun phrases may have singular or plural agreement, but plural pronouns may not:
(1) The loonies kens us
The boys know us
(2) The loonies ken us
The boys know us
(3) They ken/*kens us
I show how the theory in Adger 2006 provides a more satisfying account of this phenomenon than previous generative treatments
(such as Henry 1995, Borjars and Chapmen 1998) or usage based accounts (Hudson 2000, Pietsch 2005). I also show how the
statistical variability in the data is leading to an ongoing change in frequencies of the variable forms across the generations, leading
to a future system where this phenomenon will probably be found only in auxiliaries (a change which seems to have gone to
completion in Devonshire English).
This paper looks at the expression of possession in Gaelic. The core examples are:
(1) Tha leabhar aig Daibhidh
Be book at David
` David has a book'
(2) Tha an leabhar aig Daibhidh fada
Be the book at David long
'David's book is long'
I show that clausal possession in (1) is not, contra Freeze, Kayne and Harley, simply a locative predication within a small clause:
(3) * Be [SC book [PP at David] ]
Rather it involves an inherently case marked possessive subject with raising of a lower nominal into the higher subject position
(see Adger and Ramchand 2005 for this approach to psychological predication in Gaelic):
(4) Be [TP book [PredP [PP David] Pred t_book ] ]
Evidence for this comes from scope facts, binding, pseudoclefting, wh-question formation and a host of other phenomena. I then
extend the analysis to DP internal possession of the sort seen in (2).
These data are important because they show that what is seen on the surface is not what is actually happening. The data that show
that (4) is correct and (3) is incorrect are subtle, and it is extremely unclear whether they are accessible to the language learner,
since apparently identical locative predications are indeed available in the language:
(5) Tha Daibhidh aig bòrd fhada
Be David at table long
`David is at a long table'
This gives us a rather clear case where a surface oriented construction-grammar type approach is implausible: rather learners must
be keyed to the semantic difference between possession and loctation and able to project these onto two different syntaxes.