Grammars as Mechanisms for Real-Time Tree-Growth: Explaining Clitic Pronouns
Ruth Kempson (King’s College London),
Stergios Chatzikyriakidis (Royal Holloway University of London)
In this paper, we use the puzzle of clitic pronoun placement and clustering in Modern Greek
to argue for a shift of perspective into defining grammars as mechanisms for incremental
growth of interpretation, as argued in Dynamic Syntax: Kempson et al. (2001), Cann et al.
(2005) Chatzikyriakidis and Kempson (2011). In this framework, the concept of structural
underspecification and growth of semantic representation is taken as the core syntactic
notion, grounded in the modal Logic of Finite Trees: Blackburn and Meyer-viol (1994). With
modal statements expressing an array of more or less fixed tree-node relations, restrictions
on tree development are definable in a way not available to grammars lacking such treegrowth dynamics. One of these is the restriction that no more than one underspecified
relation of a type can be built at a time, their being indistinguishable with only this weak
characterisation, which we use to explain a range of clitic clustering restrictions found in
many clitic-pronoun systems – we take the so-called Person-Case Constraint (PCC) as
displayed in different ways in dialects of Modern Greek as our case study.
The PCC is illustrated by the inability of a dative clitic pronoun to co-appear with a 1st/2nd
person pronoun, a puzzling morpho-syntactic restriction which is displayed in slightly
different forms across all clitic pronoun displaying languages. The examples from Standard
Modern and Cypriot Greek (SMG and CG respectively) are illustrative:
1) Tu
to
edose (SMG)/ Enna
tu
to
dosi. (CG)
him.CL-DAT it.CL-ACC gave.3rd
FUT
him.CL-DAT it.CL-ACC give
‘S/He gave it to him.’
2) Mu
to
edose (SMG)/ Enna mu
to
dosi. (CG)
me.CL-DAT it.CL-ACC gave.3rd
FUT me.CL-DAT it.CL-ACC give
‘S/He gave it to me.’
3) *Mu
se
edose (SMG)/ *Enna mu
se
dosi. (CG)
me.CL-DAT you.CL-ACC gave.3rd
FUT me.CL-DAT you.CL-ACC give
‘Intended: S/He gave you to me.’
4) *Tu
me
edose (SMG)/ *Enna tu
me
dosi.(CG)
him.CL-DAT me.CL-ACC gave.3rd
FUT him.CL-DAT me.CL-ACC give
‘Intended: S/He gave me to him.’
The dominant line of analysis of the PCC within syntax has been a feature
(in)compatibility, with the challenge being taken to be finding a feature, in some sense
relating to the clitics’ content, capable of capturing the appropriate restriction across the
requisite range of variants : Anagnostopoulou (2005), Adger and Harbour (2007) among
many others. The less ambitious morphological-template alternative is a simple lexiconinternal stipulation, with no attempt to capture the cross-linguistic patterning: see e.g.
Bonet (1995), Heap (2005) among others. Contrary to these mainstream views, we argue
following Cann and Kempson (2008) and Chatzikyriakidis and Kempson (2011), that
individual clitics and clitic clusters display calcified forms of general tree growth effects,
and that the PCC follows directly from the restriction debarring more than one unfixed
node of the same type at any one time in the growth sequence. This restriction as
displayed in SMG and CG follows from two assumptions: (i) dative clitics and 1st/2nd
person pronouns encode the same structurally underspecified relation - in many
languages, the forms in question are syncretic (though this matching of morphological
and semantic underspecification is not essential to the explanation) ; (ii) 3rd person clitics
project fixed structure (these forms are rarely syncretic) and the restriction accordingly
not operative, hence the wellformedness of (1) and (2). The different facts of Pontic
Greek (PG) strongly buttress this account, this being a dialect which basically disallows
exactly the patterns that according to current feature-based analyses should be allowed,
viz. combinations of two 3rd person clitics. In this dialect, these clitics are notably all
syncreticised, and with these taken to transparently encode the requisite weak tree
relation, the restriction follows:
(5) *Edoka a
a/ *Edoka aton
a/ *Edoka ats
a
gave it.CL it.CL gave him.CL it.CL gave them.CL it. CL
‘S/He gave it to it (e.g. to the child)/ S/He gave it to him/ S/He gave it to them.’
With this type of cross-dialectal cross-language variation conventionally taken to be
beyond principled explanation, we propose that concepts of underspecification plus
update should form the central core of natural-language syntax. A notable bonus of this
move is that natural-language grammars reflect the endemic relativity of natural
languages to context much more directly than has hitherto been envisaged.
Adger, D., Harbour, D., 2007. Syntax and syncretisms of the Person Case Constraint.
Syntax 10, 2-37.
Anagnostopoulou, E., 2005. Strong and weak person restrictions: a feature checking
analysis. In L., Heggie and F., Ordonez (eds), Clitics and Affixation, 199 - 235
Amsterdam Benjamins,.
Blackburn, P., Meyer - Viol, W., 1994. Linguistics, logic and finite trees.
Bulletin of Interest Group of Pure and Applied Logics 2, 2-39.
Bonet, E., 1995. The Person Case Constraint: a morphological approach.
In: H., Harley & C. Philips (eds), The Morphology Syntax Connection, 33-52.MIT Press.
Cann, R., Kempson, R., Marten, L., 2005. The Dynamics of Language. Elsevier
Chatzikyriakidis, S. and R. Kempson 2011. Standard Modern and Pontic Greek person
restrictions: a feature-free dynamic account. Journal of Greek Linguistics 11 127-66.
Heap, D. 2005. Constraining optimality: clitic sequences and feature geometry. In
L.Heggie and F.Ordonez (eds.) Clitic and Affix Combinations. John Benjamins.
Kempson, R., Meyer-viol, W., Gabbay, D., 2001. Dynamic Syntax: The Flow of
Language Understanding. Blackwell.
Kempson, R., Cann, R., 2008. Production pressures, syntactic change and the emergence
of clitic pronouns. In Cooper, R and Kempson, R (eds.) Language in Flux, 178-220.
College Publications.