Differentiating core and peripheral syntactic data using analytic methods (FWAV)
Gertjan Postma (Meertens Institute Amsterdam)
[email protected]
Theoretical frameworks that describe natural language make a difference between phenomena that are
central to the grammar and phenomena that are peripheral, for instance the "exceptions" in traditional
grammar. Chomsky (1981) uses the "core grammar" versus "peripheral rules" opposition to validate
linguistic data. In articulated derivational models, e.g. Distributed Morphology (DM, Halle & Marantz
1991), a post-syntactic sequence of ordered modules is assumed, which helps us situating the
"exceptions" or the "periphery" within the theory itself: the later in the derivation, the more peripheral in
the Chomskyan sense (Arregi & Nevins 2012:342).
In some cases, rules from core syntax and rules from peripheral syntax give rise to similar Elanguage phenomena, e.g. head movement (core syntax) versus local dislocation (periphery). Various
scholars assume that this distinction provides a mechanism that drives language change. A peripheral
rule, introduced by adults or L2 acquirers, gets built into the core system by a subsequent generation
(Halle 1962, Weerman 1993, Lightfoot 1999, Yang 2002). In Postma (2011), an algebraic model is
proposed that simulates the relation between a peripheral phenomenon and a subsequent change in the
core grammar, within the analytic framework of Kroch's logistic function model (Kroch 1989).
Analytically, the S-curve, which represents the successful core-grammar change, was shown to be the
first integrated function of a bell-shaped curve that fuels it. (1) represents the rise of s-reflexives in 15th
century Dutch and the occurrence of the peer group sick-forms that initiated and fueled the change,
although these sick forms themselves died out, illustrating the two curves. Both curves are two-parameter
functions and the parameters of both curves are equal.
In this talk, we develop a geographic counterpart of this model. It gives us a tool to disentangle
peripheral and core-syntax data in data sets with geographic language variation. Under the assumption
that changes diffuse in space, we develop a model where a fuzzy isogloss of non-zero dimension is
represented as a spatial S-curve. The model predicts a spurious peak under this S-shaped isogloss. The
peak represents a peripheral phenomenon that can be considered as the driving force of the isogloss drift.
As an example, we study a local dislocation phenomenon in Dutch dialects, illustrated in (2),
where the standard V-Tense-AGR ordering is realized as V-AGR-Tense: AGR-metathesis. Using
syntactic argumentation, we show that core-syntax head-movement is tied to this peripheral metathesis
which drives the structural change. First, there is the observation that though AGR metathesis is scattered
in space, it precisely occurs under the isogloss that divides dialects with double paradigms (direct versus
inversion contexts, or AGRT and AGRC) from dialects with uniform AGRC. The situation is schematized
in Figure 3. Now, double paradigms are caused by absence of T-to-C (Zwart 1993). It will be shown that
AGR metathesis can also be produced by syntactic head movement, under T stranding when the verb
goes to C. This prepares a subsequent core-syntax change where T-to-C is not obligatory. This is the
successful change. The two types of spatial curves represent core and peripheral syntactic data.
References: Halle Arregi, Karlos & Andrew Nevins (2012). Morphotactics - Basque Auxiliaries and the
Structure of Spellout. Springer: Amsterdam. • Morris. 1962. ‘Phonology in generative grammar.’ Word
18: 54-72. • Halle, Morris, and Alec Marantz. 1993. Distributed Morphology and the pieces of inflection.
In: Kenneth Hale and Samuel Jay Keyser (eds) The View from Building 20: Essays in honor of Sylvain
Bromberger. Cambridge, MA: MIT Press. 111–176. • Lightfoot, David W. 1999. The development of
language: acquisition, change, and evolution. Oxford: Blackwell. • Postma, Gertjan (2011). The Impact
of Failed Changes - • Kroch, Anthony S. 1989. Reflexes of grammar in patterns of language change.
Language Variation and Change 1:199-244. • Weerman, Fred. 1993. ‘The Diachronic Consequences of
First and Second Language Acquisition: the Change from OV to VO’, Linguistics 31, 903-931. • Zwart,
Jan-Wouter. 1993. Dutch Syntax: A Minimalist Approach. PhD dissertation, University of Groningen.
Drenthe–hem/refl: predicted failure in Logist Model
% s-reflexives
100.0
fit to logist
% sick
80.0
predicted curve
rate
60.0
40.0
20.0
0.0
1400
1440
(1)
(2)
1480
1520
year
a
b
du klöp-z-de an
you knock.2sg.past PRT
du klöp-de-s an
you knock-past-2sg PRT
'you knocked on the door'
(AGR metathesis, dialect of Venlo)
(common pattern, general Limburgian)
Two dialect areas and spurious metathesis under
their isogloss
+ double
paradigms
– double
paradigms
V-to-T
V-to-T-to-C
± T-to-C
dialects with V-AGR-T
(3)
AGR.2sg intrusion in the V -T complex as an peripheral effect
on the du - jij isogloss
+ / - double paradigms
1.0
i s o g l o s s
0.8
rate
0.6
0.4
+ double paradigms
gij/jij/...
0.2
0.0
-30
(4)
(5)
AGR infixation
– double paradigms
klob-s-de
V-T-AGR
V-T-AGR
*V-AGR-T
-10
10
30
Distance from the Uerdinger isogloss in kms
[CPdu V-AGR [TP du [T-de] [VP V]]]
↑____________________|
du klöp-z-de ('you knocked')