Advances in Biolinguistics - The Human Language Faculty and Its Biological Basis

framework? Or, if we may eliminate (2), then should we also reconsider the
grounds of (3) and (4), the effects of which have been predominantly attributed
to (2)? (See also Fukui 2011 and Fukui and Narita 2014.)
The purpose of this chapter is to pursue the simpler theory of Merge in (5),
and argue that not only endocentric structures but also “exocentric” (label-free
and non-endocentric) structures are possible in human language. In Section 2,
we will fi rst introduce Fukui’s (2011) generalization concerning the distribution
between endocentric (or “asymmetric”) and exocentric (“symmetric”) structures
and explore some of its consequences. In Section 3, we will put forward a
theory of “symmetry-driven” syntax that can naturally account for Fukui’s (2011)
generalization, which we will also refi ne. We will further discuss various conse-
quences of the proposal in Section 4 and Section 5. Section 6 concludes the
chapter.

2 Fukui’s (2011) generalization

In what follows, we will explore the distribution of (non-)endocentricity in
human language without recourse to the stipulations in (2)-(4). Let us start
with cases whose syntactic and/or interpretive properties exhibit typical endo-
centricity. (7a-e) exemplify cases of the form {H, XP}, where H, an LI, typically
serves as the head.^1

(7)
C TP

b.

T vP

c.

v VP

d.

a.

V nP

e.

P nP

These are traditionally called “head-complement confi gurations.” The struc-
turally prominent LI H functions as a prime factor for characterizing the
semantic interpretation of these structures at SEM: C determines the force
of the clause (declarative, interrogative, etc.) in (7a); T feeds tense and modal
properties in (7b); V characterizes lexical and aspectual properties of the
denoted event and assigns a Theme/Patient θ-role to its complement, etc.
Generally speaking, {H, XP} structures receive what has been traditionally
called “lexical-conceptual aspects of semantics” or “d-structure interpreta-
tions,” including selectional restrictions, predicate-argument structure,
θ-marking, etc., in which the notion of endocentricity/headedness plays a
major role.
Interestingly, such endocentric {H, XP} structures are typically generated by
External Merge (EM). As pointed out by Chomsky (2004), Merge has two
modes of application: EM takes two SOs that are independent of (external to)
each other, as in (7a-e), while Internal Merge (IM) takes two SOs that are not
independent, i.e., one is a term of the other.

Feature-equilibria in syntax 11