1. Introduction
Thought has
internal structure independently of language. Humans, that is, have the
cognitive ability to construct frames (Fillmore 1982, 1985) or, in Lakoff’s
(1987) terminology, Idealized cognitive Models (ICMs), i.e., nonlinguistic
schematic conceptual representations of, roughly, objects, and events. – 301
2. Verbal and
constructional polysemy: different cognitive approaches
Though Cognitive
Linguistics cannot be said to form a single unified framework (yet), it seems
united with respect to two points: firstly, language (words and grammar)
structures nonlinguistic conceptualizations that are grounded in experience,
and secondly, language underdetermines meaning. – 302-03
3. The
caused-motion construction
3.1 The
constructional senses of the caused-motion construction described
3.2 The
caused-motion constructional subsenses: some problems
4. Arguments
against additional CAUSE and MOVE verb senses
4.1 Fusing CAUSE
to MOVE and basic verb meaning
4.2 The effect of
linguistic presence of MOVE
4.3 Locative
versus directional PPs in caused-motion constructions
5. A
compositional account
5.1 Platonic
concepts and senses in contact
Verbs only
symbolize one abstract Platonic concept (whose content is, by definition,
unpredictable) and that actualized sense is never a sense of the verb in and of
itself but, again by definition, a sense in context that includes whatever
other (non)linguistic conceptual elements helped create it; a sense, in other
words, that gets compositionally constructed by cognition, and is predictable,
given sufficiently tight contextual constraints (contextual in its broadest
sense). – 319
5.2 Oblique
complements s subpredicates rather then arguments
5.3
Compositionality: a narrow view and a broad view
“[a] language is
compositional if the meaning of a complex expression is systematically related
to the meaning of its constituents” (Coulson 1997: 3). – 332
6. Predictable or
not, that is the question
The view upheld
by the majority of the cognitive linguistic community is (i) that words are
polysemous, with a central (essentially prototypical) sense and other extending
from the central sense or from another extended sense, and (ii) that extended
senses are not predictable, only motivated. – 322
7. Concepts are
blind without percepts, percepts are vague without concepts