Paul D. Deane, _Grammar in Mind and Brain: Explorations in
Cognitive Syntax_, Cognitive Linguistics Research 2, Berlin:
Mouton de Gruyter, 1992, Pp. x + 355.
Reviewed by Patrick Farrell
According to Marcel Danesi's reading of Giambattista Vico's
_La scienza nuova_ (Danesi 1993), metaphor is crucial to understanding
language and cognition. A fundamental principle of etymology is
that abstract word meanings develop from concrete meanings via
metaphor (see also Sweetser 1990: Ch. 2 and work cited there in this
connection), as revealed by the etymology of the word _idea_
(< Greek _idein_ 'to see'). Danesi suggests further that syntax,
the most abstract aspect of linguistic conceptualization, is the
culmination of this fundamental process of metaphorical projection.
Paul Deane's _Grammar in Mind and Brain_ embodies this general
view of language and cognition and provides support for this
specific idea about syntax, as it is an attempt to elaborate a model of
syntactic knowledge based on a version of the Spatialization of
Form Hypothesis (Lakoff 1987), i.e., the idea that syntactic
structures are metaphorical projections of image schemata from
ordinary spatial cognition. More particularly, drawing on the
insights into image-schematic cognition, categorization, and
conceptual metaphor contained in works by George Lakoff and
Mark Johnson (Lakoff and Johnson 1980, Lakoff 1987, Johnson 1987)
and John Anderson's theory of cognitive processes (Anderson
1983), this study
* develops a theory of hierarchical structure and grammatical
relations according to which these are projections of PART-
WHOLE, CORE PART, and LINK schemata that apply to physical
* explores some of the consequences of such a theory for various
important issues in English syntax, including the analysis of
complementizers and auxiliary/tense elements, complementation
structures, raising, control, and _tough_-movement constructions,
the distinction between restrictive and nonrestrictive relative
clauses, and _that_-trace effects;
* provides argumentation for a cognitive approach to island
constraints on extraction in question constructions and shows how
these constraints can be explained by the interaction of a theory of
attention management and the proposed theory of syntactic
* advances and motivates (with evidence from studies of aphasia)
the hypothesis that syntactic knowledge is situated in a neural net
in the left inferior parietal lobe of the brain, which accounts for the
proposed connection between bodily experience, spatial cognition,
This endeavor is noteworthy not only for its various analytical
innovations and successes in the domain of English grammar and
its overall ambitiousness (an integrated theory of cognition,
syntactic representation and processing, and neural function), but
also for its interesting stance with respect to the question of the
autonomy of syntax. D posits representations of constituent
structure like those of standard versions of X-bar theory as well as
relations between elements in these representations that are quite
analogous to the theta-marking, case-marking, and binding of
government-binding theory. More generally, unlike in Ronald
Langacker's cognitive grammar, syntactic knowledge is not
considered to be "indisocciable from other facets of human
cognition" (Langacker 1991: 1). Rather, it is claimed to be a
specialized form of knowledge, involving a distinct
representational system. However, unlike for Noam Chomsky and
other generative linguists, it is not held to be an innate module of
the mind that functions independently of other cognitive faculties.
The proposal is that the same part of the brain that is primarily
devoted to processing spatial structure is utilized in language
processing, entailing "an implicit analogy between grammatical and
spatial knowledge" (47). Thus, D takes a kind of intermediate stance
on the issue of syntactic autonomy. Although this book is bound to
appeal primarily to the cognitive linguistics community, I see it as
an admirable attempt to build bridges between research traditions.
It contains information, ideas, and analyses that researchers
interested in various aspects of the relation between language,
mind, and brain, including syntacticians of all theoretical
orientations, should find thought-provoking, at least.
D's central empirical goal is a comprehensive analysis of island
constraints. The various case studies of grammatical phenomena in
English are intended to illustrate and motivate the proposed
theories of syntactic structure and attention, which together
provide a thorough account of constraints on extraction such as that
illustrated by the following examples.
(1) a. [S' Who [S do you believe [S' that I know _ ]]?
b. * [S' Who [S do you believe [NP the fact that I know _ ]]?
(1b), unlike (1a), involves extraction of an interrogative pronoun
from within an NP, which in transformational grammar has been
assumed to be ruled out by such principles as the Complex NP
Constraint (Ross 1967) or the Subjacency Condition (for example,
Simplifying greatly, but hopefully without missing the essence,
D's story concerning _wh_ extraction and the contrast in (1) goes as
follows. First, a fronted _wh_ phrase is located in the COMP
position of the matrix S' and the COMP position and S (the other
immediate constituent of S') are mutually linked, which is to say
they stand in the same kind of relation as verbs and their
complements, for example. Second, by virtue of the fact that these
structures are metaphorical projections of the LINK and PART-WHOLE
schemata that pertain to physical objects, certain rules of inference apply,
most importantly (2).
(2) If X is linked to Y and Y has parts, then X is linked to some
part of Y.
Third, processing a structure with a fronted _wh_ phrase involves
connecting this phrase to some element in its sister S following
(possibly multiple) applications of (2). Fourth, successful processing
of such a structure is possible only to the extent that the domains in
the structure within which the links need to be established are
highly active or salient in terms of attention received. Fifth, salience
can be achieved in a syntactic structure by spreading activation (in
a manner approximating c-command, the heads of sisters of salient
nodes are also automatically salient), by being placed in focus, or by
being automatically focused by virtue of being highly entrenched
(where being lower on a hierarchy of topicality--as in Silverstein
1981--correlates with being more entrenched).
Now, in the case of (1b), _who_ is the element placed in focus (of
which there can only be one) and it needs to be linked to an element
contained in its sister S, following (2). It cannot be linked to the
embedded verb _know_ (which it needs to be linked to) because
_know_ is contained in an NP that is salient neither by spreading
activation (it is a maximal projection rather than a head) nor by
virtue of being automatically focused (being a definite, specific NP,
it is relatively high on the topicality hierarchy). (1a) differs crucially
in that the embedded clause containing _know_ is automatically
focused and hence salient, since clauses are low on the topicality
There are several interesting aspects of this story, which is
representative of D's stories for an impressive range of further
constraints on _wh_ extraction. It is noteworthy, to begin with, that
the general analysis of _wh_ extraction is quite like standard
formalist approaches to this phenomenon, except that the
connection between the extracted element and the extraction site is
established not through language-specific mechanisms such as trace
binding or SLASH feature percolation, but through a cognitive
process that occurs in the conceptualization of physical objects as
well. This approach also provides straightforward and plausible
accounts of various apparent exceptions to the standard island
constraints. For example, the acceptability of (3) is attributed (in
part) to the relative higher topicality (and hence higher salience) of
the specific, indefinite NP out of which extraction occurs.
(3) Which restaurants have you received [NP suggestions that we
patronize _ ]?
Finally, the key proposals relevant to the analysis of extraction
possibilities are independently motivated in compelling ways. The
topicality hierarchy and its proposed organization, for example, is
shown to be supported not only by numerous previous linguistic
studies implicating such a hierarchy but also by evidence from
reflexivization in English, the directionality of metaphoric transfer,
and trends in children's conceptual development. Similarly, the
proposals concerning attention and activation spread are supported
by evidence from metonymy and anaphora.
Overall, this book is a significant addition to what is now a fairly
large body of work arguing for cognitive explanations for
phenomena that have been alternatively characterized in strictly
Anderson, John. 1983. The Architecture of Cognition. Cambridge,
MA: Harvard University Press.
Chomsky, Noam. 1986. Barriers. Cambridge, MA: MIT Press.
Danesi, Marcel. 1993. Vico, Metaphor and the Origin of Language.
Bloomington: Indiana University Press.
Johnson, Mark. 1987. The Body in the Mind: The Bodily Basis of
Meaning, Imagination and Reason. Chicago: University of Chicago
Lakoff, George. 1987. Women, Fire, and Dangerous Things: What
Categories Reveal about the Mind. Chicago: University of Chicago
Lakoff, George, and Mark Johnson. 1980. Metaphors We Live By.
Chicago: University of Chicago Press.
Langacker, Ronald W. 1991. Foundations of Cognitive Grammar,
Volume II: Descriptive Application. Stanford: Stanford University
Ross, John R. 1967. Constraints on Variables in Syntax. MIT
Silverstein, Michael. 1981. Case Marking and the Nature of
Language. Australian Journal of Linguistics 1.227-244.
Sweetser, Eve E. 1990. From Etymology to Pragmatics:
Metaphorical and Cultural Aspects of Semantic Structure.
Cambridge: Cambridge University Press.
Patrick Farrell is an assistant professor in the Linguistics
Program at the University of California, Davis.
His research interests and publications are primarily in the areas
of syntactic theory and lexical conceptual semantics.