specific language deficit and to contribute to linguistic theory through the

refinement of a descriptive instrument. However, Blumstein covertly, if not

explicitly, removes her study from this arena by occasional (although circum-
spect) references to 'psychological organization' (p. 88, footnote), 'current
generative psychological theory' (p. 108), or the 'psychological bond' between
phonemes (p. 109). Terms such as these refer to the language process, not to
taxonomy of the language product, and once invoked, displace the study from
the modest realm of description to a domain where a linguist's claims about the
implications of his data for linguistic theory must be very carefully scrutinized,
indeed.
In summary, the major strength of A Phonological Investigation of Aphasic
Speech is that it arranges and categorizes a large amount of information about
phonological dissolution in aphasia. Its major weakness is that it treats this
categorization (or description) as an explanation of the data, rather than as a
seedbed of significant hypotheses about the process of language production.

T h e University of Alberta MARY LOIS MARCKWORTH

REFERENCES
BOTHA, R.
1973 The Justification of Linguistic Hypotheses. The Hague: Mouton.
[Janua Linguarum, Series Major, 84.]
J A K O B S O N , R., 8c M. HALLE
1956 Fundamentals of Language. The Hague: Mouton. [Janua Linguarum,
Series Minor, 1.]
J A K O B S O N , R., M. H A L L E , & G. FANT
1962 Preliminaries to Speech Analysis. Cambridge, Mass.: M.I.T. Press.

The Psychology of Language: An Introduction to Psycholinguistics and Generative

Grammar, by J. A. Fodor, T. G. Bever, and M. F. Garrett. New York: McGraw-
Hill, 1974. Pp. xviii + 537.

THE FODOR, BEVER AND GARRETT (henceforth FBG) book is organized into
eight chapters, plus two short sections as preface and conclusion. With the
exception of the last chapter, on first-language learning, the first seven chapters
are fairly full and relatively self-contained treatments of the psycholinguistics
of taxonomic grammar (chapter 2), generative grammars (3), sematics (4),
the psychological reality of grammatical structures (5), sentence perception (6),
and sentence production (7). The first chapter, the Introduction, sets out the
tone of the book and provides a straightforward look at the authors' philosophi-
cal and methodological ties (indeed, the dedication does this as clearly and more
succinctly).
As a first approximation, FBG see the central question in psycholinguistics
as: 'How does the speaker-hearer employ the knowledge of his language
represented by a grammar to effect the encoding and decoding of speech?
(p. 21).' But the problem is also part of a larger and more important one, and
in fact is reminiscent of Chomsky's commitment of linguistics to cognitive
psychology. If the central problem in cognitive psychology is correctly identi-
fied as understanding how concepts are applied, then an examination of how
a speaker-hearer applies his knowledge of the abstract structure of his language

C J L / R C L 2 1 : 1 (1976)
Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 20 Nov 2018 at 05:41:52, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0008413100008124
 to the understanding and production of sentence-utterances should also serve
to throw light on a wider range of cognitive phenomena.
For FBG, then, the most exciting idea in contemporary psycholinguistics is
that it may be possible to construct an experimental mentalism, doing justice
to the richness of unobservable mental processes while also being disciplined
empirically in the sense that one expects a science to be. One major impetus to
the development of such a mentalistic psychology comes from linguistics,
specifically generative transformational linguistics in the form proposed and
elaborated by the Chomskyan paradigm. In fact the relation between the two
disciplines is the main theme of the book, and FBG provide 'a critical summary
of a decade of work on the mental processes involved in encoding and decoding
the syntactic structure of sentences' (xvii) from the point of view of this theo-
retical commitment.
The commitment to the sentence as the basic linguistic unit must, however,
revolve around more than just its centrality in the highly influential linguistic
model. The sentence must be demonstrated to be the linguistic unit in which
much of the systematicity of language and language behavior is revealed, and
much of the FBG discussion of grammar is devoted to reviewing evidence
related to this claim. While most would agree on the importance, if not the
centrality, of the sentence unit as highly revealing of the systematicity of
language structure, the same is not true for the structure of language behavior,
unless it is defined in the narrow sense that some adhere to. In fairness, one
might add that while this view does seem indeed to be a common one in much
of psycholinguistics, there are many who would reject the imperious suggestion
that it is the only one. In two short decades, linguists have seen their discipline
achieve a unified outlook under the aegis of a central issue, only to show signs
of fragmentation when it became obvious that certain crucial questions are
perforce overlooked by the monolithic nature of such singular paradigmatic
issues. One may speak more realistically of the central issue of contemporary
psycholinguistics, both paying tribute to the excellent work done earlier as well
as allowing for the possibility that the central issue may change, as it inevitably
will along with the fashions of science. After all, it is not inconceivable that the
present interdisciplinary partnership in theoretical interests may go the way of
the pseudo-partnership that evolved in the 1950's when structuralism in
linguistics found common ground with Hullean psychology and information
theory.
Some have already suggested that the one-way traffic from linguistics into
psychology may be slowing to a halt, if not reversing, while others have even
suggested that psychology and linguistics may be moving off in separate
directions entirely. This is not at all surprising, given that their goals and
methodology are so clearly different. This is not to deny that one derives
invaluable benefits from each stage. For example, in the case of Hullean
psycholinguistics, principles of learning theory were taken into a larger arena
and asked to deal with a highly complex, perhaps species-specific type of
behavior—language—and found to be seriously lacking. Hullean psycho-
linguistics was also characterized by close and fruitful interaction between
linguist and psychologist and this marks the beginning of a point of view in
which linguists make psychological claims for their grammars and psychologists
become concerned with the facts of language structure.
FBG present excellent discussions of taxonomic structural linguistics, how
it is done, what it discovered about language (observing that many of the

Reviews/Comptes rendus—127
Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 20 Nov 2018 at 05:41:52, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0008413100008124
 facts about language structure that taxonomic grammarians set out to capture
are real), and the psycholinguistics of taxonomic grammar. An equally in-
formative discussion of generative grammars follows, with instructive examples
of how five types of syntactic phenomena (affix movement, passivization,
relative clause construction, complementization, and lexical insertion) might
be handled with the mechanisms associated with the standard theory. One
thing we all have to be thankful for in this section is an absence of the cloyingly
cute sentences which have characterized so much of the literature in syntax
and semantics.
Developments in semantics within the context of generative grammar, are
surveyed, preceded by a review of traditional theories about the nature of
meaning in recent psychology and philosophy. Attention should be called to
the fact that psycholinguistics, like linguistics, has in its turn been heavily
influenced by questions which have traditionally been of interest in philosophy.
One can maintain, as Chomsky has in the case of psychology and linguistics,
that the rigid separation between the disciplines is an artificial one, but the
fact that philosophical concerns have had a substantial effect on linguistics
and psycholinguistics remains. One discipline influences another in several
ways—by virtue of the important questions posed, the way in which answers
to these questions are obtained, and the way in which the answers are stated—
and the means by which this influence is transmitted is by means of powerful
professional protagonists. The figures in linguistics are obvious; the authorship
of this book provides an example of philosophical ties in psycholinguistics, and
nowhere is the influence more clearly seen than in the chapter on semantics.
It is of some interest to note that FBG suggest that we are far from being
able to realize our goals for semantic theories of natural languages. One might
suggest that one reason this is so is because of the very nature of the goals as
specified by the semantic theory. The fact is that we are not only unclear about
the nature of the relationship between presupposition, focus, and entailment,
but also that we may not get any clearer about certain aspects of the problem.
One may well ask whether certain kinds of linguistic formulations about the
structure of a semantic theory are really at all informative about the nature of
meaning in some significant sense or are just expanded theories of collocations,
suggesting attractive explanations for some problems (e.g., the nature of
metaphor), but incomplete explanations (e.g., paraphrase) or none at all (e.g.,
inference) for others. Secondly, just how far are we from being able to under-
stand, let alone portray such facts of language use as presupposition, focus,
entailment, sentential truth value, and so forth. FBG's observation that our
proximity to realization of our goals for a semantic theory of natural languages
is not encouraging leads one to wonder what a blunt and realistic reading of this
euphemistic assessment should be. Perhaps it is the goals that need to be
restated, not the semantic theory.
An entire chapter is devoted to a review of experiments which investigate
the psychological reality of grammatical structures postulated by linguists. In
particular, the common interest of this entire tradition of experiments was to
discover whether syntactic structure, as postulated by linguistic descriptions,
in any way corresponds to the representations of or mechanisms that individuals
employ in producing, understanding, learning, or recalling sentences. Much
of this interest centered on the generative model with its concern for the
sentence as the basic linguistic unit, and it was generative grammar which
served as basic input for this line of experimentation. Curiously, many early

C J L / R C L 2 1 : 1 (1976)

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 20 Nov 2018 at 05:41:52, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0008413100008124
 experiments chose to investigate certain selected optional transformations in
the 1957 sense rather than obligatory transformations, often overlooking
crucial features like sentence length, frequency and familiarity, and semantic
considerations.
FBG's review of this line of experimentation re-confirms the psychological
reality of the taxonomy of sentences (the sentence family and intersentential
distances) implied by transformational syntax, but not necessarily the psycho-
logical reality of transformations. In fact, they observe that this is part of a
larger pattern of experimental findings on generative grammar; it turns out
that 'experiments which undertake to demonstrate the psychological reality
of the structural descriptions characteristically have better luck than those
which undertake to demonstrate the psychological reality of the operations
involved in grammatical derivations' (p. 241). Thus, structural descriptions
and many of the units specified therein (like the sentence constituent) do indeed
appear psychologically real, while the evidence for the fact that the integration
and recognition of sentences is governed by transformational processes is far
weaker in comparison. Similarly, the psychological reality of aspects of deep
structure and surface structure trees is open to interpretation in several re-
spects, but the evidence supporting these is still stronger than that supporting
the psychological reality of transformations.
The question of sentence perception raises the issue of whether we can expect
that the recognition and understanding of a sentence involves the speaker-
hearer's recovery of the structural description the grammar assigns to the
sentence. One expects that information about the grammatical structure of the
language which the sequence is a sentence of is essential information in the
recognition process. This is hardly an issue, but the question of how and in
what form this information is used is. According to FBG, the model is likely
one in which the relationship between the form in which grammars provide
linguistic structure and the form in which the recognition process does is not
a very intimate one. Specifically, the experimental results are equivocal in
terms of the psychological reality of grammatical rules, derivations, and opera-
tions in generative grammar, and are thus equally equivocal about the place
of such concepts in the process of sentence recognition. Indeed, investigation
of the derivational theory of complexity is an excellent example of this fact.
Briefly, the derivational theory of complexity hypothesized that the complexity
of a sentence can be assessed simply by the metric device of measuring the
number of grammatical rules appearing in its derivational history. A number
of experiments tested this notion, usually together with the related notion of
sentential similarity, and the results were not always supportive; in fact, in
several instances they were plainly contradictory.
Conversely, it appears that some features of the recognition process are not
predictable at all from the grammar. In fact, 'the present view is that the
sentence-recognition device has a complex structure of its own, and that
behavior in sentence-recognition tasks is not, in general, explicable as the
consequences of interactions between grammatical knowledge and unsyste-
matic variables' (p. 369). FBG go on to observe that 'the discovery that psycho-
linguistics has a subject matter—a body of phenomena which are systematic
but not explicable within the constructs manipulated by formal linguistics—
is, perhaps, the most important result of the last decade of psycholinguistic
research' (p. 369). One wonders whether this was not obvious from common
sense and whether it required a decade of serious psycholinguistic research—

Reviews/Comptes rendus—129
Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 20 Nov 2018 at 05:41:52, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0008413100008124
 after all, a model, strategy, or program which is adequate for one task is not
necessarily applicable to another task.
This of course suggests a rather different approach to the relationship between
linguistics and psychology and the goals, models, and methodologies thereof.
This has already been mentioned in the course of this review and one can only
speculate on the direction in which the relationship will move.
Just as aspects of the problem of sentence recognition are unsettled, so also
the important conceptual and empirical issues in sentence production are
unsolved, and much of what we have to say at this point is entirely speculative.
This may be tied to the fact that our knowledge of the nature of semantic
representations is also insufficient, so that we have no way of coding the under-
lying semantic intent. Suffice it to say that speakers do begin with some com-
municative intent and end with a final result which is a surface form in the
language in question. What the intervening levels of encoding are, the way in
which these are activated and operate, and the order in which they occur is
exactly what we must look to further research to elucidate. We do have con-
siderable evidence which may be interpreted in favor of surface constituent
structure as one such intervening level and we also have some idea of the
process by which such surface trees are constructed, proceeding in general from
left to right, top to bottom, and clause by clause. There would also appear to
be some evidence that this process is guided by a more abstract representation
yet, this notion being compatible with the notion of both the reality and effect
of deep structure. This underlying abstract representation also appears to be
syntactically organized (as suggested by slips of the tongue) and this organiza-
tion may be reflected in the integration and interpretation of certain aspects
of the surface organization.
What we obviously need more information on are the stimuli which may be
predictive of the form and content of sentences. There is only one systematic
experiment to date which has attempted to elicit sentence patterns by manipu-
lating various stimuli in the environment. It is discouraging that this important
aspect of sentence production has been the recipient of such niggardly interest,
but not surprising in view of the hermetic existence of the sentence in linguistic
theory. A related area, the effects of structural variables on the mental pro-
cesses involved in producing sentences, has received more attention, but no-
where near the amount that its immediate practical importance merits.
The last chapter, on first language learning, does not fit in as comfortably
with the remainder of the book and seems almost an afterthought. It is as if
the chapter had been added to make the book somewhat wider in its coverage
of the field, since developmental psycholinguistics has been of equal importance
for many. This last chapter is the one instance where one sharply recalls the
fact of an authoring troika; it is neither as well-integrated nor as complete
within itself as preceeding sections. Moreover, it makes too much of defending
man against chimp in its DesCarChomskyan recital of the uniqueness of human
language, human intelligence, and by extension, human nature.
The chapter does cover some interesting ground in taking up three topics:
taking up questions about the specificity of language skills (whether or not the
mechanisms of language learning are instances of universal principles of learn-
ing); surveying several accounts (namely, Chomsky and McNeill) as to what
the innate contribution to language learning might be; and lastly, taking up
some empirical studies of language acquisition. The selection and discussion

CJL/RCL 21:1 (1976)

Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 20 Nov 2018 at 05:41:52, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0008413100008124
 of such topics is of course congruent with FBG's position on psycholinguistics
and necessarily filtered through this viewpoint.
The book is not entirely without errors. The carets the reader is instructed to
look for in example sentence 5-11, p. 252, were left out. There are a number of
small errors: read 'affected by' for 'affect by,' p. 252; lack of apostrophe in
'speakers beliefs,' p. 402; read 'in' for 'is,' p. 420; read 'than' for 'that,' p. 446;
'rendered corresponding intelligible,' p. 462; lack of apostrophe in 'motorists
behavior,' p. 507. There are some similar errors in the first half of the book as
well; this reviewer did not trouble to itemize them until after it was apparent
that there were a small number.
This is in many ways an excellent book; it is also in many ways a contentious
book. Still, it offers a critical historical summary of what has characterized
modern psycholinguistics and the arguments that have been offered for and
against the theoretical trends of the last several decades. For this alone the
book is extremely valuable. Secondly, FBG present an excellent example of
what contemporary psycholinguistics has been and is all about for many, the
way in which it has selectively and successfully amalgamated linguistic and
psychological concerns, and the way in which these concerns have been trans-
lated into psycholinguistic research. Thirdly, it presents a critical summary of
what the results of these psycholinguistic investigations are and how these
illuminate the areas of speech production and speech perception. The fact that
the grammar is not a plausible model of the mental operations involved in
speech production and perception, leads us into the inevitable question of
what the future of psycholinguistic research will be. Certainly there are
obvious areas of investigation like the way in which the perception-production
system is internalized. But what really needs to be elucidated now is a theory
of the mental operations which are involved in speech production and percep-
tion, as well as a theory of how linguistic information is accessed in these pro-
cesses. It may be that directions may be reversed in this search, or at least
made reciprocal, with work on the psychological reality of syntactic structures
'concerned less and less with the vindication of independently motivated
linguistic analyses and more and more with choosing between candidate
analyses where the linguistic arguments are equivocal' (p. 512). Thus, 'it may
well be that only direct experimentation on psychological reality will ultimately
choose between competing syntactic theories' (p. 512). One would hasten to add
that there are areas in which the sizable amount of information that are the
result of psycholinguistic investigations should also provide principled ways of
treating certain facets of language from the very outset. The unfortunate fact
is that this is a book that will be read by many more psychologists than lin-
guists. It is the linguists who stand to gain much from a serious reading of
FBG's The Psychology of Language: An Introduction to Psycholinguistics and
Generative Grammar.

University of Victoria JOSEPH F. KESS

Reviews/Comptes rendus—131
Downloaded from https://www.cambridge.org/core. Stockholm University Library, on 20 Nov 2018 at 05:41:52, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0008413100008124