topics: |  linguistics | cognitive | semantics | chomsky

Book Review: Bridging the Chomskyan divide

This is Jackendoff's attempt to bridge the chasm between the Chomskyan view
of language and the host of cognitive evidence that runs counter to key
assumptions such as nativity and compartmentalization of syntax.  This
mounting body of evidence led Chomsky to remark sarcastically in 2000:

   there is a ton of empirical evidence to support the opposite conclusion
   on every single point that I mentioned.  Furthermore, the core
   assumption of highly productive recent work - and its pretty impressive
   achievements- is that everything I said is wrong...
		-  Noam Chomsky, Architecture of Language, 2006, p.23

Here Jackendoff, who worked with Chomsky in the divisive 1960s but has now
diverged himself, tries to save those aspects of the generative grammar
enterprise that were valuable, while sacrificing philosophical
underpinnings such as syntactocentrism.  This task is formulated in terms
of three questions he spells out in a later precis on BBS, and also in the
later book Language, Consciousness, Culture(2007):
  - What was right about generative grammar in the 1960s, that it held out
  	such promise?
  - What was wrong about it, such that it didn’t fulfill its promise?
  - How can we fix it, so as to restore its value to the other cognitive
	sciences?

Two of the ways in which the Chomskyan model revolutionized linguistics are
elaborated in chapters 2 (Mentalism) and 3 (Combinatoriality). Mentalism is
described in "the remarkable first chapter of Noam Chomsky's Aspects of the
Theory of Syntax (1965)" (p.19) and brings in focus the user's
mind (instead of its usage as among people).  This MENTALIST position in some
senses gave rise to the cognitive view of language.  Jackendoff rightly
points out however, that in the Chomskyan view, this mentalist model was
implemented in terms of a distinction between competence (what the lg-user
knows), and performance (what she actually says) - and the focus was on
competence, which was to be judged in terms of the "ideal speaker-listener"
(Chomsky, 1965, Aspects of the Theory of Syntax, p.3-4).  In practice, this
meant that actual utterances of users were to be ignored for three decades in
order to explain some constructs made up by the linguist.  This was
eventually overthrown with the emergence of corpus linguistics in the late
90s.

Acquisition and Poverty of Stimulus

Chapter 4 deals with an overview of Universal Grammar and its challenges.
After an insightful overview of the evolution of generative grammar, that
is, the various theories introduced and defended (sometimes quite rudely)
by Chomsky.

Part of its emphasis was the focus on explaining language ACQUISITION.  But
here the model perhaps runs aground on the Poverty of Stimulus argument
(which said that there isn't enough data a child can learn grammar by,
hence grammar in some form is built-in).  But this is a problem only given
the autonomous syntax view, where the child is hearing string after string,
all completely devoid of any connection to real world scenes and events,
and then one is surprised that she manages to gain a competence in
constructing new sentences.  However, if we throw away the assumption of
syntactical dominance (as Jackendoff does, calling "syntactocentricism" an
"important mistake") it may turn out that semantics holds some of the cues
to syntax.

While taking a pro-semantics view in general, Jackendoff however, does not
propose that the learning (or natively having) some semantics may in some way
make it easier to propose a solution of the syntactic dilemma; in fact, he
sticks to the PoS argument, noting how children are able to disambiguate
among the many possible grammars and induce a structure that would explain
(generate) the correct sentences in her target language.  In contrast,
Jackendoff wryly notes, the legion of trained linguists aworking for decades,
with tools and knowledge far sophisticated than the infant, have not been
able to come up with the grammar of even one language.

The linguistics wars (mid-60s)

While discussing Universal grammar, Jackendoff also talks about the duels
between UG and the rest of the field.  The description of these duels is
quite interesting - at the time, Jackendoff was one of Chomsky's students,
and he puts in an apology of sorts by stating that he was party to the "rude
behaviour displayed by both sides" (p. 74).

The debate referred to included spasms of great vituperation by Chomsky,
who was by then a leading statesman in linguistics.  The same arrogance
informs his earlier attack on Skinner, and also his debate with Piaget.  For
example, in a 1973 repudiation tirade against Lakoff, Chomsky talks
down to Lakoff in the tone of a chiding schoolmarm - at one point, Lakoff's
statement is "completely wrong", at another his "confusion is complete" - and
finally,
	Lakoff seems to have virtually no comprehension of the work he is
	discussing.
This sort of arrogant language, surely could have been avoided by a senior
spokesman for the field.

Intellectually, Jackendoff traces how the division came about, primarily
because of the hope, after Chomsky's "Standard Theory" of Aspects, 1965,
that what he had named "deep structure" would become, in essence, a model of
semantics.  This view was carried forward by Fillmore ("case grammar"),
Lakoff and Postal ("Generative Semantics") and others who proposed
increasingly elaborate (and ultimately futile) sets of rules and derivations
for including semantics.

That Chomsky might have thought of Deep structure as having links to
semantics is revealed in J's extensive quotations from C, e.g. (footnote
p. 73):
    In place of the terms "deep structure" and "surface structure", one might
    use the corresponding Humboldtian notions "inner form" and "outer form"
    of a sentence... The terms "depth grammar" are familiar in modern
    philosophy and something roughly like the sende here intended
    (cf. Wittgenstein's distinction of 'Tiefengrammatik' and
    'Oberflaechengrammatik', 1953, p. 168) [Chomsky 1965, 198-9]

    The deep structure that expresses meaning is common to all lgs, so is
    claimed by the Port-Royal grammarians), being a simplre reflections of
    the forms of thought. [Chomsky 1966, p.35]

For more on Humboldt and "inner form", see Seuren's
Western Linguistics: An Historical Introduction, where "inner form"
in Humboldt is seen to include "at least both" the grammar or syntactical
rules and the semantics or logical form.

Chomsky abandons semantics

However, Chomsky repudiated the moves to interpret deep structure as
semantics.  In a move that could be viewed as an attempt to spite his
opponents, he diluted the "Deep" in "Deep Structure" b replacing it with
"D-structure", and in his next theory, Government and Binding Theory, he
moved Semantics (now called "Logical Form") to a much later stage in the
process.

 
   Chomsky's Syntactic Structures of 1965 revolutionized the study of
   language by holding out the promise that it might be possible to construct
   rule structures that would explain the working of language at both the
   surface (usually understood to be syntactic forms) as well as "deeper"
   levels (a topic of misunderstanding deep divisions in linguistics).  His
   Government and Binding theory (1981) however, was a letdown, and took the
   wind out of the semantic sail by claiming that "Deep Structure", now
   renamed "D-structure" was just a level of syntax.  By the time of the
   Minimalist Program of 1995, many linguists had tired of it all, and the
   theory did not cause much of a stir (it is rejected by Jackendoff here).
   In general, over time, there appears to be a steady reduction of the
   transformational components [rules] and an expansion of the Lexicon along
   lines proposed by then in several other grammars (e.g. HPSG).  The
   link to semantics remains nebulous and un-worked out at each step.
   (image from http://www.bbsonline.org/Preprints/Jackendoff-07252002/Referees/index_files/figure1.jpg)

In the process, there was widespread disillusionment among those who had
flocked to Generative Grammar hoping to get some insights into meaning (see
for instance the heartfelt discussion of this episode in Ch.1 of Robin
Lakoff's The language war).

[AM:
While the association of "apple" with the fruit is arbitrary, is the term
"Deep structure" also completely arbitrary (idiomatic) or is its semantics
derived by composing the semantics of Deep with Structure?  When smart
people coin names, they try to choose something that would give a sense of
its role.  If so, in what sense was "Deep semantics" deep, esp. given
Chomsky's other statements above?

Jackendoff defends Chomsky by saying that in the earlier traditions, they
thought of "DS" as meaning because, lacking formal logic, they had no idea of
"deeper" layers of syntax.  But the lay person encountering the term would
also tend to associate something like meaning to it.  "Chomsky chooses to
interpret this practice, for better or worse, as an implicit version of his
own theory of syntax." p. 73

But this does not hold water.  If "Deep Structure" is a purely technical
term, and has no relation to "deep" and "structure", then why call it so?
Why not call it, say, "stapler" or "term137"?

The same critique applies to other coinages by Chomsky, such as "the language
organ" - which means something to the average user - something that is a
separate part of the body, like the heart - yet it is a technical term
within the theory -  actually "the language faculty" (which is also
non-neutral) which means that it is also interacting with the
conceptual-intentional system (another technical term) and the sensory-motor
system.]

Jackendoff Proposal: Three autonomous generative grammars

As an alternative to the Syntactocentric view, Jackendoff proposes a parallel
structure with three branches - Phonology, Syntax and Semantics, where each
is an "autonomous" system [the meaning of this term is a bit unclear], served
by its own grammatical structure, along with interfaces between them.  This
combined system comprises the language structure, in Jackendoff's view.

These levels are well illustrated by Jackendoff's analysis of the sentence
"The little star's beside the big star":

 
   "The little star's beside the big star": phonological and syntactic
   levels.  In Jackendoff's model, each level is the result of an autonomous
   generative system.

 
	   The semantic level represent a set of predicates.  But is this all
	   the semantics the sentence has?  This is a q. thrown up by
	   cognitive semantics, that such logically formulated models of
	   semantics cannot address.  Finally, we have a "spatial model",
	   which is how the stars may be perceived.

Analog with vision:
     A tradition that Chomsky rarely cites is the gestalt psychology of the
     1920s and 1930s (e.g. Wertheimer 1923; Koehler 1927; Koffka 1935), which
     argued for an innate basis for perception and learning.  Nativist
     thinking remained influential in Europe much longer - e.g. in the work
     of Michotte 1954.  p. 70-fn

Cognitive Mechanisms: Priming

Chapter 7 discusses some psychological links with the proposed theory.
Among the phenomenon considered are lexical access in language perception
(how a word is understood based on linking the phonological computation of
guessed syllable/word boundaries etc (on the phonological blackboard), with
the lexicon-lookup and syntactic computation.

How does it distinguish surface-similar phonological structures (e.g. "a
parent" vs "apparent")?  These may be viewed as multiple bindings
[like polysemy] competing with each other -
- semantic promiscuity of lexical access - see expts by
[Swinney 1979] and [Tanenhaus etal 1979] -  survey in
[Cutler/Clifton:1999]

Describes WINNER-TAKE-ALL - but without using that phrase - "there comes a
point when all the competing multiple analyses but one are extinguished,
because of ill-formedness, contextual bias or other factors."

Sometimes resolves multiple possibilities prematurely, as in

(4) the horse raced past the barn fell.
[But does not explain why it closes prematurely.  No frequentist
interpretations here.]

Semantic promiscuity in lexical access

Hearing the word bug, in any context, briefly fires up other
interpretations of bug - speeds up reaction times to both insect and spy ...
But after this brief moment, only one meaning remains primed.

Psychological reaction time / Priming expts: Bug and insect are related at
     the semantic level; yet the judgment on the lexical decision task
     exhibits is through the phonology of insect.  Argues for [I am not
     sure I understand this] how this implies that priming is between long
     term memory (lexicon) and the phonology, and does not involve working
     memory (blackboard) at all.

Thus four components of lexical access:
  a) Activation: call from phonological working memory to item's
  	phonological structure in LT memory [/lexicon?]  Can also be
  	activated by priming - spreading activation from associated items in
  	lexicon.
  b) Binding: LT memory items is fetched to working memory.  Binding
  	entails activation, but not vice versa.
  c) Integration: item from blackboard (WM) is combined with larger
  	construction in progress.  Claims that it is "integrated
  	promiscuously" - i.e. all item-structure linkages are considered.
  	However, promiscuity is limited to those items in WM, and not to
  	other activated concepts (a,b).
  d) Resolution: the process by which multiple linkages are pared down.
  	because it is ill-formed, or it is out-competed [J: inhibition is
  	sent "up- or downstream to the associated structure at other levels.
  	Produces a "maximally favorable" single structure at all levels.

Each of these suprocesses is realized neurally.  Thus they are not
instantaneous, but have a time course -- a rise time, before max intensity
is achieved, and this max intensity depends on the item.  This is how
reaction time in lexical decision tasks depends on word frequency [Oldfield
and Wingfield 1965 ]

Spoonerisms

Rev Spooner: "Our queer old dean" for "Our dear old queen"
Q. How do these errors occur?
Possible answer - Lashley 1951: for the kw sound to be inserted, it must
	already be present in working memory, waiting its turn.  Hence, language is
	not produced by chaining words, but there is some overall
	planning is occurring in advance.

Victoria Fromley 1971: word transposition errors
	e.g. I'd hear that if I knew it --> I'd know it if I heard it
here the verb inflections incl the irregular past tense of know are in
place.  Thus, the exch has taken place at the syntactic level.

Semantic Gradation: Cluster concepts

Part III deals with Semantics and concepts.  The chapter on Lexical
Semantics (11) includes this interesting example of a graded meaning which
combines two different parts (features) of a semantics.

From p.353

concept = a set of conditions ; examples that satisfy fewer of the conditions
are generally regarded as less typical.

example from Fillmore 1982:

10 a. Bill climbed (up) the mountain
   b. Bill climbed down the mountain
   c. The snake climbed up the tree
   d.?*  The snake climbed down the tree

Climbing involves two independent conceptual conditions: i) moving upward,
and ii) effortful grasping motions (clambering).  Now, the most likely
interpretn of (10a) satisfies both.  10b vilates (i) and 10c (ii) since
snakes can't clamber.  Yet both examples are accepted instances of
"climbing".  However, both conditions are violated in 10d which is no longer
"climbing".   Thus, neither of the two actions are nec, but either is suff.
At the same time, 10a is judged to be more protypical.

Also, climb is not ambiguous between the senses 'rise' and 'clamber'; 10a
encodes both.

Note: Cluster concepts as Gricean scalar implicature
    [  [maxim of quantity: `Don't say more or less than is required'].
Gricean view of meaning (Scalar Implicature): meaning of a term is a
(possibly very large) set of meanings.  Is interpreted based on what
other meanings are available.  Thus, "or" may
mean exclusive-or (the book is in the bedroom or the bathroom) but not
always (e.g. `He is given to drinking, or is irresponsible, or both'); it's
semantics includes both meanings, and is further refined by pragmatics.
However, if someone says "p or q" it means that one is not in a position to
say an available alternative, "p and q" - hence that you are not in a
position to assert "and". Here the stronger sense "and", implies "or" (p^q -> p|q),
but not vice versa.
[Levinson on Pragmatics:]:
Other  pairs: < and, or>, , ,
, , . For each of these, asserting the
weaker, less informative expression will implicate that the stronger
expression does not apply, as in `Some of the students are punctual',
which, other things being equal, will implicate `not all of the students
are punctual'.

similarly, "climb" is in contrast to "rise", "ascend" or "go up" in that it
expresses a) voluntariness, and b) a difficulty in the task.  This
second meaning is part of the overall meaning, and the cluster concept
merely makes it clear that it does not mean "rise" alone.  Thus,
climbing down is possible only on trees and mountains, perhaps not on
staircases.

The problem with logical models of semantics is that it does not consider
these other entities in the symbol space.

More from the precis

You can read Jackendoff's own precis the
Brains and Behavioural Sciences journal.  Actually, this becomes the first
chapter of his later book, Language, Consciousness, Culture(2007).  Here
are some excerpts:

The remarkable first chapter of Noam Chomsky's Aspects of the Theory of
Syntax (1965) set the agenda for everything that has happened in generative
linguistics since.  Three theoretical pillars support the enterprise:
mentalism, combinatoriality, and acquisition.

Mentalism.  Before Aspects, the predominant view among linguists - if it
  was even discussed - was that language is something that exists either
  as an abstraction, or in texts, or in some sense "in the community"
  (the latter being the influential view of Saussure (1915), for example).
  Chomsky urged the view that the appropriate object of study is the
  linguistic system in the mind/brain of the individual speaker.  According
  to this stance, a community has a common language by virtue of all speakers
  in the community having essentially the same linguistic system in their
  minds/brains.

  There still are linguists, especially those edging off toward semiotics and
  hermeneutics, who reject the mentalist stance and assert that the only
  sensible way to study language is in terms of the communication between
  individuals (a random example is Dufva and Lähteenmäki 1996).  But on
  the whole, the mentalistic outlook of generative grammar has continued to
  be hugely influential throughout linguistics and cognitive neuroscience.

Combinatoriality: The earliest published work in generative grammar,
  Chomsky's Syntactic Structures (1957), began with the observation that a
  language contains an arbitrarily large number of sentences.  Therefore, in
  addition to the finite list of words, a characterization of a language must
  contain a set of rules (or a grammar) that collectively describes or
  "generates" the sentences of the language.  Syntactic Structures
  showed that the rules of natural language cannot be characterized in terms
  of a finite-state Markov process, nor in terms of a context-free phrase
  structure grammar.  Chomsky proposed that the appropriate form for the
  rules of a natural language is a context-free phrase structure grammar
  supplemented by transformational rules.  Not all subsequent traditions of
  generative grammar (e.g. Head-Driven Phrase Structure Grammar (Pollard and
  Sag 1994) and Lexical-Functional Grammar (Bresnan 1982, 2001)) have
  maintained the device of transformational rules; but they all contain
  machinery designed to overcome the shortcomings of context-free grammars
  pointed out in 1957.

  An important reason for the spectacular reception of early generative
  grammar was that it went beyond merely claiming that language needs rules:
  it offered rigorous formal techniques for characterizing the rules, based
  on approaches to the foundations of mathematics and computability developed
  earlier in the century.  The technology suddenly made it possible to say
  lots of interesting things about language and ask lots of interesting
  questions.  For the first time ever it was possible to provide detailed
  descriptions of the syntax of natural languages (not only English but
  German, French, Turkish, Mohawk, Hidatsa, and Japanese were studied early
  on).  In addition, generative phonology took off rapidly, adapting elements
  of Prague School phonology of the 1930s to the new techniques.  With
  Chomsky and Halle's 1968 Sound Pattern of English as its flagship,
  generative phonology quickly supplanted the phonological theory of the
  American structuralist tradition.

Acquisition: Mentalism and combinatoriality together lead to the crucial
  question: How do children get the f-rules into their heads?  Given that the
  f-rules are unconscious, parents and peers cannot verbalize them; and even
  if they could, children would not understand, since they don't know
  language yet.  The best the environment can do for a language learner is
  provide examples of the language in a context.  From there on it is up to
  the language learner to construct the principles on his or her own -
  unconsciously of course.

  Chomsky asked the prescient question: what does the child have to
  "(f-)know in advance" in order to accomplish this feat?  He phrased
  the problem in terms of the "poverty of the stimulus": many
  different generalizations are consistent with the data presented to the
  child, but the child somehow comes up with the "right" one, i.e. the
  one that puts him or her in tune with the generalizations of the language
  community.  I like to put the problem a bit more starkly: The whole
  community of linguists, working together for decades with all sorts of
  crosslinguistic and psycholinguistic data unavailable to children, has
  still been unable to come up with a complete characterization of the
  grammar of a single natural language.  Yet every child does it by the age
  of ten or so.

  One of the goals of linguistic theory, then, is to solve this "Paradox
  of Language Acquisition" by discovering what aspects of linguistic
  f-knowledge are not learned, but rather form the basis for the child's
  learning.  The standard term for the unlearned component is Universal
  Grammar or UG, a term that again perhaps carries too much unwanted baggage.
  In particular, UG should not be confused with universals of language: it is
  rather what shapes the acquisition of language.  I prefer to think of it as
  a toolkit for constructing language, out of which the child (or better, the
  child's brain) f-selects tools appropriate to the job at hand.  If the
  language in the environment happens to have a case system (like German), UG
  will help shape the child's acquisition of case; if it has a tone system
  (like Mandarin), UG will help shape the child's acquisition of tone.
  But if the language in the environment happens to be English, which lacks
  case and tone, these parts of UG will simply be silent.

  What then is the source of language universals?  Some of them will indeed
  be determined by UG, for instance the overall "architecture" of the
  grammatical system: the parts of the mental grammar and their relations (of
  which much more below).  Other universals, especially what are often called
  "statistical" or "implicational" universals, may be the result
  of biases imposed by UG.  For instance, UG may say that if a language has a
  case system, the simplest such systems are thus-and-so; these will be
  widespread systems crosslinguistically; they will be acquired earlier by
  children; and systems may tend to change toward them over historical time.
  Other universals may be a consequence of the functional properties of any
  relatively efficient communication system: for instance, the most
  frequently used signals tend to be short.  UG doesn't have to say
  anything about these universals at all; they will come about through the
  dynamics of language use in the community (a process which of course is not
  very well understood).

  If UG is not learned, how does the child acquire it?  The only alternative
  is through the structure of the brain, which is determined through a
  combination of genetic inheritance and the biological processes resulting
  from expression of the genes, the latter in turn determined by some
  combination of inherent structure and environmental input.  Here
  contemporary science is pretty much at an impasse.  We know little about
  how genes determine brain structure and nothing about how the details of
  brain structure determine anything about language structure, even aspects
  of language as simple as speech sounds.  Filling out this part of the
  picture is a long-term challenge for cognitive neuroscience.  It is
  premature to reject the hypothesis of Universal Grammar, as some have
  (e.g. Elman et al. 1996 and Deacon 1997), arguing that we don't know how
  genes could code for language acquisition.  After all, we don't know how
  genes code for birdsong or sexual behavior or sneezing either, but we
  don't deny that there is a genetic basis behind these.

More on Deep structure and Language War

        (from Jackendoff's precis, also Chapter 1 of Jackendoff 2007 - p.33)

A fourth major point of Aspects, and the one that seeped most deeply into the
awareness of the wider public, was Deep Structure.  Deep Structure expresses
underlying syntactic regularities of sentences.  For instance, a passive
sentence like (1a) has a Deep Structure in which the noun phrases are in the
order of the corresponding active (1b).

(1)   a. The bear was chased by the lion.
      b. The lion chased the bear.

Similarly, a question such as (2a) has a Deep Structure closely resembling
that of the corresponding declarative (2b).

(2)   a. Which martini did Harry drink?
      b. Harry drank that martini.

In the years preceding Aspects, the question arose of how syntactic structure
is connected to meaning.  Following a hypothesis first proposed by Katz and
Postal (1964), Aspects made the striking claim that the relevant level of
syntax for determining meaning is Deep Structure.

In its weakest version, this claim was only that regularities of meaning are
most directly encoded in Deep Structure, and this can be seen in (1) and (2).
However, the claim was sometimes taken to imply much more: that Deep
Structure IS meaning, an interpretation that Chomsky did not at first
discourage.3 And this was the part of generative linguistics that got
everyone really excited.  For if the techniques of transformational grammar
lead us to meaning, we can uncover the nature of human thought.  Moreover, if
Deep Structure is innate - being dictated by Universal Grammar - then
linguistic theory gives us unparalleled access to the essence of human
nature.  No wonder everyone wanted to learn linguistics.

What happened next was that a group of generative linguists, notably George
Lakoff, John Robert Ross, James McCawley, and Paul Postal, pushed very hard
on the idea that Deep Structure should directly encode meaning.  The outcome,
the theory of Generative Semantics (e.g. McCawley 1968, Postal 1970, Lakoff
1971), increased the "abstractness" and complexity of Deep Structure,
to the point that the example Floyd broke the glass was famously posited to
have eight underlying clauses, each corresponding to some feature of the
semantics.  All the people who admired Aspects for what it said about meaning
loved Generative Semantics, and it swept the country.

But Chomsky himself reacted negatively, and with the aid of his then-current
students (full disclosure: present author included), argued vigorously
against Generative Semantics.  When the dust of the ensuing"Linguistics Wars"
cleared around 1973 (Newmeyer 1980, Harris 1993, Huck and Goldsmith 1995),
Chomsky had won - but with a twist: he no longer claimed that Deep Structure
was the sole level that determines meaning (Chomsky 1972).  Then, having won
the battle, he turned his attention, not to meaning, but to relatively
technical constraints on movement transformations (e.g. Chomsky 1973, 1977).

The reaction in the larger community was shock:  for one thing, at the fact
that the linguists had behaved so badly; but more substantively, at the sense
that there had been a "bait and switch."  Chomsky had promised Meaning
with a capital M and then had withdrawn the offer.

The different levels of grammar in Chomsky

1965(Aspects;SPE 1968):  Standard Theory:
		 Deep Structure --> Semantic representation

1972(Studies on Semantics in Generative Grammar): Extended ST:
	     Deep Structure
	     + surface structure --> Semantic representation

(quantifier, scope, focus, and possibly anaphora were read off from surface
structure).

1975 (Reflections on Language): Revised Extended Standard Theory
	     Surface structure --> Semantic representation

("trace of movement", t, permitted all semantic interpretn in terms of Surface
structure)

1981 (Lectures on Govt and Binding): GOVERNMENT BINDING THEORY
        S-structure --> PF and LF (the latter through move-alpha "covert movement")
			;  LF --> Semantic representation

(Major shift: two new levels added; both derived from Surface structure (now
called S-structure) through further sequences of derivational rules (now
simplified to the general form "Move alpha"), guided by multiple constraints.
One of these sequences, resulting in the Phonetic Form (PF), more or less
duplicates the old phonological component.  The other sequence, which received
far greater attention, results in a level called Logical Form (LF).  LF begins
rather modestly (Chomsky 75) as a way of explicitly encoding quantifier scope
and anaphoric relations in syntax, but by Chomsky (1986:68) it is called a
"direct representation of ... meaning," an "'interface' between language and
other cognitive systems."  Thus it takes over the semantic function assigned
to the more extreme interpretations of Deep Structure in the Aspects theory.
The remnant of Deep structure, now called D-structure, is still the locus of
syntactic formation rules and the insertion of lexical items.  Crucially the
derivational rules that connect S-structure to LF are without visible
consequences; they perform "covert" movements that account for the mismatches
between interpretation and surface form.

1993 Minimalist Program
     		(covert movt) --> LF  --> Semantic representation

(Eliminates D- and S- structure altogether.  Syntactic structures are built up
by combining lexical items according to their intrinsic lexical constraints;
the operation of combining lexical items into phrases and of combining phrases
with each other is called Merge.  Merge operations can be freely interspesed
with derivational operations.  However, at some point, the derivation splits
into two directions, one ("Spell-out") yielding PF and the other LF.  Despite
all these changes, what is preserved is (a) that syntactic structure is the
sole source of generativity in the grammar, and (b) that lexical items enter a
derivation at the point where syntactic combination is taking place.
[FN: Along with the syntactocentrism of the framework has come a
syntactocentrism of outlook in many practitioners] 111

The scientific mistake: Syntactocentrism

We now turn to what I think was an important mistake at the core of
generative grammar, one that in retrospect lies behind much of the alienation
of linguistic theory from the cognitive sciences.  Chomsky did demonstrate
that language requires a generative system that makes possible an infinite
variety of sentences.  However, he explicitly assumed, without argument
(1965: 16, 17, 75, 198), that generativity is localized in the syntactic
component of the grammar - the construction of phrases from words - and
that phonology (the organization of speech sounds) and semantics (the
organization of meaning) are purely "interpretive", that is, that their
combinatorial properties are derived strictly from the combinatoriality of
syntax.

[In 1965] As for semantics, virtually nothing was known: the only things on
the table were the rudimentary proposals of Katz and Fodor (1963) and some
promising work by people such as Bierwisch (1967, 1969) and Weinreich (1966).
So the state of the theory offered no reason to question the assumption that
all combinatorial complexity arises from syntax.

Subsequent shifts in mainstream generative linguistics stressed major
differences in outlook.  But one thing that remained unchanged was the
assumption that syntax is the sole source of combinatoriality.

[Shifts in Chomsky - Standard Theory to G-B to Minimalist Program]

These shifts alter the components of syntax and their relation to sound and
meaning.  What remains constant throughout, though, is that (a) there is an
initial stage of derivation in which words or morphemes are combined into
syntactic structures; (b) these structures are then massaged by various
syntactic operations; and (c) certain syntactic structures are shipped off to
phonology/phonetics to be pronounced and other syntactic structures are
shipped off to "semantic interpretation" to be understood.  In short,
syntax is the source of all linguistic organization.

I believe that this assumption of "syntactocentrism" - which, I
repeat, was never explicitly grounded - was an important  mistake at the
heart of the field.4