Introduction: from Turing to Chomsky
Noam Chomsky: Language & Mind
Noam Chomsky: Language &
Mind (1997: Into the Classroom Media)
- [Traditionally] the study of language has often been pursued in
fairly intimate relation to investigations of higher mental faculties
more generally, to what we can call problems
- The recent revival of many traditional ideas and concerns
followed a similar course. It was part of a more general
reconsideration of questions of mind that is sometimes know as the cognitive revolution of the 1950s.
- There was clearly an important shift ... from the study of
behavior and the products of behavior ... to the inner mechanisms that
enter into thoughts and action and interpretation. So, from the
cognitive perspective, behavior and its products are not the object of
inquiry, they are simply data. Of whatever interest they are,
their interest would be derivative. They would be of interest
insofar as they provide evidence about the inner mechanisms --
mechanisms of mind -- how they operate in executing actions or
interpreting experience, and the ways in which they come to assume the
forms they do from the embryo to maturity.
- The approach was criticized at the time as mentalistic. ... It's
mentalistic in that it's concerned with what we might call the "mental"
aspects of the world which stand along side of others, the mechanical,
and optical, chemical, and other aspects of the world. It
undertakes to study a real object in the natural world, primarily the
brain. ... And in this manner to move the study of mind
towards eventual integration into ... the biological sciences.
- In general the mentalistic approach is a step toward scientific
Properties of Language
- The human faculty of language seems to be a real species property, that is it varies
very little among humans and it is without significant analog elsewhere.
- Perhaps the closest analogs are found among insects, that would
be an evolutionary distance of roughly a billion years. For
example, the communication system of bees shares with humans an
unusal property ... sometimes called "displaced reference".
- [= reference to things not
present to the senses]
- The fact that say I can describe to you the arrangement of
furniture in my home, and you can understand what I'm saying, though
it's not in my sensory field or yours.
- Bees can ... describe to other bees ... the direction and
position and quality of a flower though the flower is not in the
sensory domain of any of them.
- That ... very elementary property of human language seems not to
be found elsewhere in nature, and it seems to be extremely hard to
induce experimentally in other animals. It's been tried, but
without much luck.
- [H]uman language is ... biologically isolated in its essential
properties ... and it also seems to be quite a recent development from
an evolutionary perspective.
- There is no serious reason today to challenge the Cartesian view
that the ability to use linguistic signs to express freely formed
thoughts marks, in Descartes words, "the true distinction between man
and animal, or machine." And that's true whether by "machine" we
mean the automata that captured the imagination of the 17th and 18th
century, or those that are providing a stimulus to thought about these
matters today. (Whether constructively or not, is another
question I will put aside.)
- [T]he faculty of language enters quite crucially into just about
every aspect of human life, human thought, and human interaction.
- a very elementary structural property that also appears to be
- the property that's present in it's purest form in the natural
numbers ... one, two, three, and so on, but no natural numbers in
don't learn that property. Unless the mind already possesses it,
at least possesses the basic principles, no amount of evidence could
provide them. So knowledge about that system comes to us from the
"original hand of nature," as David Hume put it, part of what we
would now call our biological endowment.
- Galilleo ... regarded "the discovery of means to communicate our
most secret thoughts to any other person with 24 little characters [the
alphabet]" as "the greatest of all human inventions." The
invention succeeds because of the discrete infinity of the language
that it reflects, that these characters are used to represent.
- [T]he authors of the Port
Royal Grammar and Logic were struck by what they called "the
marvelous invention of the means to construct from a few dozen sounds
an infinity of expressions that enable us to reveal to others with no
access to our minds what we think and imagine and feel.
- from a contemporary standpoint it is not an invention
- a product of biological evolution, about which, essentially,
nothing is understood.
The Language Organ
- [The language] faculty can reasonably be regarded as a kind of
language organ in the sense in which scientsts refer to the visual
system, or the immune system, or the circulatory systems, as "organs."
- Organ: a subsystem of a
much more complex structure, with distinctive characteristics.
And one identifies it in the hope of coming to approach an
understanding of the full complexity by investigating components of
this kind and their interaction.
- Each language is the result of the interplay of two factors:
- the initial state
- the course of experience
- The initial state of the language faculty can be thought of as a
kind of input/output device
- output = language obtained (that is the state of the language
faculty obtained ... the output is internal).
- The input and the output are open to examination. The course of
experience can be investigated to whatever detail one likes, and the
properties of the language that are acquired -- the state attained --
they can also be investigated. And what you learn in this way, by
investigating input and output, that can tell you a lot about the ...
the genetically determined initial state that mediates between them.
Universal Grammar (UG)
- a really uncontroversial fact, namely that the initial state is
common to the species. For example, if my children had grown up
in Tokyo, they would be talking whatever it is that kids talk in the
streets in Tokyo. ... So, it follows, then, that evidence
about Japanese bears directly on assumptions about the initial state
for English. And so on, much more generally.
- UG: "the theory of the initial state of the language faculty"
- Eventual problems for the biology of language
- How do the genes determine the initial state?
- What are the brain mechanisms that are involved?
- How do these states of the language faculty interact with other
systems of the mind and brain?
Domain Specific Cognitive Faculties
- This approach ... concerned with the faculty of language, with
the initial state, the states that it assumes. We can call it an internalist and modular approach ... meaning it's
concerned with the internal nature of a component of higher
mental faculties and it assumes that this array of higher mental
faculties is heterogeneous, that it's a complex of specialized
cognitive systems, much like the rest of the body, or for that matter
the rest of organic nature.
- There is supporting evidence ... for instance ... about dissociation of language and
other cogntive abilities. That means we find cases of language
abilities intact and other cognitive abilities not intact, and, in
fact, conversely. [Dissociation under pathology: Goes to
- There are also developmental
dissociations ... take, say my granddaugher and her kitten.
- Humans and cats ... will, apparently, treat [certain
navigational] problems quite similarly. Which means that some
domain specific cognitive system is developing in its own intrinsic
way, but more or less common across mammals, apparently, and probably
using similar mechanisms.
- We can also expect ... that my granddaughter's faculty of
language will begin to function ... in such a way as to identify the
category of language behavior, seperate it out, carry out segmentation,
find bigger units, do the kinds of analysis, and so on, that lead quite
early on to interpreting quite complex expressions novel to her
experience and later on producing them freely formed thoughts.
- And we can also be reasonably certain that won't happen with
her pet kitten who is subjected to essentially the same
- Maturation v. Learning
- Maturation: All of this just seems to happen. Rather like
growing tails, and fur, and so on. It isn't anything that the
child does. There's no way of teaching it. Nobody would know how
to teach it. It's just something that kind of happens to the
child, and to the kitten.
- It's questionable ... whether we should even use the term
"learning" to refer to these things. ... The conventional
properties that are associated with learning ... don't seem to
hold in these cases.
Language and Grammar
- We can refer to L [the
state of his language faculty] Peter's language. ...
Understood in this way a language is something like "a way to speak and
understand" ... It's a cognitive system that stores
information. It's embedded in other systems -- performance
systems -- that access and use that information.
- [W]hen I'm referring to language I mean internalized language in
that sense -- state of the language
- Grammar: the theory of
- Peters language ... determines an infinite array of
expressions. Each of those expressions has its sound and its
- In technical terms one says that Peter's language generates the
set of expressions.
- Therefore, a theory of Peter's language is called a generative
- ["Creativity" or "Productivity" of
Language: to be explained
- novelty: ability to
understand and generate new
- unboundedness: the
capacity to understand and generated an unlimited number of sentences
- Computational explantion of
"creativity" or "productivity": example
shows how infinitely many things can be said & recognized
shows how infinitely many things can be meant & understood.]
The Cognitive Revolution
- Generative grammar arose in the context of what is called the
- The first cognitive revolution of the 17 & 18 century ... was
part of the general scientific revolution that very radically modified
our understanding of the world. It was reconized at that time
that language involved the infinite
use of finite means.
- By the mid 20th century ... great advances in the formal sciences
had provided the appropriate concepts ... that made it possible to
address the traditional questions ... to try to give a precise account
of the computational principles that generate the expressions of the
language and in that manner to capture at least partially the idea of
infinite use of formal meaning. I stress "partial" for reasons I
will come back to.
- Aquistion studies revealed that from the earliest stages the
child knows vastly more than experience provides ... that is more than
experience provides without reliance on highly specific mechanisms to
get from scattered experience to what you actually know.
- UG: to show that all languages are essentially variations on the
- Conditions of adequacy for grammars: serious tension
- Descriptive adequacy: gives a full and accurate account of the
actual properties of the language. (Complexifying)
- Explanatory adequacy: shows how the particular language derives
from the initial state under the boundary conditions of experience.
- The assumption that a language is a complex system of rules ...
considerations of explanatory adequacy suggest that picture is very
unlikely to be correct.
- 1960-1980: attempt to discover general properties of rule systems
which we attribute to the faculty of language itself ... the initial
state ... in the hope that when we do that the residue that's left over
will be more simple and uniform. Crystalized in the Principles and Parameters approach.
Principles and Parameters
- According to this approach there aren't any rules and there
aren't any constructions.
- Rules are decomposed into ... invariant principles of the
language faculty which interact and yield the properties of
expressions. The constructions [N, V, NP, etc.] are held not to
exist. they're just taxonomic artifacts ... kind of like terrestrial mammal.
- Higginbotham's image: language is like ... a fixed network that
had a switchbox associated with it. The network is the universal
priciples. The switchbox is - a finite number of switches
that can be turned on and off -- those are the parameters. If you
set the switches one way you get Bantu. If you set them another
way you get Hungarian.
- Gives an outline of what could be a genuine theory of language
for the first time.
- modern generative grammar has sought to address certain concerns
that animated the tradition. In particular, the Cartesian idea
that the true distinction between humans and other creatures and
machines, is the ability to act in the manner that they took to be most
clearly illustrated in the ordinary use of language, though
characteristic of human action generally, that is without any finite
limits, influenced but not determined by internal states, appropriate
to situations but not caused by situations or internal states, hence
uncaused, coherent and evoking thoughts that the hearer might have
expressed, and so on.
- Generative grammar seeks to discover the mechanisms that are
used, therefore contributing to the study of how they're used in the
creative fashion of normal life. How they are used is the problem
that intrigued the Cartesians and those who followed them who talked
about the infinite use of finite means. Notice use.
- These branches of sciences [study of the motor and visual
systems] do not raise the question of how people decide to look at a
book, or pick it up [just as generative grammar doesn't raise the
question of how people decide what to say]
- It is these capacities -- manifested most strikingly in language
use -- that are at the heart of traditional concerns. For
Descartes, they are "the noblest thing we can have and all that truly
belongs to us." They're the defining characteristics of human
beings and the essence of mind, from his point of view.
- About a half a century before Descartes, the Spanish
Philosopher/Physician Juan Huarte observed that this "generative
faculty of ordinary human understanding and action" -- his term for it
-- "the generative faculty is foreign to beasts and plants, although
it's a lower form of understanding that falls short of true exercise of
the creative imagination. And it's important to stress that even
this lower form falls far beyond our theoretical reach apart from the
study of the mechanisms that enter into it.
- [M]any mysteries still lie beyond the form of human inquiry we
call "science." It's a conclusion which we shouldn't find
surprising, in my opinion, if we consider humans to be part of the
organic world, and perhaps one we should not find distressing either.
Appendix: Phrase Structure Grammar
- S -> NP + VP
- S -> S + CONJ + S
- NP -> PN
- NP -> CN
- VP -> VI
- VP -> VT + NP
- PN -> spot
- VI -> barks
- VT -> eats
- CN -> dogfood
- CONJ -> and
- Example sentences
- Spot barks.
- Spot eats dogfood.
- Dogfood eats Spot.
- Dogfood barks.
- Spot barks and spot eats dogfood. [1 and 2]<>
- <>Spot barks and spot eats dogfood and Spot barks and
dogfood. [5 and 5]