Introduction to Cognitive
Linguistics
LECTURE 4
Last time…
The fundamental arguments
The argument for mental grammar
Patterns + elements of the lexicon
Expressive variety (finite means – infinite
combinations)
Prescriptive grammar vs. mental grammar
Unconscious principles
Mental grammar
This capacity to combine words
into acceptable patterns and
create a limitless number of
novel sentences is usually called
mental grammar.
Watch it: it is not the ‘socially
prescribed’ grammar that
distinguishes one dialect from
another.
Watch it 2: it is largely governed
by unconscious processes.
The tip of the iceberg metaphor
Accessability of Mental Grammar and the
Unconscious Mind
Freud compared our mind to an iceberg, he claimed
that parts of the mind are not accessible and that the
mind could be visualized as an iceberg.
The unconsciousness of Mental Grammar is even
more radical than Freud’s notion: mental grammar
is not available to consciousness under any
conditions.
The difference between Freud and Mental
Grammar
Freud believes that unconscious beliefs can
be made conscious by way of psychoanalysis.
However, this is NOT possible with the rules
of language as they are simply not available
for introspection.
If you don’t believe this, try to devise a technique
through which a native speaker would eventually
become ‘aware’ of what’s going on in his or her mind
while producing a short sentence…
The Argument for Innate Knowledge
The way children learn to talk implies
that the human brain contains a
genetically determined specialization
for language
How do children do it? Many people
simply assume that the parents teach
them. Obviously, parents often teach
their kids words… and not much more
than that.
… Not even all words…
“This is ANY, Amy! Say, HOWEVER,
little girl!”
Even for those words that they do learn…
How are they able to do it?
Is it really easy to understand
that each chair of this world
belongs to the category/concept
of “CHAIR”?
I don’t think so. It is a
tremendously complex task!
And one in which there are
many errors, initially.
How about these..?
Or these..?
The Argument for Innate Knowledge
The idea that parents teach their children language
is so omnipresent, that the language is called
native or mother tongue.
But does it actually come from mothers?
No.
It comes from the environment.
Regardless of the child’s nationality, race, color, or creed, the
language it is immersed into in its early life will become its
native language.
This raises a serious question: if this is so, then there must be
something common to all of us, a human nature which allows
us to master just about any language we are exposed to early
on, regardless of our origin.
A linguistic democracy? Perhaps. The truth is: there are no
good or bad languages, and every one of us could have
acquired any of them as our mother tongue. Only if we’d been
given a chance…
Problems with the ‘parental guidance’ idea:
If parents teach children their
language,
1. We learn the rule of language:
e.g. cat -> cats,
wug -> wugs
break -> breaked
(overregularisation)
2. ‘Poverty of the Stimulus’ (Chomsky)
What a child is exposed to is usually
less than perfect language. To say
the least.
Three Main Approaches to Language
Acquisition
Empiricism: behaviorism (Watson, Skinner)
neobehaviorism, ‘connectionism’: (Rumelhart et al.
1986)
Nativism: ”Language Acquisition Device” (LAD),
”Universal Grammar” (UG) (Chomsky), Pinker (1994)
The Language Instinct
Constructivism: Piaget, Vygotsky, Bruner
Tomasello (2003) Constructing a Language, moving
by analogy from item-based phrases and word
islands to richer constructions
Empiricism:
Language is learned entirely from experience. The
mind of the child is a ‘tabula rasa’ (Locke). Learning
through general-purpose mechanisms:
habituation,
conditioning,
association-formation
Rationalism:
Language (and other characteristics of the human
mind) is innate (Descartes)
LAD, UG (Chomsky)
”Language of Thought” (Fodor)
Constructivism:
language is acquired through a developmental
process that leads to higher complexity, emerging
through interaction with the (social) environment,
influenced by but not determined by the genes
(Kant)
Epigenesis (Piaget)
Socio-cultural internalization (Vygotsky)
We’ll adopt the Rationalist Concept:
Why?
BECAUSE we appear to know more than we
actually seem to learn. It’s not clear what sorts of
stimuli could lead us to our adult language
knowledge.
“A striking property of language acquisition is that children attain
knowledge which, quite literally, infinitely surpasses their actual
experience.”
The Language Organ, Anderson and Lightfoot 2002:18
Children Do Learn Words, but…
As a result of parental instruction children do learn words,
but not relevant grammatical patterns as well.
David McNeill:
Child: Nobody don’t like me.
Mother: No, say ‘nobody likes me’
Child: Nobody don’t like me.
(Eight repetitions of this dialogue follow…)
Mother: No, listen carefully! Say “nobody likes me”
Oh! Nobody don’t likes me.
Kids Resist Instruction
Braime (1971)
Want other one spoon, daddy.
You mean, you want the other spoon.
Yes, I want other one spoon, please Daddy.
Can you say ‘the other spoon’?
Other…one…spoon
Say ‘other’
Other
‘Spoon’
Spoon
‘Other spoon’
Other…spoon. Now give me other one spoon?
In other words, children are impervious to correction and
conscious instruction.
Some Prescriptive Rules are Taught…
… and they hardly work.
Such as the rule that a preposition must never
end a sentence (or something you must never
end a sentence with!)
This rule has been mocked by no less a
persona than Winston Churchill:
“This is a rule up with which we should not
put.”
(Winston Churchill)
(In Serbian: would you ever say “Da ju je on
bio video na vreme, on nikada ne bi bio
zakasnio.”)
Some Simple Facts about Language
Only humans have a form of communication which
we call Language
There are approximately 6,000 languages in the
world.
Any normal child growing up (say, from prenatal to
infancy to 5 y.o.) in any language environment will
master the local language.
More Simple Facts
Many exceptional children, i.e. blind, deaf,
cognitively deficient, neurologically impaired, etc.
may exhibit essentially normal language
development
(The opposite is also possible: perfectly intelligent
children and adults may suffer from such conditions
that they end up with insurmountable language
problems)
Mastery of Language is achieved without
explicit instruction
Ever tried to teach an adult beginner some
English?
“Normal” adults have great
difficulty achieving competence,
let alone fluency, in a second
Language, despite:
Greater cognitive sophistication than
children
Explicit instruction in classrooms
This means that children have
something that adults lack! And
we will discuss this later on in our
course!
The “authorities” teach unnecessary things such as
the use of prepositions
However, there are some very complex rules which
are never taught:
A. Joan appeared to Moira to like herself.
B. Joan appeared to Moira to like her.
C. Joan appealed to Moira to like herself.
D. Joan appealed to Moira to like her.
You “Know” Unconsciously that
Each of these sentences has a different combination of
who is to like whom.
How do we get to know which interpretation is correct,
without any prior thinking?
It appears that we possess some special, unconscious,
knowledge about ordinary and reflexive pronouns, and
about the two verbs such as ‘appear’ and ‘appeal’.
Yet, no one would ever bother to teach native speakers
something like this in an English language course. So,
this particular piece of their knowledge of language does
not come from conscious instruction.
Jackendoff emphasizes that
No one is ever taught about contrasts between these
issues (pronouns and verbs such as ordinary vs.
reflexive, or ‘appear’ and ‘appeal’).
Regardless, this aspect of English grammar appears
to be deeply ingrained, much more so than the
already played out proscription about not placing
prepositions at the end of a sentence.
Some other examples may include what you might
have seen in a generative syntax course:
John is eager to please. vs
John is easy to please.
Another Striking Example of Deeply Ingrained
Knowledge
How many times do I have to tell you that the book
is not by Thomson, but by Martin-bloody-net!
The guy who screwed us was not the retailer but the
manu-fuckin-facturer, remember!
What is interesting in such examples is that we have
very clear intuitions how to use the infixes, without
ever being taught on the topic.
This phenomenon in English is known as expletive
infixation and is very interesting from a purely
linguistic perspective.
More Infixes
Uni-goddam-versity
Manu-fuckin-facturer
*Jacken-bloody-doff
*Ele-goddam-phant
Native speakers know (“feel”) that the first two are
acceptable, but the second two are not. Why? What
might be the underlying rule?
It is ABSOLUTELY clear that none of us was ever taught
the pattern where it is possible to insert an expletive infix
into English words.
… and yet, we know it…
The Principle is
That the infix sounds right only when it immediately
precedes the syllable of the word with the main
stress.
Thus we can conclude that much that we know about
the grammatical patterns of English has not been
taught.
Children simply have to figure out the patterns of
language. In other words they have to construct
their own mental grammar out of any input they get
(and this input is usually very impoverished,
remember. So their task is twice that astonishing).
The Argument for Innate Knowledge
The way children learn to talk implies that the
human brain contains a genetically determined
specialization for language
What we’ve seen so far suggests that a child is able to
figure out something that thousands of linguists
throughout the world have been trying for decades to
figure out…
This may well be called the Paradox of Language
Acquisition.
The Paradox of Language Acquisition
First: what the child ends up with is a mental
grammar that is inaccessible to consciousness.
Second: a substantial part of the language-learning
process is also unconscious, so linguists can neither
directly observe it nor ask children about it.
All children hear are sentences, they must
unconsciously discover for themselves the patterns.
Children Have a Head Start
Yes, they have a head start on linguists and
that’s why they can construct their mental
grammars.
They appear to have some sort of ‘hint’
about how to go about constructing the
grammar. That’s why it takes only about 5-8
years of work for an average kid, while
linguists haven’t been able to come up with
anything so far.
Head Start can be UG?
Children are equipped with a body of innate
knowledge pertaining to language.
Using that knowledge they can find patterns in the
stream of language they are exposed to, and they can
use these patterns as mental grammar.
Because this innate knowledge must be enough to
construct a mental grammar for any language of the
world, it is called UG (universal grammar).
So many grammars!
Prescriptive vs descriptive
Universal grammar: our biological, genetically-
transmitted capacity to pick up just any language.
Mental grammar: our perfect knowledge of the
language that has become our mother tongue (when
we are five years old or above).
Transformational-generative grammar: the
currently dominant linguistic theory trying to
formally describe the two human capacities above.
What Exactly is Universal Grammar?
What do children know (unconsciously) about
language in advance of language learning? (What is
UG?)
How do they use UG to construct a mental grammar?
How do they acquire Universal Grammar in the first
place?
How do they acquire Universal Grammar?
How can there be innate knowledge (knowledge that
is not learned)?
Like the teeth, or body hair, or walking, Universal
Grammar could just as well develop at some time
after birth (conditioned by a biological timetable;
grammatical patterns start getting acquired about
the age of 2).
So, it seems to be a part of our biological heritage. In
other words, it is genetic.
Innate Knowledge: the Mechanism
Determination of brain
structure by genetic
information
Determination of mental
functioning by brain structure
The brain is determined by the
DNA (its anatomical structure
and organization)
Chomsky: … we don’t learn to
have arms rather than wings…
The same goes for language…
Innate Knowledge: The Structure
The ability to learn language is rooted in our biology,
a genetic characteristic of the human species.
Just like an opposable thumb and a pelvis adapted
for upright stance.
This means drawing on biological precedents in
explaining language.
The Genetic Hypothesis
The mechanism for acquiring innate knowledge is
genetic transmission, through the medium of brain
structure.
Knowledge of language is determined by brain
structure, so it is present only when the supporting
brain structures are present (ca. the age of two).
Gene for Language: single gene on chromosome 7
region q31 structure named FOXP2?
Vargha-Khadem et al, 'Praxic and nonverbal cognitive
deficits in a large family with a genetically transmitted
speech and language disorder', Proc Nat Acad Sci USA 92,
930 – 933 (1995)
Vargha-Khadem et al,'Neural basis of an inherited speech
and language disorder', Proc Nat Acad Sci USA 95, 12695 –
12700 (1998)
Enard et al, 'Molecular evolution of FOXP2, a gene involved
in speech and language', Nature 418, 869 - 872, (2002)
