ED0013 – Second Language Teaching and Learning
Module 1 Essay 1: Second Language Learning
"Pattern Detection Theory of Language Acquisition: Exploring
Alternatives to Chomsky”
By Gareth Williams
Student Number: 139022140
Date of submission: January 31st, 2014
MA Applied Linguistics and TESOL
1
Contents
1. Introduction
page 3
2. Description of Chomsky’s LAD and UG
page 3
2.1 Effectiveness of Chomsky’s Theories
page 4
3. Critique of Chomsky
page 6
3.1 Objections to Chomsky 1: Not Falsifiable
page 6
3.2 Objections to Chomsky 2: Not Locatable
page 6
3.3 Objections to Chomsky 3: The “Module” View of the
Brain page 7
3
3.4 Objections to Chomsky 4: Brain Plasticity
page 8
4. A Pattern Detection Theory of Language Acquisition
(PDTLA) page 9
4.1 Implications for LAD
page 11
4.2 Implications for UG
page 12
4.3 Limitations of PDTLA
page 12
5. Conclusion
page 14
Bibliography
page 15
4
1. Introduction
Chomsky’s theories on the nature of language and how it is
acquired are arguably the most influential of the last 50
years. Any alternative theory of language acquisition must
address his ideas at some point.
This essay is divided into two halves. The first half of the
essay is organized into three sections. The first section
will be dedicated to a description of Chomsky’s theories of
the “Language Acquisition Device” (LAD) and “Universal
Grammar” (UG). The second section will outline the
effectiveness of Chomsky’s theories. In the third section
four objections to his theories will be raised: that the LAD
and UG are neither falsifiable nor locatable, that Chomsky’s
theories adhere to a modular theory of the brain and that
they fail to explain the phenomenon of brain plasticity.
The second half of the essay will attempt to point towards
an alternative theory of language acquisition (the Pattern
Detection Theory of Language Acquisition) that explains the
phenomenon of brain plasticity. This section of the essay
5
will be dedicated to an attempt to indicate the implications
that arise for Chomsky’s theories of LAD and UG. Finally,
the essay will conclude with an outline of the limitations
of the PDTLA.
2. Description of Chomsky’s LAD and UG
Chomsky is concerned with a description of a person’s
knowledge of a first language (their competence) and not
with the use of their first language (their performance)
(Larsen-Freeman, 2011). He believes that the human brain has
a specially designated part that provides the foundation for
language use. Chomsky claims that the potential for learning
language is a unique biological aspect of the human brain
(Chomsky, 2005). He refers to this biological endowment as
the LAD (Chomsky, 2002) and describes the LAD as a
specialized device whose function is distinct from other
kinds of learning (Chomsky, 2002).
Chomsky also describes every child as being the heir of a
physical system of ‘Universal Grammar’ (UG) that is
activated by exposure to the language around them (Chomsky,
6
2002). As an illustration, it is useful to think of UG as
being a computer’s motherboard with circuitry laid across
it. When a child is born, that motherboard is largely
dormant. As the child is exposed to the input of her mother
tongue, parts of the motherboard light up. Chomsky believes
that all human beings have the same motherboard but that
exposure to different languages will start up different
parts of the system.
There are two main points here. First, humans are
biologically endowed with a specialized LAD whose function
is distinct from other kinds of learning. Second that all
languages’ grammars are a subset of UG – a sense of which is
hardwired into every human.
2.1 Effectiveness of Chomsky’s Theor ies
Chomsky’s theory is important because it identified and
dealt with the limitations of Behaviorism. Behaviorism was
the idea that people are conditioned to learn language in a
stimulus and response fashion (Larsen-Freeman, 2011). That
is, a child will attempt some sort of communication and, on
7
the basis of positive or negative reinforcement, learn their
first language. Thus, children learn by being trained to
perform communicative acts.
Chomsky points to the massive number of utterances at the
disposal of a human language user. Humans have the capacity
to create unique and limitless expressions of their
thoughts, appropriate to their circumstances and
consistently understandable by other people (Chomsky, 2002).
Chomsky goes on to state that the breadth of input which
children receive is not sufficient to explain the breadth of
communication that children end up exhibiting. The input
that they are exposed to is limited, often badly formed,
full of false starts and frequently arrives to the listener
in fragments (Larsen-Freeman, 2011).
Chomsky also points to the 100% success rate among children
for learning their first language, no matter what their IQ,
and at a time when they have a minimal capacity to hold
abstract concepts in their minds. This capacity for language
seems uninfluenced by adverse environmental conditions
8
(Chomsky, 2002). He maintains that only the existence of a
specialized LAD could explain this phenomenon and the
conclusion that such a specialized device exists is
difficult to avoid (Chomsky, 2002).
Regardless of whether or not Chomsky’s views are adhered to,
it is clear that children are developing some sort of system
of rules, inferences and hypotheses to leverage their
limited input into the ability to produce a vast array of
utterances. Chomsky believes that every child is the
biological heir of such a system and that this system is a
unique biological aspect of the human brain, which operates
as one of many separate components of a child’s brain
(Chomsky, 2002).
3. Critique of Chomsky
3.1 Objections to Chomsky 1: Not Falsifiable
Chomsky’s theories are not universally accepted. The first
objection this essay will address is the issue that the
existence of UG is not falsifiable. A scientific theory
requires that there are certain conditions under which it
9
might hypothetically be falsified (Popper, 2005) The lack of
any such conditions is a sign that a given theory lacks
scientific rigor (Popper, 2005). Just such a criticism can
be leveled at the idea of Universal Grammar. If someone
found empirical evidence of a language somewhere in the
world that was different to all the other languages, this
would not be sufficient grounds to falsify Chomsky’s theory.
An adherent of Chomsky could simply point out that the new
discovery was an aspect of the Universal Grammar that hadn’t
up until then been identified.
3.2 Objections to Chomsky 2: Not Locatable
Secondly, it is strange to posit the existence of a device
that we have no direct knowledge of. No one can point out
exactly where the LAD is. Chomsky says that the existence of
young human language learners is sufficient evidence of the
LAD. According to Chomsky there is no controversy in the
existence of the LAD because humans learn language where
other animals fail to do so (Chomsky, 2002). He is also
implying that the only plausible explanation for language
10
acquisition is the existence of a specialized LAD. However,
the idea that humans acquire language with some sort of
device is self-evident. This assertion does not prove that
the LAD is a specialized component of the brain. Stating
that there is a specialized LAD requires that we describe more
of the nature of that organ – including its location.
3.3 Objections to Chomsky 3: The “Module” View of the Brain
During the thirty or so years from the 1950s to the 1980s,
when Chomsky was developing his theory of language, the
common perception of the brain was that it was a static
entity divided into different sections with rigid assigned
roles for each part (Plaza, Gatignol, Leroy, & Duffau,
2009). Language operations were considered to be the
function of specialized parts of the brain (Plaza et al.,
2009). Chomsky appears to still assume the view that there
is a genetically determined component of the brain dedicated
to the structure and use of language (Chomsky, 2002). This
“module” view of the brain has since been shown to
inadequately describe language operations. Those areas that
11
had been thought to be exclusively specialized for language
have been shown to be involved with other thought processes
and speech functions have been found to be not exclusively
derived from the Wernicke-Broca’s language network (Plaza et
al., 2009).
This is not, in itself, a problem for positing a LAD. It
would seem possible for the LAD to exist – but for it to be
more complicated than was once believed and for it to exist
as multiple parts of the left hemisphere of the brain,
coordinated together (Carstairs-McCarthy, 2000).
12
3.4 Objections to Chomsky 4: Brain Plasticity
A more serious objection arises from recent developments in
the understanding of brain plasticity. Chomsky has asserted
that language acquisition is a special type of learning
distinct from other kinds of learning (Chomsky, 2002). Brain
plasticity implies that this is not the case.
Historically, one of the most important parts of the brain
associated with language processing was known as Broca’s
anterior frontal area (Plaza et al., 2009). This area was
held to be solely responsible for language production. A
least one recent study has found, however, that due to brain
plasticity, it is possible for the human brain to learn to
adapt most of its language processing so that Broca’s area
is not needed (Plaza et al., 2009).
These insights show that different parts of the brain can be
re-wired to take on different functions. If language can be
learned and used by a different part of the brain, than
would under normal developmental conditions function as the
language centre, the suggestion is that there is no
13
specialized LAD inside the brain. It suggests instead that the
potential for language acquisition is present throughout the
brain.
The fact that humans learn language with a brain (a ‘device’
in Chomsky’s lexicon) is not in dispute. The question is
whether the acquisition of language is a different kind of
learning than learning other cognitive skills. If there were
such a thing as a specialized LAD, whose function is
distinct from other kinds of learning, then other parts of
the brain should not be able to acquire the LAD’s functions.
However, brain plasticity demonstrates that other parts of
the brain can acquire the LAD’s functions. This indicates
that language acquisition does not have a special status
when compared to other kinds of learning.
Chomsky’s theory could survive if it turned out that the
brain adaptation was also part of the LAD – just a
heretofore-dormant part. However, it could then be argued
that he would not be describing a specialized LAD so much as
he would be describing a general function of the brain.
14
Further examination of the implications of brain plasticity
on language acquisition would be of benefit towards a clear
explanation of this issue.
4. A Pattern Detection Theory of Language Acquisition
An alternative theory of language must have equal or greater
explanatory power than the combination of Chomsky’s LAD and
UG if it is to be considered a more accurate model of
language acquisition. The idea of a LAD, as a specialized
part of the brain, comes into question when we note that the
brain exhibits plasticity. Different parts of the brain have
the potential to take up alternative processing roles as
required. When an individual, for example, takes an injury
to an area of the brain traditionally associated with
language processing, it has been found that it is possible
for other parts of the brain to take over those language
functions (Plaza et al., 2009).
One explanation for why this might be so is that the entire
human brain’s neocortex is essentially a pattern recognition
machine (Kurzweil, 2013). If multiple parts of the brain are
15
performing the same function – pattern detection – this
would explain why different parts of the brain can, under
certain conditions, take on alternative roles, if injury or
other circumstances necessitate.
Detecting patterns and making inferences based on them could
be argued to be an essential element of intelligence:
“Intelligence predicts, that is its essence; the same
intelligence that allows us to plan, hope, imagine, and
hypothesize also allows us to worry and anticipate negative
outcomes” (Doidge, 2007, p. 164). Furthermore, it might be
argued that failing to infer a pattern is a failure of
intelligence – and comes with disastrous evolutionary
consequences.
Imagine early humans who are exposed to a pattern that a
certain colored berry is fatally poisonous. The humans who
perceive this pattern survive. Even if they are wrong about
the berry being poisonous, they will only miss out on a
little fiber as a result of their overly cautious approach.
The humans that fail to identify the pattern, however, are
16
more likely to die and fail to pass on their genes (Shermer,
2008).
This example illustrates the point that failing to detect
existing patterns can be far more dangerous than detecting
patterns that don’t exist. Inferring incorrectly from past
instances will, more often than not, be a harmless mistake
resulting from a generally cautious outlook. The
evolutionary cost of such a mistake is minimal compared to
the evolutionary benefit of successfully detecting real
patterns (Shermer, 2008). This is a plausible evolutionary
explanation for why humans exhibit marked pattern detection
abilities.
Humans have a hyper-tendency to see patterns everywhere. We
have passed this tendency down through millions of years of
evolution. This proffers an explanation of what we do, as
children, with the language we hear. We identify the
patterns, infer from them, and then get on with producing
unique utterances of our own.
17
Raymond Kurzweil (2013) goes further than saying that we
have a tendency to assign patterns to our environment. He
believes that all human thinking and intelligence can be
described in terms of pattern detection. It is his
contention that the entire neocortex can be described as a
system of hierarchical pattern recognition. This theory
would describe all learning, including first language
acquisition, as pattern recognition. If that were the case,
it would have implications for Chomsky’s LAD and UG.
4.1 Implications for LAD
Describing the entire brain in terms of it being a pattern
recognition device explains the phenomenon of brain
plasticity. The brain’s essential nature, pattern
recognition, allows different sections of it to retain the
potential to perform a multiplicity of functions. Where
necessary, it is possible for one part of the brain to
acquire the functions of another. One of those functions is
language learning. If language learning shares its essential
nature, pattern detection, with every other kind of human
18
learning, this would cast doubt upon the existence of a
specialized component of the brain dedicated to language
acquisition. The PDTLA explains brain plasticity. The LAD
does not.
19
4.2 Implications for UG
Linguists often describe language in terms of patterns: “A
learner’s mental grammar is a system in its own right. It
has its own patterns and regularities” (Thornbury, 2006, p.
92). “[Language development] involves processes of
integrating new forms and patterns” (Spada, 2006, p. 189).
“[Learners] are able to see the patterns in the language”
(Hedge, 2000, p. 146). Therefore, one way to describe
language is simply as streams of vocabulary arranged along
certain changeable and imperfect rules. Those rules are the
patterns that language appears in. Similarly, we can
describe those patterns as embodying the grammar of every
language. According to Kurzweil’s theory, human brains can
be described as massive pattern detection agents. Therefore,
the child learning language for the first time is operating
primarily as a pattern identifier and producer.
The universal element of Chomsky’s Universal Grammar could
simply be that language comes in the form of predictable and
consistent input and the human brain is uniquely suited to
20
pick up on that patterned input. However, the capacity for
acquiring grammar should not be seen as a special biological
endowment. It is more likely to be a general capacity
(pattern detection) in common with every other kind of human
cognitive ability.
4.3 Limitations of Pattern Detection Theory of Language
Acquisition
Plaza et al. (2009) note that although their patient no
longer used Broca’s area, a degree of linguistic fragility
persisted. Brain plasticity is not universal – the examples
of successful adaptation of language functions are minimal.
Furthermore, successful examples are not perfect –
suggesting that the brain is not 100% successful at adapting
different functions to different parts of the brain.
Chomsky has never said that the LAD or the UG constitutes
the Wernicke-Broca’s area. In fact, he has never said where
the LAD or the UG are. It might turn out that the LAD and
the UG are a complex system of interactions between
21
different parts of the brain – in particular the left
hemisphere (Carstairs-McCarthy, 2000).
PDTLA does not explain the age of onset. As Chomsky has
pointed out, children have a 100% success rate in learning
their first language. After an indeterminate number of
years, the age of onset, learning a second language becomes
markedly less successful. PDTLA does not explain why a
child’s pattern detection ability is so effective for first
language acquisition and less so for learning other
languages later on in life.
PDTLA does not explain why children can think in a
particularly abstract manner when it comes to language but
not with other aspects of their thinking. A 1973 study of
children between the ages of four and seven years showed
that they could detect the pattern in a matrix sequence –
noting that in previous studies it was believed only
children eight years or older could do so (Gaines, 1973).
The implication is that children younger than four cannot
detect patterns in a matrix sequence. However, children as
22
young as 18 months are clearly developing patterned language
output of at least 2 words (Thornbury, 2005). It appears
that children’s capacity for abstract thinking is more
pronounced with regards to language use than with other
logical skills.
Furthermore, it can be argued that not all thinking is
pattern detection (emotions, compulsivity, moral
selflessness, etc.) (Marcus, 2012). It stands to reason that
it is at least possible for language acquisition to fall
outside the realm of pattern detection as well.
5. Conclusion
Chomsky’s theories of language are widely accepted as being
important for any informed understanding in the area of
human language development. His ideas, however, have been
met with increasing criticism over the years. This essay
acknowledges Chomsky’s contribution to our understanding of
language acquisition, while also pointing to some serious
objections to his theories.
23
This essay also attempts to describe an alternative theory
of language acquisition based upon the human brain’s general
pattern detection capabilities. Although a robust
alternative theory of language acquisition remains elusive,
it is hoped that the questions and ideas generated here may
contribute to the ongoing debate on the subject of language
acquisition.
[Word Count 2966]
24
References
Carstairs-McCarthy, A. (2000). Broca's area and language
evolution. Behavioral and Brain Sciences, 23(01), 28-29.
Chomsky, N. (2002). On Nature and Language. Cambridge, GBR:
Cambridge University Press.
Chomsky, N. (2005). Three Factors in Language Design.
Linguistic Inquiry, 36(1), 1-22.
Doidge, N. (2007). The brain that changes itself : stories of personal
triumph from the frontiers of brain science. New York: Viking.
Gaines, R. (1973). Matrices and pattern detection by young
children. Developmental Psychology, 9(2), 143-150.
Hedge, T. (2000). Teaching and Learning in the Language Classroom.
Oxford: OUP.
Kurzweil, R. (2013). How to Create a Mind. New York: Penguin.
Larsen-Freeman, D. (2011). Grammar. In R. C. D. Nunan (Ed.),
The Cambridge Guide to Teaching English to Speakers of Other
Languages. Cambridge: CUP.
Marcus, G. (2012). Ray Kurzweil's Dubious New Theory of Mind
http://www.newyorker.com/online/blogs/books/2012/11/ray
25
-kurzweils-dubious-new-theory-of-mind.html. Retrieved
December, 2013
Plaza, M., Gatignol, P., Leroy, M., & Duffau, H. (2009).
Speaking without Broca's area after tumor resection.
Neurocase, 15(4), 294-310.
Popper, K. (2005). The Logic of Scientific Discovery. London and New
York: Taylor & Francis e-Library.
Shermer, M. (2008). Patternicity: Finding Meaningful
Patterns in Meaningless Noise
http://www.scientificamerican.com/article/patternicity-
finding-meaningful-patterns/. Retrieved December,
2013
Spada, P. L. N. (2006). How Languages are Learned. Oxford: OUP.
Thornbury, S. (2005). Uncovering Grammar. Oxford: Macmillan.
Thornbury, S. (2006). An A-Z of ELT. Oxford: Macmillan.
26