Critical Period

The Critical Period Hypothesis:

There is a biological period during which
language can be acquired easily, perfectly, and
without an accent; after this time, it is difficult, if
not impossible, to learn language perfectly and
without an accent.

Critical Period: First Language

1.

Victor
feral child, France, 1799, 12 years old
no language, receptive to forest sounds
Dr. Jean Itard, 5 years tutoring; Victor learned
lait and O Dieu! but never used them
communicatively.
Itard, J. (1932). The wild boy of Aveyron. NY:
Century.
LEnfant Sauvage (the Wild Child), Francois
Truffaut, 1970

Critical Period: First Language

2. Genie
1970, California, 13.5 years old, isolated since
20 months (tied to bed by psychotic father),
beaten if she vocalized, father spoke only in
grunts.
After 5 years of education, she could speak,
though slowly, and with greater-than-normal
gaps between hearing and comprehension,
overuse of formulaic languagerecognizably
different from native speakers.
Rymer, R. (1993). Genie: An abused childs
flight from silence. London: Michael Joseph.
3

Critical Period: First Language

3. Deaf Children
Born to hearing parents, sometimes deprived
of exposure to sign language in infancy.
Newport, E. (1990). Maturational constraints on
language learning. Cognitive Science ,14, 11
28.
group 1: exposure since birth
group 2: exposure since school (age 4 6)
group 3: exposure after age 12.
Results showed decreasing grammaticality in
ASL among the three groups.
4

Critical Period: Second Language

1. Patkowski, M. (1980). The sensitive period for
the acquisition of syntax in a second language.
Language Learning, 30, 449472.
SL informants: group 1 began learning English
before puberty; group 2 after puberty.
Speech recorded, transcribed, rated by native
speakers on a scale from 0 (no knowledge of
English) to 5 (educated native speaker).
note that this eliminates the phonological
variable (i.e., accent).

Critical Period: Second Language

1. Patkowski, M. (1980). The sensitive period for
the acquisition of syntax in a second language.
Language Learning, 30, 449472.
What does the
irregular shape of
the distribution
mean?
Nearly everyone
rated like a native
speaker; success
in SLA is
inevitable before
puberty.

Critical Period: Second Language

1. Patkowski, M. (1980). The sensitive period for
the acquisition of syntax in a second language.
Language Learning, 30, 449472.
What does the
normal shape of
the distribution
mean?
Results vary
widely; some do
well, others do
not; success is
not inevitable
after the puberty.

Critical Period: Second Language

2. Johnson, J. & Newport, E. (1989). Critical period
effects in second language learning: The
influence of maturational state on the acquisition
of English as a second language. Cognitive
Psychology, 21, 6099.
Chinese and Korean subjects with varying ages
of arrival in the United States; all highly
educated (students and professors at
universities).
Grammaticality judgment test wide range of
morphology / syntax rules, multiple sentences,
some correct, some incorrect. Subjects make
judgments about correctness.
8

Critical Period: Second Language

2. Johnson, J. & Newport, E. (1989). Critical period
effects in second language learning. Cognitive
Psychology, 21, 6099.
What does r =
-0.87 mean?
What does this
correlation
suggest?
A pattern re:
success in SLA
(linearity), even
before puberty.
9

Critical Period: Second Language

2. Johnson, J. & Newport, E. (1989). Critical period
effects in second language learning. Cognitive
Psychology, 21, 6099.
What does r =
-0.16 mean?
What does lack
of correlation
suggest?
Learners have
widely varying
degrees of
success in SLA
10
after puberty.

Critical Period: Second Language

3. DeKeyser, R. (2000). The robustness of critical
period effects in second language acquisition.
Studies in Second Language Acquisition, 22,
499533.
This study builds on and extends Johnson and
Newport (1989), both methodologically and
conceptually.

11

Critical Period: DeKeyser (2000)

Questions for groups
1. What variable does DeKeyser add to J&N, and
how is it measured?
2. Why does DeKeyser add the variable?
3. How does DeKeyser modify J&Ns instrument
for data collection, and why?
4. What changes does DeKeyser make in the
group from which data is collected, and why?
5. How are DeKeysers age-related results similar
to, and different from, J&Ns (Question #1)?
6. What does the addition of the variable (point 2)
allow him to explain relative to Question # 2?

12

Critical Period: DeKeyser Questions

1. What variable does DeKeyser add to J&N, and
how is it measured?
Aptitude, or analytic verbal ability (p. 506).
Measured by the Modern Language Aptitude
Test. Participants completed the MLAT test (20,
5-way, multiple-choice questions) after the other
instrument and the background questionnaire.

13

Critical Period: DeKeyser Questions

2. Why does DeKeyser add the variable?
In an effort to explain the exceptions in
previous studies and observations people or
participants who are very successful with
language learning, but began the process of
language learning as adults.
Exceptions may explain the right tail in
Patkowskis distribution

14

Critical Period: DeKeyser Questions

2. Why does DeKeyser add the variable?
In an effort to explain the exceptions in
previous studies and observations people or
participants who are very successful with
language learning, but began the process of
language learning as adults.
Or the outliers in J&N

15

Critical Period: DeKeyser Questions

2. Why does DeKeyser add the variable?
In an effort to explain the exceptions in
previous studies and observations people or
participants who are very successful with
language learning, but began the process of
language learning as adults.
Or the impressionistic data in other studies
(Coppieters, 1987) or the partial overlap of the
native and nonnative distributions in Birdsong
(1992) (p. 507).

16

Critical Period: DeKeyser Questions

3. How does DeKeyser modify J&Ns instrument
for data collection, and why?
Grammaticality judgment task instrument
shortened from 276 to 200 items / sentences.
Original test may have been too long for the
participants to concentrate on every item (p.
502).

17

Critical Period: DeKeyser Questions

4. What changes does DeKeyser make in the
group from which data is collected, and why?
Native speakers of Hungarian, in and around
Pittsburgh
Long period of residence (10+ years), to
eliminate possible confusion between age of
arrival and age of test taking in J&N, where
period of residence was only minimum of 5
years.
Wide range in age of arrival and socioeconomic
status. Why socioeconomic status?
a first approximation of verbal ability [i.e.,
aptitude] (p. 508).
18

Critical Period: DeKeyser Questions

5. How are
DeKeysers
results similar
to, and
different from,
J&Ns results
(Question
#1)?
R = -0.63

19

Critical Period: DeKeyser Questions

5. How are DeKeysers results similar to J&Ns?
The correlation between age of arrival and test
score is remarkably similar.
What does this suggest?
An age-related effect in SLA.

r = -0.77

r = -0.63

20

Critical Period: DeKeyser Questions

5. How are DeKeysers results different than J&N?
Whereas J&N found a strong correlation
between age of arrival and test score before
puberty (r = -0.87), DeKeyser did not (r = -0.26).
What does this suggest?

J&N; r = -0.87

DeKeyser; r = -0.26

21

Critical Period: DeKeyser Questions

5. How are DeKeysers results different than J&N?
It might betray a lack of disruption in English
proficiency between early and later arrivals; the
uninterrupted downward linear trend might fail to
illustrate the end of Critical Period.

J&N; r = -0.87

DeKeyser; r = -0.26

22

Critical Period: DeKeyser Questions

5. How are DeKeysers results different than J&N?
Or, it might illustrate the different shapes of the
distribution before and after puberty which,
while different from J&N, nevertheless suggest a
disruption signaling the end of the CP.

J&N; r = -0.87

DeKeyser; r = -0.26

23

Critical Period: DeKeyser Questions

5. How are DeKeysers results different than J&N?
On the other hand, DeKeysers results did show
a difference between early and late arrivals
relative to aptitude.
Early arrivals showed no correlation between
English proficiency and aptitude (r = 0.07, ns).
Late arrivals showed a moderate correlation (r =
0.33; p < .05).
What does this mean?
That early arrivals did not need above average
aptitude to achieve high English proficiency.
Something else facilitated their success, maybe
Critical Period effects.
24

Critical Period: DeKeyser Questions

6. What does the
addition of the
aptitude variable
allow him to
explain relative to
Question # 2?
The exceptions
to CP predications
i.e., adult
learners who
achieve very high
levels of SL
proficiency.

25

Critical Period: DeKeyser Questions

Among [those] who
started acquiring
English after age 16
but obtained a high
score on the test
(over 175), all but
one had an aptitude
score of 6 or above
(p. 514). Only the
adults with above
average aptitude
eventually became
near native (p. 515).

26

Critical Period: Explanations

1. Neurological Explanation
Brain lateralization refers to the separation of
brain functions into right and left. Language is
controlled by the logical / analytical left side.
Lateralization is completed at around the time of
puberty. Before lateralization, the brain is
plastic, meaning that it is still developing and
dynamic, and that brain functions and paths are
not yet firm.
Some researchers have suggested that
lateralization constrains the ability to learn
language that it constitutes the close of the
Critical Period.
27

Critical Period: Explanations

2. Cognitive Explanation
Cf. Piagets theory of child development.
Children advance from the Concrete operational
stage (ages 7 to 11), wherein they begin to think
logically about concrete events, to the Formal
operational stage (age 11 ff.) wherein they
develop abstract reasoning ability.
Researchers have suggested that there is a
relationship between Piagets concrete and
formal stages and the ability to learn language.
Gaining the ability to think abstractly, which of
course means we approach tasks like language
learning differently, the Critical Period closes. 28

Critical Period: Explanations

3. Psychomotor Explanation
The Critical Period may be related to our
physical development. More specifically, it may
close with when we have completed the process
of learning how to form the sounds of our native
language.
African clicks?

29

Critical Period: Explanations

4. Affective Explanation
At a certain point in their development, children
become conscious of themselves relative to
other people. This consciousness affects their
confidence, degree of extro / introversion,
attitudes, inhibitions, and other affective aspects
(i.e., emotions and feelings).
Researchers relate the CP to this development
of affective consciousness. Before it, children
are open and uninhibited, and they learn
language easily and perfectly. After it after the
CP closes their emotions inhibit language
learning and render the result imperfect.
30

Critical Period: Explanations

5. Socio-Biological Explanation
Some scholars suggest a relationship between
the process of sexual maturation and our ability
to learn language.
At the point of sexual maturation (puberty), we
lose the ability to learn a new language without
an accent. The accent that we do acquire
marks us for potential mates, making us un/
attractive to them, thereby maintaining the purity
of the gene pool for future generations.
think hominids, race / ethnicity, social class
http://www.youtube.com/watch?v=_vF9g37FCmk&feature=a
utoplay&list=PL037415A3348CAFC5&lf=results_main&play
31
next=2

Critical Period: Hakuta, Bialystok, & Wiley

Recall how DeKeysers results differed from J&Ns.
The uninterrupted downward linear trend perhaps
failed to illustrate the end of CP. If the CP does
not end, then is there such a thing as a critical
period?
This is exactly what Hakuta et al. wonder.

J&N; r = -0.87

DeKeyser; r = -0.26

32