The Critical Period Hypothesis: Some Problems: January 2007
The Critical Period Hypothesis: Some Problems: January 2007
The Critical Period Hypothesis: Some Problems: January 2007
net/publication/28165671
CITATIONS READS
27 6,649
1 author:
David Singleton
Trinity College Dublin
77 PUBLICATIONS 1,397 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Beyond Age Effects: Variables Trumping the Age Factor in Instructional L2 Learning View project
All content following this page was uploaded by David Singleton on 24 July 2014.
1. Introductory
In this paper I shall claim that to speak in terms of the Critical Period Hypothesis (CPH) is
misleading, since there is vast variation in the ways in which the critical period (CP) for language
acquisition is understood. I shall in addition point out that such variation also relates to the ways
in which the purported CP is interpreted in terms of its practical implications for L2 instruction.
My contention is that such variation fatally undermines the status of the CPH as a scientific
hypothesis.
The notion of CP has its origins in the biological sciences. One example of a CP that is
often cited is behaviour exhibited by ducklings whereby they follow the first moving object they
perceive after hatching:
This following behavior only occurs within a certain time period after hatching, after which point the
ducklings develop a fear of strange objects and retreat instead of following. Within these time limits
is the critical period for the following behavior.
(De Villiers & De Villiers 1978: 210)
A critical period for the development of binocularity may begin when central nervous system cells
driven by each eye grow and compete for cortical synapses (Wiesel & Hubel, 1963). This critical
period may end when the degree and extent of competitive synaptogenesis diminishes and stabilizes,
perhaps regulated in part by the amine system ( Kasamatsu & Pettigrew, 1979). The critical period for
development of binocularity may take place between weeks 4 and 12 in the cat; 1 and 9 in certain
monkeys; and years 1 and 3 in man.
(Almli and Finger 1987: 126)
On the basis of the above examples, CPs can be characterised as being of limited duration
within well-defined and predictable termini and as being related to very specific capacities or
behaviours. We shall see that the precise termini proposed for maturational constraints on
language acquisition by CP advocates vary across quite a wide range and that there is no
consensus either regarding the particular acquisitional capacities that are deemed to be affected by
such constraints. We shall further see that researchers diverge markedly in respect of their
identification of the underlying causes of maturational constraints on language acquisition.
3. CP Termini
Penfield is widely seen as at least a proto-CP theorist and so let us begin our discussion of
the termini of the CP with him. Penfield suggests that 'for the purposes of learning languages, the
human brain becomes progressively stiff and rigid after the age of nine' (Penfield and Roberts
1959: 236) and that 'when languages are taken up for the first time in the second decade of life, it
is difficult ... to achieve a good result ... because it is unphysiological' (Penfield and Roberts
1959: 255).
SINGLETON, David, “The critical period hypothesis: some problems”
Lenneberg, for his part, proposes (1967) not only an end point for the CP (puberty), but
also a point at which it supposedly has its onset (age two), the intervening period purportedly
coinciding with the lateralization process - the specialization of the dominant hemisphere of the
brain for language functions. Lenneberg further suggests (1967: 142) that the developmental stage
on which language acquisition is predicated ‘is quickly outgrown at the age of puberty,’ the point
at which he posits the lateralization process to be complete. With respect to L2 acquisition, he
asserts (ibid.: 176) that after puberty ‘the incidence of “language-learning-blocks” rapidly
increases’, ‘foreign languages have to be … learned through a conscious and labored effort, and
‘[f]oreign accents cannot be overcome easily’.
There has been a great deal of debate about lateralization, much of it focusing on the
completion point. Molfese (1977: 206f.), for example, suggests that the lateralization of certain
'low-level' functions of phonetic and/or phonological character may in fact be 'complete by the
first year of life,’ while the sensorimotor cognitive structures underlying the child's early use of
syntax and semantics may be 'represented in both hemispheres,’ in which case 'his meaningful
linguistic utterances will be mediated by both hemispheres, although perhaps not to the same
degree.’ This approach predicts a multiplicity of CPs, with the phonetic/phonological CP being
posited as coming to an end very early.
Other researchers propose an earlier end of the phonetic/phonological CP on different
grounds. Seliger (1978) argues that as well as a lateralization process there is a localization
process within the dominant hemisphere. He suggests that phonetic/phonological functions are
localized by puberty and that syntactic functions are localized subsequently, thus remaining
acquirable until later in life. Yet another advocate of the multiple critical periods perspective is
Diller (198l: 76), who notes that pyramidal cells, which he associates with phonetic/phonological
acquisition, develop by age 6/8, whereas stellate cells, which he links with higher order functions,
mature over two to three decades; he infers that authentic L2 accents can be acquired only by
young children whereas cognitive aspects of L2s can be learned by relatively mature people.
Scovel (1988: 101) also distinguishes pronunciation from other areas of language,
claiming that it shows maturational effects because it has a ‘neuromuscular basis.’ He suggests
that acquiring other aspects of language is fundamentally different from learning pronunciation,
because vocabulary, morphosyntax, etc.do not, unlike phonetics/phonology, have a ‘physical
reality.’ He claims that those who begin to be exposed to an L2 after age 12 cannot ever ‘pass
themselves off as native speakers phonologically’ (Scovel 1988: 185; cf. Scovel, 2000, 2006).
Other L2 researchers claim that the offset of the CP is progressive, this process beginning
around age 6 or 7. Johnson and Newport (1989) infer from their research that there is a specific
maturational phase – up to about 7 years – which is particularly favourable for language learning
and a second phase - from about 7 years to about puberty – during which the language learning
capacity declines gradually but subsequent to which there is a very abrupt deterioration. Long
(1990) accepts Johnson and Newport’s evidence in relation to an early beginning to the decine of
the language acquiring capacity and agrees with Scovel’s suggestion that age 12 constitutes the
point beyond which a native-like L2 accent cannot be acquired. He also asserts that the
prerequisite for the acquisition of L2 morphology and syntax to native levels is exposure to the
L2 before age 15.
Ruben (1997) takes a more radical line. On the basis of studies of the effects of very early
temporary hearing impairment, he concludes that the CP for phonetics/phonology ends around the
twelfth month of infancy. Ruben further reads the research literature as indicating that the CP for
syntax ends in the the fourth year of life, and for semantics in the fifteenth or sixteenth year of
life. Hyltenstam and Abrahamsson also favour a very early CP offset. Their review of evidence of
differences between even very early L2 acquirers and native speakers, leads them to a certain
dubiousness about the CPH, and to state in one publication that the CP may be ‘une chimère’
(2003b: 122). Elsewhere they speculate that the language learning mechanism may be ‘designed
in such a way that it … inevitably and quickly deteriorates from birth’ (2003a: 575), and for this
reason ‘nativelike proficiency in a second language is unattainable’ (ibid.: 578).
A succinct comment on the state of the evidence regarding CP termini comes from Aram et
al. (1997): ‘the end of the critical period for language in humans has proven… difficult to find,
with estimates ranging from 1 year of age to adolescence’ (p. 285). The foregoing survey, though
brief and selective amply suffices to confirm the well-foundedness of this comment.
As we have seen, whereas Lenneberg (1967) and others see the CP as applying to language
acquisition across the board, Scovel (1988) limits its application to the phonetic/phonological
sphere. The notion that maturational constraints impact only on the acquisition of
phonetics/phonology has a long history. Thus, for example, Dunkel & Pillet (1957) found that
adult L2 beginners outperformed elementary school pupils in comprehension tests, but that
younger beginners’ pronunciation was 'superior to that achieved by older beginners' (p. 148),
concluding that pronunciation is 'the most rewarding aspect’ of early L2 learning. Fathman (1975)
and Fathman & Precup (1983) reached similar conclusions in studies of immigrants acquiring
English in the US.
Other researchers have hypothesized constraints of a different nature on the operation of
the CP. For instance, Martohardjono & Flynn (1995) claim that, whereas aspects of L2
proficiency unrelated to innate elements may be subject to age-related degradation, aspects
supported by biological endowment (Universal Grammar – UG) are likely to resist such
degradation. Other researchers (e.g. Bley-Vroman 1989; Schachter 1988) claim that post-
pubertal language learning has no access to UG, explaining L2 learning without UG in terms of
general problem solving mechanisms (seen as not constrained by maturation) plus L1 knowledge
(Bley-Vroman 1989). There is also a ‘partial access’ position on this issue. Thus, Hawkins (2003;
Hawkins & Yuet-hung Chan 1997) suggests that some features of UG are inaccessible to late L2
learners, resulting in mental representations which diverge from those of the native speaker.
A further – non UG-based - distinction is proposed by Dekeyser (2000, 2003a). In his
(2000) study, the adult beginners who scored within the range of the child beginners evinced
higher levels of verbal analytical ability, an ability which, according to him, played no role in the
performance of the child beginners. DeKeyser concludes that maturational constraints apply only
to implicit language learning mechanisms - that ‘somewhere between early childhood and puberty
children gradually lose the ability to learn a language successfully through implicit mechanisms
only’ (DeKeyser 2003a: 335).
As in the case of the survey of proposed termini for the CP, the above tour d’horizon of
proposals with respect to the domains affected by the CP is very far from exhaustive. However,
once again, even a very brief survey of this kind reveals quite a range of different points of view
in this matter.
5. Causes of the CP
As is already clear from previous discussion, many of the widely cited proposals regarding
causes of the CP refer to neurobiology. Penfield (Penfield and Roberts 1959) talks about
decreasing cerebral plasticity; Lenneberg (1967) and Molfese (1977) refer to the process of
lateralization of language functions; Seliger (1978) evokes the localization of specific language
subfunctions within the dominant hemisphere; and Diller (1981) bases his arguments on the
maturation timetables of different types of brain cell. A further perspective is offered by
Pulvermüller and Schumann (1994), who note that, as the brain matures, the axons of neurons are
progressively wrapped by glial cells – a process labelled myelination – after the substance myelin
contained in the glial cells. Pulvermüller and Schumann claim that myelination reduces plasticity
in the language areas of the brain until around puberty, that plasticity then remains low, and that
this is what explains age effects in language acquisition.
Other treatments of the neurobiological dimension of this issue have derived from studies
using brain-imaging techniques. Thus, Kim et al. (1997) investigated spatial representation of L1
and L2 in the cortex of early and late bilinguals during a sentence-generation task. What they
found was that in respect of Broca’s area late bilinguals showed two adjacent centres of activation
for L1 and L2, whereas in the early bilinguals a single area of activation for both languages
emerged. Similarly, Wattendorf et al. (2001) looked at subjects (i) exposed to two languages
before age 3 and to a third after age 10, and (ii) exposed to only one language up to age 10 and
had then to two further languages. In the early bilinguals the zones in Broca’s area activated by
the use of the first two languages overlapped, whereas in the late multilinguals the pattern of
activation was more diffuse. Whether such effects are related to age of onset or to degree of
proficiency is still a matter of dispute.
In sum, there is ‘no consensus’ (Hyltenstam & Abrahamsson 2003a: 563) regarding
neurobiological interpretations of age effects. Nor, of course, are other varieties of explanation
exempt from neurobiological implications; the issue is that of directionality. Recent work in the
neurosciences (see e.g. Bialystok and Hakuta 1999; Gazzaniga 1992; Robertson 1999) shows that
brain structure and organization are as likely to reflect different kinds of experience as to
determine these experiences, which means that features of structure and organization may result
from rather than cause different cognitive, affective and other aspects of language acquisition –
including level of proficiency (see e.g. Perani et al. 1998; Abutalebi et al. 2001).
Some researchers’ explanations for the CP have focused not on neurobiology but on
Piagetian psychology, and in particular on the impact of the onset of ‘formal operations’ (FO)
during adolescence. Krashen (1975) homes in on the development of an interest in general
'systems' and 'theories' rather than ad hoc solutions, suggesting that this 'general tendency of
adolescents to construct theories' (Inhelder and Piaget 1958: 336) might inhibit the 'natural' and
complete acquisition of an L2. Rosansky (1975) focuses on the awareness of contradictions
associated with FO, which she sees as being deleterious to language acquisition. Felix (1981)
sees the ‘problem-solving cognitive structures’ associated with FO as competing with the
language-acquiring structures of the language faculty.
A more recent perspective on the role of general cognitive factors in relation to the age
question is provided by DeKeyser (2000, 2003a, 2003b, 2006), who posits an inescapable decline
in language acquiring capacity because of cognitive maturation. He suggests that there a
diminishing capacity for the implicit learning of complex abstract systems – including language
but not excluding other complex systems. (cf. Bialystok (1997, 2002; Bialystok & Hakuta 1994,
1999; Hakuta, Bialystok & Wiley 2003).
Explanations for the CP also refer to affective-motivational factors. For example, Krashen
(1982: 216) claims that the 'affective filter' (cf. Dulay and Burt 1977) is strengthened at puberty
thanks to the onset of formal operations. His view (1985: 13) is that ‘[w]hile the filter may exist
for the child second language acquirer, it is rarely, in natural informal language acquisition
situations, high enough to prevent native-like levels of attainment,’ whereas ‘[f]or the adult it
rarely goes low enough to allow native-like attainment.’
Another batch of explanations are Freudian in inspiration. An early Freudian account of
age effects in L2 acquisition is that of Stengel (1939) account, in terms of – inter alia:
identification, the desire to be like others, which he sees as underlying the phenomenon of
'echolalia' in children, which plays no role in normal adult language learning (pp. 471f.); the
super-ego, which, he says plays an inhibiting role in the adult but is relatively undeveloped in the
child (p. 473); and narcissism which Stengel claims occasions the ‘sense of shame' many adults
feel when they start to use a new language (p. 476). Another concept borrowed from Freud is that
of hardening ego-boundaries. Taylor’s (1974: 33) approach to maturational constraints is
informed by this perspective but it is particularly emphasized in the work of Guiora, who (1972,
1992; Guiora, Brannon and Dull 1972; Guiora, Beit-Hallahmi, Brannon, Dull and Scovel 1972)
introduces the construct of language ego, Guiora’s proposals claim that learning an L2 requires
the taking on of a new identity and that in order for this to happen language ego boundaries have
to be permeable, which, typically, they are not in the adult.
Schumann also takes affective-motivational factors into account in his discussion of the
CPH. In his earlier work indeed (e.g. 1975, 1978) he related post-pubertal L2 learning difficulties
principally to ‘social and psychological changes’ (Schumann 1975: 229) and to social and
psychological distance. In his more recent discussion of these issues he continues to see affective-
motivational factors as playing a countervailing role in respect of neurobiological maturational
constraints (Schumann 1994; cf. Moyer 1999, 2004).
Once again it is worth emphasising that the above short account of relevant research, like
those in the previous sections, could be extended significantly (see, e.g. Singleton & Ryan 2004:
Chapter 5). However, despite its curtailed and partial nature, the foregoing overview reveals a
wide variety of opinions on the matter of causes of the CP.
We come now to the question of practical implications of the CPH. We can begin by noting
that advocacy of early L2 instruction in the 1950s-60s was much influenced by the ideas of
Penfield regarding the difficulty of acquiring a language in the second decade of life (Stern 1983:
132). It is clear then that the notion of a critical period had a major impact on decisions
concerning early L2 instruction.
Some present-day CPH theorists take essentially the same line as Penfield. Spada, for
example, recently made the following assertion in an interview:
If the goal for learning/teaching a foreign language is to obtain the highest level of second
language skills ... there is support for the argument that ‘earlier is better.’ This support, found in
the critical period hypothesis literature, is based on the claim that biological and maturational
factors constrain language learning beyond a certain age.
(ReVEL 2004)
However, this is by no means a unanimous view among those who favour the CPH.
Johnson & Newport (1989) conclude that ‘the learning which occurs in the formal language
classroom may be unlike the learning which occurs during immersion, such that early instruction
does not necessarily have the advantage for ultimate performance that is held by early immersion’
(p. 81). DeKeyser takes a rather similar line:
Rather than suggesting the importance of starting early, [age differences] indicate that the
instructional approach should be different depending on age: full-scale immersion is necessary for
children to capitalize on their implicit learning skills, and formal rule teaching is necessary for
adults to draw on their explicit learning skills.’
(DeKeyser 2003a: 335)
Contrariwise, there are CPH sceptics who energetically support the early introduction of
L2s into the curriculum. For example, Ekstrand (1971) Genesee (1978) and Hatch (1983) all
argue for early L2 instruction, but not because they accept the idea of maturational constraints,
but rather on the basis of their belief in the importance of factors such as the desirability of as
long an exposure to the L2 as possible and the importance of laying an early foundation to L2
learning to maximise the ground that can be covered.
In other words, the consensus which is lacking with regard to the termini, the domains of
operation and the causes of the CPH is also completely absent in respect of what implications it
might or might not have for L2 instruction.
7. Concluding remarks
The variety of propositions to be found under the CPH heading mean that the CPH
cannot plausibly be regarded as a scientific hypothesis. Nor does the option of reducing the
various versions of the CPH to a single summary form help. Such a summary form would look
something like this: ‘for some reason, the language acquiring capacity, or some aspect or aspects
thereof, is operative only for a period which ends some time between perinatality and puberty.’
This is not a hypothesis either; it is at best an extremely vague promissory note.
In the context of research into social and emotional development, the child psychologist
Ross Thompson voices the following conclusion about the CP concept.
While the metaphor of critical periods in the organization of neural systems in the visual cortex or
of imprinting in lower species may offer an attractive heuristic to students of human development,
the complexity of the behavioral systems to which these concepts are applied in young children
makes it difficult, if not impossible, to identify the parameters of sensitive periods with
appropriate specificity.
(Thompson 2001: 87)
These seem to me to be words that CPH theorists would do well to ponder on.
References
Abutalebi, Jubin, Stefano Cappa and Daniela Perani (2001). The bilingual brain as revealed by
functional neuroimaging. Bilingualism: Language and Cognition 4: 179-190.
Almli, C. Robert and Stanley Finger (1987) Neural insult and critical period concepts. In Sensitive
Periods in Development: Interdisciplinary Perspectives, Marc H. Bornstein (ed.), 123-
143. Hillsdale, NJ: Lawrence Erlbaum.
Aram, Dorothy. Elizabeth Bates, Julie Eisele, Judi Fenson, Ruth Nass, Donna Thal and Doris
Trauner (1997). From first words to grammar in children with focal brain injury.
Developmental Neuropsychology 13: 275-343.
Bialystok, Ellen (1997). The structure of age: In search of barriers to second language acquisition.
Second Language Research 13: 116-137.
Bialystok, Ellen (2002). On the reliability of robustness: A reply to DeKeyser. Studies in
Second Language Acquisition 24: 481-488.
Bialystok, Ellen and Kenji Hakuta (1994). In Other Words: The Science and Psychology of
Second Language Acquisition. New York: Basic Books.
Bialystok, Ellen and Kenji Hakuta (1999). Confounded age: Linguistic and cognitive factors in
age differences for second language acquisition. In Second Language Acquisition and the
Critical Period Hypothesis, David Birdsong (ed.), 161-181. Mahwah, NJ: Lawrence
Erlbaum.
Bley-Vroman, Robert (1989). What is the logical problem of foreign language learning? In
Linguistic Perspectives on Second Language Acquisition, Susan M. Gass and Jacquelyn
Schachter (eds.), 41-68. Cambridge: Cambridge University Press.
DeKeyser, Robert (2000). The robustness of critical period effects in second language acquisition.
Studies in Second Language Acquisition 22: 499-533.
DeKeyser, Robert (2003a). Implicit and explicit learning. In The Handbook of Second Language
Acquisition (eds.), Catherine Doughty and Michael H. Long (eds.), 313-348. Malden,
MA: Blackwell.
DeKeyser, Robert (2003b). Confusion about confounding: The critical period and other age-
related aspects of second language learning. Paper presented at ELIA VIII (Encuentros de
Linguistica Inglesa Aplicada), El factor edad en la adquisición y enseñanza de L2,
Seville.
DeKeyser, Robert (2006). A critique of recent arguments against the critical period hypothesis.
In Age in L2 Acquisition and Teaching, Christián Abello-Contesse, Rubén Chacón-
Beltrán, M. Dolores López-Jiménez, M. Mar Torreblanca- López (eds), 49-58. Bern:
Peter Lang.
De Villiers, Jill and Peter De Villiers (1978). Language Acquisition. Cambridge, MA: Harvard
University Press.
Diller, Karl (1981). 'Natural methods' of foreign language teaching: Can they exist? What criteria
must they meet? In Native Language and Foreign Language Acquisition, Harris Winnitz
(ed.), 75-91. New York: The New York Academy of Sciences.
Dulay, Heidi and Marina Burt (1977). Remarks on creativity in language acquisition. In
Viewpoints on English as a Second Language, Marina Burt, Heidi Dulay and Mary
Finnochiaro (eds), 95-126. New York: Regents.
Dunkel, Harold B. and RogerA. Pillet (1957). A second year of French in the elementary school.
Elementary School Journal 58, 142-151.
Fathman, Ann (1975). The relationship between age and second language productive ability.
Language Learning 25: 245-53.
Fathman, Ann and Lois Precup (1983). Influences of age and setting on second language oral
proficiency. Second Language Acquisition Studies, Kathleen M. Bailey, Michael H. Long
and Sabrina Peck (eds.), 151-161. Rowley, MA: Newbury House.
Felix, Sascha (1981). Competing cognitive structures in second language acquisition. Paper
presented at the European-North American Workshop on Cross-Linguistic Second
Language Acquisition Research, Lake Arrowhead CA, September. Cited and discussed in
Birgit Harley (1986) Age in Second Language Acquisition. Clevedon: Multilingual
Matters, 12.
Gazzaniga, Michael S. (1992). Nature’s Mind: The Biological Roots of Thinking, Emotions,
Sexuality, Language, and Intelligence. New York: Basic Books.
Genesee, Fred (1978). Is there an optimal age for starting second language instruction? McGill
Journal of Education 13: 145-54.
Guiora, Alexander (1972). Construct validity and transpositional research: Toward an empirical
study of psychoanalytic concepts. Comprehensive Psychiatry 13: 139-150.
Guiora, Alexander (1992). Notes on the psychology of language – the Gent lectures. Scientia
Paedogogica Experimentalis 29 (Supplement): vii-62.
Guiora, Alexander, Benjamin Beit-Hallahmi, Robert Brannon, Cecelia Dull and Thomas Scovel
(1972). The effects of experimentally induced changes in ego states on pronunciation
ability in a second language: An exploratory study. Comprehensive Psychiatry 13: 421-
428.
Guiora, Alexander, Robert Brannon and Cecelia Dull (1972). Empathy and second language
learning. Language Learning 22: 111-130.
Hakuta, Kenji, Ellen Bialystok and Edward Wiley (2003). Critical evidence: A test of the Critical
Period Hypothesis for second-language acquisition. Psychological Science 14: 31-38.
Hatch, Evelyn (1983). Psycholinguistics: A Second Language Perspective. Rowley, MA:
Newbury House.