Emergent Fossilization by Brian MacWhinney
Emergent Fossilization
The following
hypotheses attempt to capture the Age of Arrival (AoA) and fossilization
patterns using clearly stated mechanisms for processing and learning.
- The lateralization hypothesis (Lenneberg, 1967): Lenneberg
claimed that AoA effects arise from maturational changes in the brain
related to lateralization of function. The idea here is that the massive
changes that occur at puberty serve to terminate the language learning
abilities of the child by finalizing the process of lateralization.
However, this account has not received much empirical support.
- The neural commitment hypothesis (Lenneberg, 1967): Lenneberg
proposed that another mechanism for AoA effects involves a progressive
maturational commitment of language areas to linguistic functioning. This
hypothesis would claim that some specific brain region or set of regions
is undergoing commitment in parallel with the period during which AoA
effects are demonstrated for learners.
- The parameter-setting hypothesis (Flynn, 1996): Generative
linguists such as Flynn have proposed that first language learning depends
on a process of parameter setting specified by universal grammar (UG).
Some UG researchers have favored a second model that holds that the
parameters of UG that were available to the child are no longer available
to the adult older second language learner.
- The metabolic hypothesis (Pinker, 1994): Pinker has
proposed an innovative account of AoA effects based on notions from
cognitive neuroscience and evolution. He suggests (p.293) that “a decline
in metabolic rate and the number of neurons during early school-age years”
is a probable cause of the loss of language-learning ability. However,
there is no evidence for any overall loss in metabolic activity of the
type Pinker is suggesting.
- The reproductive fitness
hypothesis
(Hurford & Kirby, 1999): They reason that, over the course of human
evolution, the attainment of complete fluency in a first language was a
major determinant of reproductive fitness. If a child had not successfully
acquired language by the age of sexual maturity, they would not be as
attractive to a sexual partner and would therefore be less likely to
produce offspring. Conversely, those who had acquired a high level of
language ability would be highly attractive and would reproduce…The wide
range of individual differences in successful mastery of a language after
early children does indeed suggest that a variety of fairly recent
evolutionary pressures have been operating to produce the observed
population diversity.
- The aging hypothesis (Barkow et al., 1992): Proponents
of this hypothesis believe that as they age, most people begin to
experience a marked slowdown in metabolic activity, energy, and
flexibility. Core cognitive functions such as the storage of new memories
and the retrieval of old memories can be disrupted by degeneration in the
hippocampus and temporal lobe…Aging is certainly not going to account for
all age-related effects, since it tells us nothing about changes before
45. However, it is clearly a contributory factor to some fossilization.
- The fragile rate hypothesis (Birdsong, 2005): Birdsong
suggests that, with increasing age, learners may have more problem
acquiring irregular forms as opposed to regular forms. He suggests that
irregular forms may include not only words with irregular inflections, but
also irregular use of particles and prepositions in phrasal verbs.
Birdsong seeks to ground the decline in learning of irregulars on
neuroanatomical changes in the parts of the brain subserving the
declarative memory system. There is little evidence to support his
hypothesis.
- The starting small hypothesis (Elman, 1993): Newport (1990)
argues that “language learning declines over maturation precisely because
cognitive abilities increase.” The idea is that children have a smaller
short-term memory span and that this shorter span makes it more difficult
for them to store large chunks of utterances as formulaic items. As a
result, children are forced to analyze language into its pieces.
- The entrenchment hypothesis (Marchman, 1992): Those who
advocate the entrenchment hypothesis believe that when we practice a given
skill thousands of times, we soon find that it has become automated or
entrenched. The more we continue to practice that skill, the deeper the
entrenchment and the more difficult it becomes to vary or block the use of
the skill. Entrenchment occurs in neural networks when a high frequency
pattern is presented continuously in the input training data.
- The entrenchment and balance
hypothesis
(MacWhinney, 2005): According to MacWhinney, catastrophic interference can
be solved by systems that emphasize the lexical and item-based nature of
second language learning. Moreover, these lexically-grounded systems can
also illustrate another important aspect of second language learning. This
is the parasitic nature of L2 learning when L2 is already well
consolidated. Parasitism occurs because the L1 form is already well
consolidated and entrenched by the time the learner tries to add the L2
form to the map. But what happens when both L1 and L2 are acquired
simultaneously during childhood…In this way, simultaneous bilingual
acquisition tends to minimize the misleading effects of transfer and
parasitism.
- The social stratification
hypothesis
(MacWhinney, 2006): Acculturation during adulthood involves a far greater
diversity of situations. In some cases, immigrants may marry into the L1
community, thereby guaranteeing basic acceptance and access to input.
However, even in these cases, they will not be treated in as supportive a
manner as a parent treats a child. When they are overtly corrected, they
will feel a certain loss of prestige that can strain social relations.
Outside of marriage, L2 acculturation may succeed through work groups and
casual social groups.
- The compensatory strategies
hypothesis (MacWhinney,
2006): The variance we find in adult L2 attainment must certainly be
explained in large part by variation in the social contexts facing,
immigrants and the ways in which they deal with these contexts socially.
However, even in unfavorable social situations, learners can make use of
compensatory learning strategies. These strategies are designed to
directly combat the effects of increased L1 entrenchment, as well as the effects
of biological again. There are at least three major strategies: input
maximization, recoding, and resonance.
i.
Input
maximization: receiving input from books, dictionaries, movies, lectures, practice using the language as much
as possible.
ii.
Recoding:
representing the new word orthographically.
iii.
Resonance:
establishing a series of associative relations between words and meanings that
can allow the learner to maintain a vivid image of the word until the relations
are consolidated.
References:
Barkow, J.,
Cosmides, L. and Tooby, J. (eds.) (1992) The Adapted Mind: Evolutionary Psychology and the
Generation of Culture. New York: Oxford
University Press
Birdsong, D.
(2005) Interpreting age effects in second language acquisition. In J.F. Kroll and A.M.B. deGroot (eds) Handbook
of Bilingualism: Psycholinguistic
Approaches (pp. 109-127). New York: Oxford University Press.
Elman, J.
(1993) Incremental learning, or the importance of starting small. Cognition 48, 71-99.
Flynn, S.
(1996) A parameter-setting approach to second language acquisition. In W.C. Ritchie and T.K. Bhatia (eds.) Handbook
of Second Language Acquisition
(pp. 121-158). San Diego: Academic Press.
Hurford, J.
and Kirby, S. (1999). Co-evolution of language size and the critical period. In D. Birdsong (ed.) Second
Language Acquisition and the Critical Period
Hypothesis (pp. 39 – 63). Mahwah, NJ: Lawrence Erlbaum Associates.
Lenneberg, E.
H. (1967) Biological Foundations of Language. New York: Wiley.
MacWhinney, B.
(2005) A unified model of language acquisition. In J.F. Kroll and A.M.B. deGroot (eds) Handbook of
Bilingualism: Psycholinguistic Approaches
(pp. 49-67). New York: Oxford University Press.
MacWhinney, B.
(2006) Emergent Fossilization. In Han, Z. & Odlin, T. (eds) Studies of Fossilization in Second
Language Acquisition. Clevedon: Multilingual
Matters.
Marchman, V.
(1992) Constraint on plasticity in a connectionist model of the English past tense. Journal of
Cognitive Neuroscience 4, 215-234.
Pinker, S.
(1994) The Language Instinct. New York: William Morrow.
I certainly hope to push through and not let fossilization happen to me. This is a nice breakdown of the different theories.
ReplyDeleteI see similarities between the way natural brains and neural nets (NNs) may work. I imagine that a brain is like a NN that may use low-level confirmation bias to ignore new evidence in favor of comfortably maintain previously trained networks.
Thank for commenting on the Emergent Fossilization post. It seems that you are still learning Spanish as a second/foreign language. I have no doubt that you will push through and not fossilize in any of the five domains: syntax, morphology, phonology, semantics, and pragmatics.
ReplyDeleteI like your NNs parallel to learning a second language. There is evidence in neurolinguistics that when an adult learns a second language, new neurons fire in the brain and accretive language acquisition happens. You make a valid point that second language learners achieve limited proficiency in their target language because they may overlook patterns they have not internalized yet and rely on the ones they have already established.