child is something that everyone does, like walking. By contrast, adults differ widely in their ability to learn a new
language, somefinding it relativelyeasy and somefinding it nearly impossible. That is, adult language learning is more
like playing chess or the stock market or a musical instrument, domains in which individuals differ widely in talent.
This wide variation is documented in Klein and Perdue's (1997) long-term study of second language acquisition by
immigrant workers. (There is some anecdotal evidence that early multilingualism enhances talent at adult language
learning.)
It should not be said, of course, that adults cannot learn second languagesat all. Even the least talented learn some
words and expressions. The norm, given adequate exposure and motivation, is at least some degree offluency, but
with many errors in pronunciation and in grammaticalfine points.
The difficulty of language learning for adults is sometimes attributed to their greater self-consciousness and/or lack of
the child's innocent motivation.While these may be contributing factors, I imagine we are all familiar with non-native
speakers of our own language who have the most outgoing, bubbly, unselfconscious personalities, but have not in
many years managed toachieveanythingnear nativefluency. Contrariwise, many of us were terriblyself-conscious and
inhibited as children, but still we managed to learn to speak our native languages quitefluently, thank you. (We have
already discussed in section 4.6 a related putativeexplanation:the idea that children learn language better because they
are simply not as smart as adults.)
Lenneberg (1967) ascribes the disparity between children and adults to a biological “critical period” of brain
development. He characterizes this as a time window in which the Language Acquisition Device is availableto a child,
and he cites as biological analogies the critical periods for maturation of binocular vision, for learning species-specific
songs in certain bird species, and for imprinting infants of various species on their parents. More general examples of
biological timing include puberty and development of teeth.
Elissa Newport (1990) refines the thesis of a critical period, showing thatfluency does not drop off sharply at a
particular age. Rather,fluency in certain aspects of second language acquisition correlates inversely with the age at
which one starts to learn. On average, people who start at age six get better than people who start at twelve, and still
better than people who start at eighteen; after eighteen the curveflattens out to adult levels of incompetence.
Twoindependentkinds ofdata confir mthedecayof languagelearning proficiencywithage. Thefirstcomes from deaf
individuals whosefirst exposure to sign language (and thereforefirst exposure toanylanguage) comes relativelylate in
life. The results (Newport 1990) parallel those for late second language acquisition. All speakers achieve some
competence with sign, but the greatest