L-2 acquisition of gender

[Annette Karmiloff-Smith]

>Date: Wed, 28 Jul 1999 09:03:37 -0700 (PDT)
>From: "Dan I. SLOBIN" <slobin at cogsci.Berkeley.EDU>
>To: Annette Karmiloff-Smith <a.karmiloff-smith at ich.ucl.ac.uk>
>Subject: Re: L-2 acquisition of gender
>MIME-Version: 1.0


>Excerpt from: Slobin, D. I. (1996). From "thought and language" to
>"thinking for speaking."  In J. J. Gumperz & S. Levinson (Eds.),
>_Rethinking linguistic relativity_ (pp. 70-96).  Cambridge: Cambridge
>University Press.
>
>
>FIRST-LANGUAGE THINKING IN SECOND-LANGUAGE SPEAKING
>
>	There is something dissatisfying in limiting ourselves to evidence
>that is so bound up with the acquisition and use of native languages.  I
>have also suggested that the stability of grammaticized categories in
>historical language change can be taken as evidence of the cognitive
>importance of those categories for speakers. In conclusion, I would like
>to point to another type of evidence that seems to support my proposal
>that the ways in which learning a language as a child constrains one's
>sensitivity to what Sapir called "the possible contents of experience as
>expressed in linguistic terms."
>
>	Consider the small collection of linguistically encoded
>perspectives that we have been examining:  temporal contours of events
>marked by aspectual forms, movement and trajectories in space (and also
>indication of definiteness of participants mentioned in connected
>discourse). These are precisely the sorts of things that make it so hard
>to master the grammar of a second language.  For example, it is very hard
>for English-speakers to grasp the Spanish perfective/imperfective
>distinction that is lacking in our native language.  In fact, we seem
>never to fully master this system in Spanish. By contrast, we have little
>difficulty in figuring out how to use the Spanish progressive and perfect,
>or the Spanish definite and indefinite articles--since we have already
>learned how to make decisions about the linguistic expression of these
>notions in English.  But there is nothing inherently easy or hard about
>*any* of these Spanish distinctions.  For example, native French speakers
>have no trouble with the Spanish imperfective, since they have a similar
>category in French; but the progressive and perfect pose problems to them,
>since these are not French ways of looking at events.  Turkish speakers
>have difficulty with definite and indefinite articles in learning to speak
>Spanish, English, and German, since there are no definite articles in
>Turkish.  German speakers of English use the progressive where they should
>use simple present, although Turks do not make this error in English,
>since Turkish uses progressive aspect and German does not.  Spanish
>learners of English object that we make too many obscure distinctions with
>our large collection of locative prepositions and particles.  And so on.
>In brief, each native language has trained its speakers to pay different
>kinds of attention to events and experiences when talking about them.
>This training is carried out in childhood and is exceptionally resistant
>to restructuring in adult second-language acquisition.
>
>	Much of value for the thinking for speaking hypothesis could be
>learned from a systematic study of those systems in particular second
>languages that speakers of particular first languages find especially
>difficult to master.  Suggestive data of precisely this sort come from a
>European Science Foundation project, "Second Language Acquisition by Adult
>Immigrants" (Perdue, in press).  Two examples, one from the domain of time
>and the other from the domain of space, are instructive.
>
>	Consider Italian- and Punjabi-speaking immigrants to Britain
>(Bhardwaj, Dietrich, & Noyau, 1988).  Italian and English are both
>"tense-prominent" languages--that is, every finite clause must be
>grammatically marked as to its deictic relation to the moment of speaking.
>And Italian immigrants readily acquire English past-tense forms.  This
>makes it possible for them to construct narratives from a
>situationally-external perspective, relating a succession of past events
>as seen from the present, as is done in Italian.  These speakers make far
>more frequent use of tense-marking than of the English progressive aspect.
>Punjabi, by contrast, is an "aspect-prominent" language, and Punjabi
>immigrants make heavy use of the English progressive to narrate events
>"from within," from the perspective of the protagonist, in analogous
>fashion with the narrative use of the Punjabi imperfective.
>
>	In the domain of space the influence of Punjabi on learners' early
>organization of English is striking (Becker, Carroll, & Kelly, 1988).  In
>Punjabi, spatial locations are regions named by nouns, analogous to
>English expressions such as "on THE TOP of the pile" and "at THE BACK the
>house".  Punjabi learners of English often treat prepositions as nouns,
>producing forms such as "put THE ON please", "put THE DOWN chair", and
>"pull THE UP" (Becker et al., p. 69).  English relational terms have
>apparently been reanalyzed as names of locations.  In addition, Punjabi
>focuses on states as the results of processes (somewhat like the Spanish
>use of participles discussed above).  This pattern also transfers to
>English.  For example, a newspaper lying on a table is referred to as "put
>in the table" by a Punjabi-speaker.  The investigators suggest that "he
>imagines that the newspaper was put there by someone.  In Punjabi one says
>exactly the same thing" (Becker et al., p. 73).
>
>	The European Science Foundation team concludes (Bhardwaj et al.,
>1988, p. 86):
>
>	"The influence of the lexico-grammatical systems of both the SL
>[source language] and the TL [target language] can be observed in the
>acquisition process.  The picture which emerges is quite a simple one--an
>adult acquirer tries to discover in the TL a system that is similar to
>that of his SL, and if he does not discover any, he tries to construct
>one; but since it is the TL material he has to use, the outcome is
>invariably a hybrid which is an autonomous system (often consisting of
>loosely or tightly integrated sub-systems) which partakes of some features
>of both the `parent' systems but is identical to neither."
>
>	I propose that the grammaticized categories that are most
>susceptible to SL influence have something important in common: *they
>cannot be experienced directly in our perceptual, sensorimotor, and
>practical dealings with the world*.  To be sure, all human beings
>experience sequences of events that have particular temporal contours, put
>objects in locations, and so on.  Indeed, animals do the same.  But only
>language requires us to *categorize* events as ongoing or completed,
>objects as at rest or as at the end point of a trajectory, and so forth.
>Other categories seem to be less dependent on purely verbal
>categorization. I would imagine, for example, that if your language lacked
>a plural marker, you would not have insurmountable difficulty in learning
>to mark the category of plurality in a second language, since this concept
>is evident to the nonlinguistic mind and eye.  Or if your language lacked
>an instrumental marker it should not be difficult to learn to add a
>grammatical inflection to nouns that name objects manipulated as
>instruments.  Plurality and manipulation are notions that are obvious to
>the senses. [*]
>
>----------------------------------
>[*] You may have difficulty remembering to use these markers on every
>occasion--as Chinese speakers of English do not always mark the plural, to
>take one possible example. But this is a matter of automatizing attention,
>which may be difficult in adulthood.  What I am proposing is that some
>grammaticized categories may be obvious on nonlinguistic grounds.  For
>such categories, the problem in second-language learning is not to make
>the proper conceptual distinction, but to treat it as obligatory.
>----------------------------------
>
>	But there is nothing in everyday sensorimotor interactions with
>the world that changes when you describe an event as "She WENT to work" or
>"She HAS GONE to work," or when you refer to the same object in successive
>utterances as "A*car" and "THE car."  Distinctions of aspect,
>definiteness, voice, and the like, are par excellence, distinctions that
>can only be learned through language, and have no other use except to be
>expressed in language.  They are not categories of thought in general, but
>categories of thinking for speaking. It seems that once our minds have
>been trained in taking particular points of view for the purposes of
>speaking, it is exceptionally difficult to be retrained.
>
>	It is interesting that Wilhelm von Humboldt anticipated these
>questions as well.  He wrote (1836/1988, p. 60):
>
>"To learn a foreign language should therefore be to acquire a new
>standpoint in the world-view hitherto possessed, and in fact to a certain
>extent this is so, since every language contains the whole conceptual
>fabric and mode of presentation of a portion of mankind.  But because we
>always carry over, more or less, our own world-view, and even our own
>language-view, this outcome is not purely and completely experienced."
>
>If we substitute von Humboldt's term, "world-view," with my proposed term,
>"thinking for speaking," we have here a powerful statement about the role
>of language in what Sapir called those "contents of experience [that are]
>capable of expression in linguistic terms."
>
>CONCLUSION
>
>	In sum, we can only talk and understand one another in terms of a
>particular language.  The language or languages that we learn in childhood
>are not neutral coding systems of an objective reality.  Rather, each one
>is a subjective orientation to the world of human experience, and this
>orientation *affects the ways in which we think while we are speaking*.
>