Transcript Slide 1

Department of Language and Linguistics
Language Day Conference for Teachers – 17th September 2011
Why don’t they learn things I’ve taught them? – understanding how
second language knowledge develops in classroom learners
Roger Hawkins
1
You may have come across exchanges like the following of a parent
trying to teach a child first language learner some aspect of the TL
(from de Villiers and de Villiers 1979):
Adult: Say ‘tur’.
Child: Tur.
Adult: Say ‘tle’.
Child: Tle.
Adult: Say ‘turtle’.
Child: Kurka.
The child appears not to learn something the adult is trying teach
her/him
2
A similar observation can be made of classroom L2 learners:
“A teacher has drilled students in the structure known as indirect
questions:
Do you know where my book is?
Do you know what time it is?
Did he tell you what time it is?
As a direct result of the drills, all students in the class were able to
produce the structure correctly in class. After class, a student came up
to the teacher and asked, ‘Do you know where is Mrs Irving?’ In other
words, only minutes after the class, in spontaneous speech, the student
used the structure practised in class incorrectly.”
(From Gass and Selinker 2008, p.15)
3
What is the grammatical property of English the learner has apparently
failed to acquire?
In embedded questions (indirect questions) tense-marked verbs do not
precede subjects:
I wonder where his book[subject] was/*I wonder where was his book
She wonders what Mrs Irving[subject] likes/*She wonders what does Mrs
Irving like?
4
The example illustrates an important characteristic of classroom L2
learners:
They can establish two distinct types of knowledge
explicit (conscious, verbalisable) knowledge
implicit (unconscious, not necessarily verbalisable) knowledge
40 years of empirical research on second language acquisition has
uncovered a number of the characteristics of explicit and implicit
knowledge.
5
- Explicit knowledge can be used in certain kinds of task but not all:
Typically accessible in classroom drills, grammar exercises, written
translation
Explicit knowledge of the form of embedded questions in English is what
allowed all learners in the Gass & Selinker example to perform to criterion
in the class drills
6
- Explicit knowledge is not easily accessible in spontaneous, natural
communicative uses of language (conversation, discourse where the
focus is on meaning (could be written or spoken), non-scripted
monologues)
The implicit knowledge of the learner who spontaneously asked ‘Do
you know where is Mrs Irving?’ in the Gass & Selinker example does
not distinguish main from embedded clauses for question formation.
7
-Implicit knowledge of an L2 develops according to a kind of ‘internal
mental timetable’.
Certain properties emerge in a learner’s mental grammar before
others
This is true not just of ‘immersion’ learners, but also classroom learners
who have traditional instruction + practice exposure
(examples to follow shortly)
- Implicit knowledge often appears to make ‘mistakes’; learners
produce things they have not heard in the input they get
Crucially, these mistakes are often evidence of acquisition, and not
failure to acquire
8
Examples of a timetable for the development of implicit knowledge of
properties of an L2 in classroom learners:
1. In L2 English, subject-verb agreement on forms of the copula be
emerge before agreement on main verbs.
L1 speaker of Japanese orally retelling an aurally-presented story:
Mr Jones always read the weather forecast in the morning newspaper.
When it forecast raining he bring umbrella with him. … Unfortunately
sometimes weather forecast is not correct. When he arrive at work he …
his clothes are wet …
9
2. In L2 French, implicit knowledge of the gender of articles goes
through a stage where one form of the article is restricted to the correct
set of nouns, while the other is overgeneralised.
A typical pattern is a speaker using la (fem) only with feminine nouns, but
le (masc) with both masculine and feminine nouns:
La balle (f), la chose (f), le carré (m), le mur (m), *le balle (f), *le chose (f)
Careful counting of the use of articles in contexts where implicit
knowledge is accessed shows that use is systematic, although it may
appear superficially random
3. Development of implicit knowledge of word order in German by
classroom L2 learners is in stages.
10
In German main clauses, non-finite parts of the verb (participles,
particles) always come at the end:
Johann hat ein Buch gekauft
Johann has a book bought
Linguists refer to this as ‘verb separation’
In main clauses the finite (tensed) part of the verb always appears in
second position:
Heute hat Johann ein Buch gekauft
Today has Johann a book bought
Linguists refer to this as ‘verb second’
11
In subordinate clauses, both parts of the verb appear at the end of the
clause:
Er glaubt, dass Johann ein Buch gekauft hat
He believes that Johann a book bought has
Linguists refer to this as ‘verb final’
In terms of implicit knowledge, L2 learners acquire these properties in
the order:
verb separation → verb second → verb final
12
However, in classrooms word order is often presented to learners in the
order:
verb second → verb separation → verb final
Ellis (1989) – 39 English-speaking adult ab initio German learners –
teaching of grammar + communicative activities
Despite teaching in the order
verb second → verb separation → verb final
on a picture description task, accuracy of word order was:
verb separation → verb second → verb final
13
Let me now give you an example of how implicit knowledge in L2
learners can give rise to ‘mistakes’ that in fact indicate that acquisition
is going on
Repeated observation: L2 speakers of English who know the passive
construction produce and accept sentences like (1a-c) (examples from Zobl,
1989: 204 and Yip, 1995: 130)
1a. My mother was died when I was just a baby
b. The most memorable experience of my life was happened 15 years ago
c. Rush hour traffic can be vanished because working at home is a new
version
14
Here passive morphology is supplied with verbs which for native
speakers disallow it: intransitive unaccusative verbs like die, happen and
vanish:
2a. My mother died when I was just a baby
b. The most memorable experience of my life happened 15 years ago
c. Rush hour traffic can vanish because working at home is a new
version
Passive morphology is used significantly more often with:
unaccusatives: die, happen, disappear, break, sink, melt
than:
unergatives: cry, shout, cough, laugh, run, walk
15
Intransitive verbs fall into two classes:
(i) unergatives: the subject has similar A man shouted (shout – unergative)
semantic properties to the subject of The diamond sparkled (sparkle – unergative)
transitive verbs: typically the subject
causes or gives rise to the event
described by the verb.
(ii) unaccusatives: the subject has
A man appeared (occur – unaccusative)
similar semantic properties to the
The traffic vanished (vanish – unaccusative)
object of transitive verbs: typically the
subject is affected by the event
described by the verb.
16
Because the single argument with unaccusative verbs has ‘object
properties’, a number of linguists have argued that at an abstract level
of representation the single argument IS an object, in contrast to
unergatives:
3a
b
e happened this experience 15 years ago (unaccusative)
The student shouted loudly (unergative)
The object of the unaccusative must move to the empty subject
position.
Notice that the passive in English involves the movement of an object to
a subject position:
17
4a
Everyone enjoyed the experience
b
e was enjoyed the experience (by everyone)
c
The experience was enjoyed (by everyone)
We now see that the L2 learners are simply extending the passive
construction in English to the case of arguments in the object
position of unaccusative intransitives.
They do not do this with unergatives because unergatives do not have
objects.
18
This ‘misuse’ of the passive in the implicit mental grammars of L2
learners shows:
(a) That they have acquired the English passive rule (although have not
yet learned that it is restricted to transitive verbs)
(b) That they know that the abstract representations of unaccusative and
intransitive verbs are different
19
Returning to our question: “Why don’t L2 learners acquire the things
I’ve taught them?”
In relation to explicit knowledge they may well learn what you
teach them:
(‘tense-marked verbs do not precede subjects in embedded
questions’, ‘chose in French is la and not le’, ‘tensed verbs always
come second in German main clauses’, etc)
And they can use this knowledge in some circumstances.
20
In natural communicative tasks, however, learners access implicit
knowledge, and this is apparently not directly affected by teaching.
Implicit knowledge develops according to its own timetable (S-V
agreement on copula be before agreement on main verbs (English);
over-generalisation of one gender-marked article and undergeneralisation of the other (French); verb separation → verb second →
verb final (German))
Learners will show patterns of use that may well be non-target-like
(over-generalisation of the passive to unaccusative intransitive verbs
(whose single argument has ‘object’ properties))
These patterns are common to all learners (classroom or immersion)
These non-target-like patterns are evidence of acquisition (not failure to
learn)
21
There is a debate about whether explicit knowledge might ‘feed’ implicit
knowledge indirectly, after a delay
There is some evidence that the implicit knowledge of tutored learners
gets to more target-like levels than non-tutored learners
Explicit knowledge might have the effect of highlighting the kind of
properties of the input that the implicit language acquisition device can
make use of.
So don’t give up on teaching learners about the language just yet!
22