psychology is known as transfer. He transfers the habit systems of
his native language to the foreign tongue . . . .
When this transfer occurs, some of the units and patterns trans
ferred will function satisfactorily in the foreign language and will
not constitute a learning problem. Other units and patterns will
not function satisfactorily in the foreign language. Against these
the student will have to learn the new units and patterns. These
constitute the real learning problems.
These learning problems turn out to be matters of form, mean
ing, distribution, or a combination of these. They can be pre
dicted and described in most cases by a systematic linguistic
comparison of the two language structures . . . .
The theory assumes that testing control of the problems is test
ing control of the language. Problems are those units and patterns
which do not have a counterpart in the native language or that
have counterparts with structurally different distribution or
meanmg.
[>
In what terms does Lado describe knowledge of language?
Give some examples of the kinds of knowledge he means.
[>
In situations where the test population is drawn from learners
of diverse linguistic background, what problems would arise
in practice if you based the design of language tests on con
trasts between the language being tested and the language of
the test takers? How might one get around this difficulty?
Text 3
B E R N A R D s P o L s KY :
'Introduction: Linguists and language
testers' in B. Spolsky (ed.):
Approaches to Language Testing.
[Advances in Language Testing Series:
2] Center for Applied
Linguistics r978, pages v-vi
Spolsky distinguishes three historical periods of modern lang
uage testing up to the time of his writing: the pre-scientific,
the psychometric-structuralist, and the integrative-socioling
uistic. He discusses the first two of these in this extract.
The pre-scientific period (or trend, for it still holds sway in many
Chia sẻ với bạn bè của bạn: