An Introduction to Applied Linguistics

3 PROTOT YPE THEORY
On the face of it, these two extreme views, the Lee and the Banville, of what it means
to be a native speaker, support the prototype theory which the psychologist Eleanor
Rosch developed. Escudero and Sharwood Smith (2001) consider that her theory has
relevance for second language acquisition (SLA); they suggest it is ‘the best way of
introducing the necessary precision’ to the concept of native speaker. ‘We argue,’
they write:
that reformulating the concept along the lines first suggested by Eleanor Rosch,
that is using prototype theory, should provide the best way of introducing the
necessary precision. This has the consequence that native-speakerhood becomes a
gradient term with, respectively, core (prototypical) and peripheral features. This
allows researchers to be more precise about what they, or the particular theoretical
approach they adopt, claim to be the essential and non-essential feautures of
nativeness and the necessary rigour is thereby achieved.
(Escudero and Sharwood Smith 2001: 275)
Prototype theory, then, is an approach to categorisation. Instead of saying of a
phenomenon: ‘Yes, this is an instance of category X’ or ‘No, this is not an instance’
we can say: ‘This instance does satisfy criteria a, b and c for membership but it does
not satisfy criteria d and e.’ The canonical example for the theory is, apparently, the
The applied linguistics challenge 151
02 pages 001-202:Layout 1 31/5/07 09:31 Page 151

---

category ‘bird’. So when shown an instance of something bird-like, you run through
the criteria for ‘bird’ and you say: ‘it’s small, has feathers, two legs, two wings and a
beak/bill, lays eggs.’ But if the instance in view cannot fly (for example a barnyard
fowl) then you compute: ‘less bird-like’ than, say, a sparrow, while an ostrich is even
less bird-like because it is not small and it cannot fly. This sounds like a neat idea.
But hold on! While it may work for birds, will it work for native speakers? After all,
Escudero and Sharwood-Smith have themselves recognised the problem they face in
achieving ‘the best way of introducing the necessary precision’. It is, as they say, that
‘precise definition of the term “native speaker” is extremely difficult and therefore
usually avoided even though the concept is vital in SLA as in many other domains
dealing with language ability’. Most researchers they continue, ‘rely on the
assumption that there is a common understanding of what a native speaker is’ (2001:
275). Precisely, there is their problem. The assumption Escudero and Sharwood
Smith seem to be making is that we already know what is meant by core features of
native speaker. But we don’t. No doubt what they are proposing is reaching towards
some consensus as to what the criteria for the core might be. But native speakers are
not birds – we could offer the headline: 
Words Not Birds
. The term ‘native speaker’ is
used in different ways which may in part explain why definition is difficult.