Corpora: re: ignorance among the media

Jock McNaught (jock@ccl.umist.ac.uk)
Fri, 7 Nov 1997 12:24:47 +0000 (GMT)

Hmm, the responses on this one so far indicate that I should be more
charitable in my interpretation of 'so-called corpus' and should look upon
'so-called' as a means of introducing a technical term the reader might be
unfamiliar with. I am familiar with that usage of 'so-called', however in
the context of this particular article I think it an unlikely reading,
given the generally somewhat aggressive tenor of the article.

It was not clear from the responses whether colleagues had in fact read
the article or were simply pointing out there was an alternative
interpretation.

It does not appear to be normal Guardian style to refer factually to terms
through the use of 'so-called' although, ironically, a collection of Guardian
texts would be needed to verify my impression!

A rapid trawl through the British National Corpus reveals that many cases
of 'so-called' are used in a disparaging fashion (as in 'the so-called
accident'). Very few appear to be used to introduce a technical term with
no connotation attached. The other major usage is to introduce what may be
a technical term but is one that the writer patently does not agree with,
implying he does not really recognise this as a term, considers it a jargon
word as opposed to a term, considers it an infelicitous coinage, wishes to
distance himself from the group who use that word as a term, etc. Notice
that this can be the case even where one might think there is indeed a
terminological reading to be obtained. Consider:

'Leucanthemella serotina is what I was brought up to call Chrysanthemum
ulginosum, the so-called moon daisy of Hungary.'

This writer (writing in a 'Gardener's World' publication) would appear to
deprecate the use of this particular popular name - other popular names in
the same text are not given this treatment.

I did not look at uses where 'so-called' came after the focussed item (as
in 'the Great Arch, so-called because...') where some explanatory function
is involved.

I stress this was a rapid trawl that did not involve deep analysis thus
must remain an impression of contemporary usage. However, I got some
personal satisfaction from using a corpus to investigate the probability
of an anti-corpus tirade having been detected.

I guess I am not convinced yet by the 'reasonable doubt' argument given
that so far I appear to be the only one to have seen the evidence of the
article itself.

However, no point in prolonging this discussion. Nevertheless, if colleagues at
Birmingham and Collins got the same reading as me, that would be interesting
to learn about. Now, where did I lay my copy of the so-called The Guardian?

Jock

-- 

John McNaught jock@ccl.umist.ac.uk Centre for Computational Linguistics UMIST PO Box 88 Sackville Street Manchester, UK tel: +44.161.200.3098 (direct) M60 1QD fax: +44.161.200.3099