Orð og tunga - 01.06.2005, Síða 54

52
Orð og tunga
late a rule that allows for secondary or alternative forms for limited
variations in spelling (i.e. the presence or absence of accents in a word
like múslimi ('Muslim') as opposed to a more radical form of variation
such as píramíði as opposed to pýramídi, etc. ('pyramid')); such sec-
ondary forms are, however, usually ignored in bilingual dictionaries,
either from a desire to be normative, or out of a desire not to confuse
the L2 user.
3.1.3 Specialised vocabulary for translators
Specialised vocabulary is particularly needed by translators of "theme-
specific" texts or for that matter by translators of fiction, where there
may suddenly or consistently be archaic or specialised vocabulary to
be dealt with. I must state, however, since there was no marking of
usage areas or semantic fields such as 'maritime', 'technical', 'archaic',
either in the printed form of Iðunn 1989 or in its underlying data, there
is no effective way of testing this dictionary's performance in semantic
areas such as media and communications, trade, and culture - areas
in which I might expect it to be weakest, because these are areas in
which a great deal has occurred socially and linguistically during the
last fifteen years.
The future of lexicography for professional translators, as indeed
for other advanced users, surely lies to a large extent with web diction-
aries - preferably connected to text corpora which have user-friendly
corpus-processing tools (cf. below, section 4).
3.2 Plans for revision of Iðunn 1989: conclusion
For all of the points I have just mentioned, new active simplex words
and the whole range of compounds with potential variants, the final
decision on what to include will ideally be largely dependent on a re-
liable and analysable frequency list for Modern Icelandic which still
does not exist, preferably one in which the texts on which it is based
are themselves electronically available or at least identifiable. This fre-
quency list should ideally be made up of a distinguishable variety of
written texts, distinguishable not only in terms of the text type, but
also chronologically (say, over a period from 1900 to the present day),
supported by a major variety of spoken texts (radio tapes are obvi-
ously important in this respect but cannot replace projects such as
ISTAL (see Ásta Svavarsdóttir 2003)).