Orð og tunga - 01.06.2001, Blaðsíða 66

56
Orð og tunga
a reasonable representation of the contents of nineteenth-century scholarly dic-
tionaries (Cleasby/Vigfusson and Fritzner) while incorporating the new material
introduced in the supplementary volume to one of them (Hpdnebp 1972).
c) The list thus arrived at, consisting of the compound words in ONP’s archives and
the vocabulary of NO, could now be augmented by the total vocabulary of ONP
itself through comparison with the dictionary slips. This was the task of Jóhannes.
Under considerable pressure of time hasty assessments had to be made. Apart
from the appreciable number of additions to the word-list in the form of lexical
items that had not been previously registered, it was necessary to check: i) that
the tokens were correctly lemmatized (organised under the correct headword); ii)
that the headwords were spelt in accordance with ONP’s standard normalisation
(see §6 below); iii) that the grammatical details were an accurate representation
of the usages recorded on the slips (wordclass and gender). Furthermore a con-
siderable number of selections had to be made. This is not the place to give a
full presentation of the limits governing ONP’s corpus, but a short explanation
is appropriate (see §5 below).7 During the process a number of what had been
registered as single words were divided up into two or more headwords, and some
items that had previously been separate were combined into one (see §5 below).
Consultations sometimes had to be carried out with the regular editorial staff,
and when everything was in place the number of handwritten slips designated as
representatives of each headword was counted by the secretarial assistants, and
this information recorded in the database file (for the statistical results, see §7
below).
5 Corpus considerations
ONP does not treat place-names or personal names (except cognomina/nicknames).
Poetic vocabulary and unassimilated foreign words (as distinct from assimilated, or
partly-assimilated, loanwords) are registered in the printed volumes but without analysis
or explanation. These delimitings of what areas of vocabulary ONP is to cover have been
the subject of much consideration throughout the history of the project and were finally
crystallized when the index volume, Registre II Indices, was produced in 1989. Factors
of time and money were significant, but important too was an awareness that ONP’s
excerpting in some of these debated areas was insufficient. Quite a number of slips have
been written in the course of time containing onomastic material and foreign words, yet
the excerpting was too irregular and inconsistent for a scholarly analysis. From the point
of view of planning it is, however, necessary to know how many slips of this type there
are in the ONP archives. Consequently there is a detailed system in ONP’s database
procedure for tagging the various types of deselected or “marginal” material. This had
to be executed in full detail during the third phase (c) of the preparation of the word-list.
7See also Nfigle // Key, p. 30.