Orð og tunga - 01.06.2001, Síða 65

Jóhannes Sigtryggsson & Chris. Sanders: The Copenhagen Old Norse Word-list 55
word-file to extract the headwords that have been selected for editing and publication.
This consultation of the word-file is the very start of the process, and, in order to avoid
duplication of data, this word-file has been transmuted into the word-list that is being
described here. The word-file also contains all of the information that generates the
section at the end of a standard dictionary entry introduced by the abbreviation Gloss.
for “glossaries”,5 but the preparation of that material is not part of the present project.
4 How the word-list was produced
ONP’s raw material is the handwritten slips that have amassed over the years. These
contain the citations on which the grammatical (syntactic) and semantic analysis is
based, and each has in its top right-hand corner the headword that was deemed ap-
propriate for each token at the time that it was excerpted. Until now, no detailed
count had been made of these slips, only approximate estimates of their quantity.
There was an additional complication in that in the course of time two collections
had developed: one that represented the corpus as it was at the beginning of the
sixties (‘gammel samling’ // old collection), and another containing the slips that had
subsequently accrued (‘ny samling’ // new collection). These two collections had to be
combined before compilation of the word-list could start, and this was the first work-
phase (a). Secondly (b), a primitive list of words was prepared on the basis of ONP’s
existing database file of compound words and the dictionary Norr0n ordbok (1975)
(subsequently NO); thirdly (c), this initial list was compared with ONP’s collection of
dictionary slips. This final phase was the cornerstone of the project, since it was here
that significant additions were made and earlier inaccuracies put to rights.
a) In the process of combining the old and new dictionary collections, improvements
were made to the physical organisation of the slips. With the help of differently
coloured dividers, inserted between the filecards, it is now possible to quickly
identify different types of slips and different types of vocabulary. This is invalua-
ble for those who subsequently key in the slips and for dictionary editors and
visiting researchers who need to consult not only the computer files but also the
physical archives. It also greatly facilitated the third phase (c) of producing the
word-list, where quick identification of the nature of the material that was to be
compared with the inventory of words in NO was an important factor.
b) Before the word-list in its present form was conceived, ONP had already prepared
a database file of the compound words that had been registered in the course of
excerpting. All the compounds in this file were copied to form the germ of the
new word-list. The vocabulary of the dictionary NO was selected as the primary
supplement to the embryonic word-list, so its inventory of headwords was additi-
onally keyed in by the secretarial assistant.6 NO was chosen because it contains
5See pp. 46-47 in Npgle // Key (1995), the explanatory booklet distributed with ONP 1 : a-bam.
6The NO vocabulary for the letters d-h had been keyed in before the start of the project.