Íslenskt mál og almenn málfræði - 13.07.1981, Qupperneq 44

40
M. P. Bames
article on fyri, on the other hand, is almost totally free from interference
by the lexicographer. It is as though the dictionary slips had been tossed
into three heaps, marked dative, accusative and adverbial, and entered
more or less in the order they lay. The only indication of an awareness
of finer distinctions is the occasional: ved verber, ved (nogle) adjektiver,
which blur rather than clarify important rules. It is a mark of the weak-
ness of this article that a better understanding of the use of fyri can be
obtained by consulting the article on the Icelandic preposition fyrir in
Blöndal (1920-4). There are of course differences between Faroese and
Icelandic usage, and Blöndal’s categorisation contains certain discrep-
ancies and contradictions, but his is by far the more useful and system-
atic of the two accounts. It is, however, clearly unsatisfactory in the
long run to have to approach one language through another, and the
aim of the present study will be to provide a more systematic description
of the semantics and morphological case government of the Faroese
preposition fyri than has hitherto been given. This description should
benefit the persistent leamer, and I trust it may also advance in a
modest way our understanding of a small part of Faroese grammar.
Since I am not a native speaker of Faroese, my study has been
mainly corpus-based. I have no wish to enter the controversy over
corpus-based vs. intuitive linguistics, and have indeed on numerous
occasions tapped the intuitive resources of native speakers, but for my
present purpose working from a corpus does seem to have certain
advantages. The examples of usage stem from a wide cross-section of
people using their native language in a variety of situations and styles,
completely oblivious to the linguistic problems they might be causing or
solving. It is language in the raw, warts and all. Specially selected sen-
tences, grammatical and non-grammatical, chosen to prove or disprove
a particular point, are useful when dealing with a language for which
much data has already been made available and on which many studies
have been done, but in the case of Faroese, as we have seen, even the
basic data is mostly lacking. A corpus-based study of performance is
clearly the pre-requisite for theories about competence, if in fact the
two can properly be distinguished. Speakers of Faroese in my experi-
ence display a high degree of uncertainty about what is and what is not
permissible in their language, and I have on occasion found it consider-
ably easier to make sense of their performance than their linguistic
intuition. No doubt this is at least in part due to the situation of Faroese,