Rit (Vísindafélag Íslendinga) - 01.06.1970, Page 303

301
ciples of phonological explanation were determined in large
part by study of the Indo-European languages, so those for
syntax should now be. Accordingly, further syntactic analysis
of the Nordic languages would be highly important. Among
other things, we would like to determine as many features as
possible which we would expect in SOV languages, and their
status in the syntax of the Nordic languages. For example,
we would also expect to find in SOV languages relative clauses
preceding their antecedents. This is an infrequent pattern in
Old Norse, attested in a few passages of verse (Nygaard 1905:
271). We would also expect to find suffixation in SOV lan-
guages. For this feature we may cite the reflexive or medio-
passive, the postposed articles, the suffixed personal pronouns,
and the suffixed negatives. Again we might conclude from the
short existence of the last two devices that suffixation is not
an integral feature of the Nordic languages; the negative, for
example, is attested primarily in verse, in legal texts, and the
like.
But if we assume that the proposed syntactic framework is
valid, and also the characterization of the Nordic languages
as SVO, how are we to account for the unexpected syntactic
features? In an excellent paper which has not yet been pub-
lished, entitled ‘Is Amharic an SOV Language?,’ Emmon
Bach has examined syntactic features in Amharic, which as
his title indicates has clause order SOV. Yet, it also has syn-
tactic features characteristic of SVO (or VSO) languages,
such as its gapping rules, prepositions rather than postposi-
tions, and others. Bach concludes that in deep syntactic struc-
ture Amharic is really an SVO language, and he proposes
that its SOV order belongs to the surface structure, its gap-
ping and other aberrant features to the deep structure, which
like that of other Semitic languages has the verb precede the
object. Moreover, he recalls in accounting for the discrepancy
between the two syntactic levels the long-established view of
substrata influences on Amharic from indigenous languages
of Ethiopia and suggests that the aberrant surface structure