Rit (Vísindafélag Íslendinga) - 01.06.1970, Page 515

5i3
feature. But, tenseness being relevant in the phonemic system
of German, the Polish speaker will subconsciously try to
produce a phonetic substitution for it. Thus the German
phoneme /e/, which the German speaker will render as a
tense [e:], will in the pronunciation of a Pole untrained in
phonetics automatically be turned into the diphthong [ei], as
in ^fdsiben] instead of [deiban]. As the quantity is irrelevant in
the German vowel system, a long vowel could be substituted
for a tense one, as is done in the case of /a/ vs. /a/ (Szulc 1966:
425-426). Unfortunately, the Polish system has no long vowels.
That is why, instead, a diphthong containing the two Polish
sounds nearest to Germ. [e:] is substituted, to compensate for
the length. The product of this transformation, i.e., the diph-
thong [eí], though phonetically totally diíferent from the expec-
ted (i.e., normative) allophone, can still serve to realize the
German phoneme /e/, as it does not collide with any other
phoneme of the German vowel system. The pronunciation
^['lEÍbEn] for [de^bon] is what the Germans would call
normwidrig', nevertheless, it does not disarrange the phonemic
system of their language, as it makes communication possible.
The subphonemic interference can thus be expressed by
the following formula:
Yx (A) -»■ Yix (A) / (B)
which reads as follows: The allophone y of the phoneme x
in the sound system of A goes over to the allophone yx of the
same sound system under the impact of the sound system of B.
From what has been said we can draw the conclusion that
interference phenomena in the phonemic (sub)system can be
strictly predicted, as the speaker cannot alter its restrictedness,
except subphonemically. Nevertheless, this system allows him
a moderately free choice of allophones to realize it.
Out of the remaining subsystems of the signal plane, viz.,
the accentual, rhythmical, and intonational ones, it is the
intonation that offers the speaker the greatest possibility of
choice. That is why it is subject to far-reaching interference,
Proceedings — 33