ALLOPHONIC VARIATION IN SPONTANEOUS MONOLOGUES (AN EXPERIMENTAL STUDY BASED ON GERMANIC LANGUAGES)

The current paper focuses on the dependence of allophonic variation on the informational structure of an utterance in German and English languages. The experiment includes two parts: the German program and the English program. The German program observes acoustic characteristics of intervocalic plosive consonants /d/ and /g/ in theme and rheme parts of German spontaneous monologues. The English program studies mean duration and F1, F2 frequencies of stressed monophthongs in theme and rheme parts of British political speeches. The subjects are male German and English speakers without any notable features of belonging to a particular dialect within Germany and the UK. The following patterns were found. The German program: Both intervocalic plosive consonants /d/ and /g/ become spirantized when occur in theme parts of the speech - they lose their closure and change acoustic articulatory characteristics as affected by neighboring vowels. Along with spirantization consonant /g/ comes out to be nasalized and sonantized within theme parts of the utterance. Plosive consonants /d/ and /g/ segmented from rheme parts of the speech have some traces of intervocalic context, but save their complete closure. The attained results show the weakening of consonants in theme parts of the speech in comparison with those in rheme parts. The English program: The majority of the monophthongs increase their duration within the rheme. At that, the slightest difference in quantity is found among front and central vowels and the greatest among the back ones. At the mean, the monophthongs within the rheme turn out to be 17 % longer than within the theme. Mean duration of the monophthongs [e] (S3), [I] (S1, S2), [A] (S2) within the rheme and the theme parts concurs. These results can be explained by the position of the words in the utterance and the phonetic context of the mentioned vowels. The change of vowel duration in relation to their position in the informative structure of the utterance correlates with their formants shifts. According to the spectrographic data mean F1 frequency of [I] (S1, S2), [i.], [e] (S1, S2), [ae], [A] in rheme parts is higher than F1 of those in theme parts. F1 frequency of monophthongs [Lambda], [alpha:] (S1, S2), [(sic):] (S2, s3), [U] (S2, S3), [u.] (S1-S3) and [epsilon:] (S1, S2) in rheme parts is lower than F1 of those in theme parts. The analysis of F2 frequency shows that vowels [I], [i.], [e], [epsilon] in rheme parts are more front, and vowels [Lambda] (S1, S2), [alpha:], [A] (S1, S2), [(sic):], [U] (S2, S3), [u.] (S2, D3) are more back. The data obtained shows that rheme parts play an important role in phoneme features maintenance.