Realizations of /x/ in intervocalic contexts in Southern Polish .
Grzegorz Nawrocki* and Wiktor Gonet^
*PWSZ Tarnów and ^UMCS Lublin
Traditional Polish phonetics and phonology views the phoneme /x/ as invariably velar, with a palatalised allophone occurring before the vowel /i/ or glide/j/ (Wi¶niewski). Alternatively, Cracow Regressive Obstruent Voicing Rule postulates the existence of a voiced velar allophone /x/.The same variant is ascribed to speakers from eastern frontiers of Poland as a realization of ‘h’-spelling (Wi¶niewski). This is extensively commented upon by Nitsch, who signals the existence of glottal articulations of the phoneme locally, but seems to dismiss them as doubtful. Some studies show possible voiced velar allophony intervocallicaly (Gonet), as well as the phonological possibility of voiceless glottal articulations in pre-vocalic contexts (Gussmann).
The problem of intervocalic variants of /x/ has often attracted attention, as apparently its phonetic value did not agree with what is proposed in literature. In areas surrounding Tarnów, the sound differs dramatically from standard Polish /x/, sharing common features with glottal sounds in languages like Irish, Scots Gaelic, Arabic, Slovak, or German, where such articulations are phonemic and contrast with velar ones. A comparison of the situation in Polish with that Gaelic and Arabic reveals that the sound often pronounced in the intervocalic positions in the area of Tarnów is clearly not velar, whether voiceless or voiced . This observation constituted motivation to analyse the problem by gathering data and investigating it acoustically and statistically.
The speech samples were acquired from 16 speakers of both genders, aged between 20 and 30, living in and outside Tarnów who were asked to read a list of 27 Polish words with the phoneme /x/ between vowels. The items were arranged in a random order, the informants did not know the purpose of the procedure, and they were asked to read the list in a casual way. The samples were recorded on an audio tape at a small classroom setting, using a dictaphone. Afterwards they were digitalized at 44 kHz and analyzed acoustically with the use of SFS/Wasp 1.0 and Speech Analyzer 1.5. The results were subjected to a 2-way ANOVA .
The results of the acoustic analysis display variability in pronunciations ranging from purely voiceless velar to fully or partially voiced glottal. Acoustic spectra of the articulations show close affinity with typical acoustic properties of both sounds found in literature. A two-way ANOVA ('Gender ' by 'Place of articulation' ) shows a statistically significant difference for places of articulation (FA=37.69, F(2;42)=5.18 at p < 0.01 ), while the difference for gender is non-significant (FB=0, F(2;42)=5.18 at p < 0.01 ). The interaction of place of articulation and gender (FAB=9.97, F(2;42)=5.18 at p < 0.01 ) is statistically significant.
The high and statistically significant occurrence of glottal articulations in males seems to indicate positional allophony within the phoneme /x/ . Despite the significance of Gender, the variability appears to be conditioned geographically, as the female informants who come from more eastern regions favoured velar articulations which generally lowered the frequency of occurrence of glottal sound for females. The vocalic context was non –high front for most words and here the articulations were mainly glottal. However, too few words with front high vowels made it impossible to test this factor statistically. What might be involved here is assimilation, or lenition conditioned geographically.
Catford, J.C. (1988) A Practical Introduction to Phonetics. Oxford: Clarendon Press.
Gimson, A.C. (1989) An Introduction to the Pronunciation of English. London: Arnold.
Gonet, W. (2001) Obstruent Voicing in English and Polish.A Pedagogical Perspective. International Journal of English Studies. Vol.1 no1 73-92. University of Murcia.
Gordon, M., Barthmaier,. P., Sands,. K. (2001) A Cross-linguistic Study of Voiceless Fricatives. Santa Barbara: University of California.
Gussman, E. (2002) Phonology. Cambridge: University Press.
Cyran, E. (1996) The Parametric Occurrence of Elements in Phonological Systems. A Festschrift for Edmund Gussmann. 76-98. Lublin: Redakcja Wydawnictw KUL.
Lass, R. (1984) Phonology. Cambridge: CUP.
Nitsch, K. (1994) ¦wiat mowy polskiej. Warszawa: PWN .
Watson, K. (2002)The realization of Final/t/ in Liverpool English. Durham Working Papers In Linguistics.Vol.8 195-205 Durham: The University of Durham
Wells, J., House,.J. (1995) The Sounds of the International Phonetic Alphabet. London: Phonetics and Linguistics,U.C.L.
Wi¶niewski, M. (1998) Zarys fonetyki I fonologii współczesnego języka polskiego. Toruń: Wydawnictwo Uniwersytetu Mikołaj Kopernika.
Wróbel, H. (1995) Fonetyka i fonologia.Kraków: Wydawnictwo Instytutu Języka Polskiego. PAN.