Academia.eduAcademia.edu

Figure 7 – uploaded by Daniel Recasens

See full PDFdownloadDownload figure
To summarize, a trend for initial /l/ to exhibit a higher F2 than final /I/ appears to be at work for moderately dark and moderately clear varieties of /V but not for strongly dark /I/. However, not all dialects with a strongly dark variety of /l/ conform Table 7: F2 frequency difference between initial and final /I/ in several language and dialects. See Figure 4 and body of the text for the language abbreviations.

Table 7 To summarize, a trend for initial /l/ to exhibit a higher F2 than final /I/ appears to be at work for moderately dark and moderately clear varieties of /V but not for strongly dark /I/. However, not all dialects with a strongly dark variety of /l/ conform Table 7: F2 frequency difference between initial and final /I/ in several language and dialects. See Figure 4 and body of the text for the language abbreviations.

Related Figures (7)

Languages and dialects have been classified into groups A through G as follows (not all the literature references from which the articulatory data have been taken are mentioned below): Table 1: Number of items subjected to analysis classified according to consonant and language group. See text for details. Cont= linguopalatal contact data; config= sagittal tongue configuration data. oral stop /c/ (also /j/), and the nasal stop /n/. This literature survey includes linguopalatal contac data obtained by means of static palatography and electropalatography (EPG), as well as sagitta lingual configuration data obtained by means of cineradiography (X-ray) and magnetic resonance imaging (MRI). Overall, data for 238 consonan realizations produced by speakers of differen languages and dialects have been analyzed (see Table 1). The target consonants occur mostly in a low vowel context, and have phonemic or a well- defined allophonic status in the languages under investigation.
Languages and dialects have been classified into groups A through G as follows (not all the literature references from which the articulatory data have been taken are mentioned below): Table 1: Number of items subjected to analysis classified according to consonant and language group. See text for details. Cont= linguopalatal contact data; config= sagittal tongue configuration data. oral stop /c/ (also /j/), and the nasal stop /n/. This literature survey includes linguopalatal contac data obtained by means of static palatography and electropalatography (EPG), as well as sagitta lingual configuration data obtained by means of cineradiography (X-ray) and magnetic resonance imaging (MRI). Overall, data for 238 consonan realizations produced by speakers of differen languages and dialects have been analyzed (see Table 1). The target consonants occur mostly in a low vowel context, and have phonemic or a well- defined allophonic status in the languages under investigation.
Figure 1 presents cross-language tongue contact percentages at the four articulatory zones for the consonants /A, c, p/ produced with a complete or central closure (top graph), and for the fricatives /¢, ¢/ exhibiting a central constriction (bottom graph). In addition to the presence of less tongue contact for fricatives than for laterals and stops, the figure shows differences in constriction fronting varying in the progression /A/ > /p/ > /c/ (closure occurs at Figure 1: Cross-linguistic linguopalatal contact percentages for laterals and stops (top graph) and for fricatives (bottom graph). The analysis criteria of the linguopalatal contact data are described next. Regarding the static palatography data, linguopalatal contact percentages were calculated at four equidistant horizontal lines traced across the palate contact surface at the alveolar, postalveolo-prepalatal, mediopalatal and postpalatal zones. As for the electropalatographic data, contact percentages were calculated for the rows of electrodes 2, 4, 6 and 8 of the artificial palate which are located in the same four articulatory zones mentioned above. Tongue contact was taken to occur at those electrodes exhibiting more than 80% of activation across tokens.
Figure 1 presents cross-language tongue contact percentages at the four articulatory zones for the consonants /A, c, p/ produced with a complete or central closure (top graph), and for the fricatives /¢, ¢/ exhibiting a central constriction (bottom graph). In addition to the presence of less tongue contact for fricatives than for laterals and stops, the figure shows differences in constriction fronting varying in the progression /A/ > /p/ > /c/ (closure occurs at Figure 1: Cross-linguistic linguopalatal contact percentages for laterals and stops (top graph) and for fricatives (bottom graph). The analysis criteria of the linguopalatal contact data are described next. Regarding the static palatography data, linguopalatal contact percentages were calculated at four equidistant horizontal lines traced across the palate contact surface at the alveolar, postalveolo-prepalatal, mediopalatal and postpalatal zones. As for the electropalatographic data, contact percentages were calculated for the rows of electrodes 2, 4, 6 and 8 of the artificial palate which are located in the same four articulatory zones mentioned above. Tongue contact was taken to occur at those electrodes exhibiting more than 80% of activation across tokens.
Figure 3: Linguopalatal contact percentages for /c/ (top) and /p/ (bottom) as a function of the language groups A, B and D. See text for details.
Figure 3: Linguopalatal contact percentages for /c/ (top) and /p/ (bottom) as a function of the language groups A, B and D. See text for details.
In order to investigate the allophonic patterns for these /I/ types, data for word/utterance initial and word/utterance final /I/ next to /i/ taken from the same literature sources referred to above will also be reported. An initial prediction is that strongly dark /I/ ought to exhibit similar strongly dark realizations initially and finally since it is highly constrained and therefore ought not to be affected much by context or position. This hypothesis is consistent with linguopalatal contact data showing no clear articulatory differences as a function of syllable position for other highly Figure 5: F2 frequency for /I/ in the sequence /ili/ in several language and dialects. Sp: Spanish; Fr: French; Va: Valencian Catalan; ECa: Eastern Catalan; Ne: Newcastle English; Le: Leeds English; Ma: Majorcan Catalan; Am: American English.
In order to investigate the allophonic patterns for these /I/ types, data for word/utterance initial and word/utterance final /I/ next to /i/ taken from the same literature sources referred to above will also be reported. An initial prediction is that strongly dark /I/ ought to exhibit similar strongly dark realizations initially and finally since it is highly constrained and therefore ought not to be affected much by context or position. This hypothesis is consistent with linguopalatal contact data showing no clear articulatory differences as a function of syllable position for other highly Figure 5: F2 frequency for /I/ in the sequence /ili/ in several language and dialects. Sp: Spanish; Fr: French; Va: Valencian Catalan; ECa: Eastern Catalan; Ne: Newcastle English; Le: Leeds English; Ma: Majorcan Catalan; Am: American English.
Figure 6: F2 frequency for initial /I/ (black bars) and final /l/ (grey bars) in several language and dialects. See Figure 4 and body of the text for the language abbreviations.
Figure 6: F2 frequency for initial /I/ (black bars) and final /l/ (grey bars) in several language and dialects. See Figure 4 and body of the text for the language abbreviations.
Related topics:
PsychologyComputer SciencePhoneticsSpeech Production
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved