Tesis doctoral en la redInfoling 1.18 (2021)
This dissertation revolves around three research questions with three goals: 1) examine the lenition patterns and allophonic variation of /p t k/ in an eastern Andalusian community; 2) analyze the degree of phonetic overlap between /p t k/ and /b d ɡ/; and 3) investigate the role played by closure duration and degree of constriction on the perception of voiced and voiceless stops.
In order to address the first research question, twelve native speakers completed a semi-directed interview, from which 2,400 tokens of /p t k/ were extracted. Four acoustic correlates were analyzed for each token: closure duration, VOT, IntRatio, and degree of voicing. The results revealed a significant amount of allophonic variation in the production of /p t k/: 13% of the tokens were phonetically voiced, 42% were partially voiced, and 45% were fully voiceless. The results also suggest that /p t k/ lenition is strongly conditioned by place of articulation, phonetic context, stress, and speech rate. Specifically, /p t k/ are more lenited in intervocalic position, in unstressed syllables, and at higher speech rates. Additionally, with the exception of the analysis of VOT, the results revealed that /k/ tends to undergo more lenition than /p/ and /t/.
In order to address the second research question, an acoustic comparison of /p t k/ and /b d ɡ/ was carried out. The results suggest that, in general, /p t k/ are longer and more constricted than /b d ɡ/. However, a significant degree of phonetic overlap was found in velars since no significant differences were found regarding the duration of phonetically voiced /k/ and /ɡ/.
Finally, a perceptual experiment was conducted to address the third research question. A native speaker recorded stimuli containing intervocalic /b d ɡ/, whose duration was digitally increased, and their intensity was reduced. A perceptual identification task that included both the original and the manipulated stimuli was completed by 97 participants. The results of the perceptual task suggest a significant incipient effect of segmental duration and intensity on the perception of voiced and voiceless stops. The implications of these findings are further discussed in Chapter 6.
CHAPTER 1: INTRODUCTION
1.1. Consonant lenition in the target variety
1.2. Diachronic variation of stop consonants in Spanish
1.3. Stop consonants in modern Spanish and theoretical motivation and contribution of this dissertation
CHAPTER 2: STOP CONSONANTS IN SPANISH: CLASSIFICATION AND PHONETIC PROPERTIES
2.1. Phonetic properties of Spanish stops
2.1.1. Activity of the vocal folds
2.1.2. Manner of articulation
2.1.3. Place of articulation
2.2. Allophonic variation of Spanish stops
2.2.1. Allophonic variation of voiceless /p t k/
2.2.2. Allophonic variation of voiced /b d ɡ/
2.3. The overlap of lenited /p t k/ and /b d ɡ/
CHAPTER 3: RESEARCH QUESTIONS AND HYPOTHESES
CHAPTER 4: LENITION OF /p t k/ AND OVERLAP OF CATEGORIES IN SPEECH PRODUCTION
4.3. Lenition of /p t k/: Analysis
4.3.1. Acoustic correlates (dependent variables)
4.3.2. Linguistic factors (independent variables)
4.3.3. Descriptive statistics of dependent variables and selection of statistical analyses
4.4. Lenition of /p t k/: Results
4.4.1. Descriptive statistics of IVs
4.4.3. Closure duration
220.127.116.11. Descriptive results for voicing (all data)
18.104.22.168. Voicing results from ordinal regression (all data)
22.214.171.124. Voicing results from linear regression (partially voiced data only)
4.4.6. Summary of main findings
4.5. Comparison of /p t k/ and /b d ɡ/: Analysis
4.6. Comparison of /p t k/ and /b d ɡ/: Results
CHAPTER 5: PERCEPTUAL ANALYSIS
CHAPTER 6: DISCUSSION AND CONCLUDING REMARKS
6.1. Research question 1: What are the patterns of /p t k/ lenition in eastern Andalusian Spanish? And what linguistic factors contribute to such lenition?
6.1.1. Articulatory and physiological considerations and how they relate to place of articulation
6.1.2. Additional articulatory considerations and how they relate phonetic context, stress, and speech rate
6.1.3. The role of word position and its implications for sound change
6.2. Research question 2: When realizations of /p t k/ are lenited and voiced, are there acoustic differences with respect to /b d ɡ/?
6.3. Research question 3: To what extent do native speakers of Andalusian Spanish rely on duration and degree of constriction to perceptually categorize /p t k/ and /b d ɡ/ when the stimulus is phonetically voiced?
6.4. Concluding remarks and future directions
Institución: The University of Texas at Arlington (EE.UU)