Ambiguous Vowels Research Articles

Phonological and Orthographic Influences on Children's Vowel Spelling

Vowel representations are particularly difficult for children to learn because most vowel phonemes can be spelled in several different ways. Children in Grades 1, 2, and 3 spelled nonwords with an ambiguous vowel and reported their spelling strategies. Analysis of the children's spellings and strategy reports revealed a shift in relying solely on phonological information to considering orthographic information for making vowel letter choices. Implications for vowel spelling development are discussed.

Phonological representation affects phonetic perception.

This study examines to what extent phonological representations affect word identification in Dutch. Dutch has an underlying contrast in voicing that is neutralized word-finally. Also, vowels are lengthened approximately 20 ms when preceding medial voiced consonants. Vowel length can therefore be a cue to voicing. The present study investigates whether the vowel length cue influences listeners when hearing stimuli with ambiguous vowel duration in an identical, neutralized, consonantal context but where the underlying representation of the consonant differs in voicing. A vowel length continuum ([at] to [a:t]) was made by shortening a long vowel in 12 steps. To this continuum, initial consonants were added to create four phonetic endpoints with opposite underlying voicing patterns: /zat/, /zaad/, /stad/, /staat/. Results of a vowel categorization task showed that, using an identical vowel length continuum, the crossover boundary in the zat–zaad continuum occurs at a significantly shorter vowel length than that in the stad–staat continuum. These results provide evidence that listeners use the underlying phonological representation in the perception and identification of words.

The phonological representation of [voice] in speech perception.

This paper examines to what extent phonological representations affect word identification. The contrast under investigation involves the interaction between voicing and vowel length in Dutch. Dutch has underlying contrasts both in obstruent voicing and in vowel length. The voicing contrast is neutralized on the surface in syllable-final position. Also, both long and short vowels are lengthened by some 25 msec when followed by medial voiced obstruents. The present study investigated whether this vowel length cue influenced listeners when hearing stimuli with ambiguous vowel duration in an identical, neutralized consonantal context in which the underlying representation of the obstruent following the vowel differed in voicing. A vowel length continuum ([at] to [a:t]) was created and appended to initial consonants to make two pairs of real words. Each pair differed in vowel length with opposite underlying final consonant representations: /zat/-/za:d/, and /stad/-/sta:t/. Our question was whether the vowel category boundaries would be different in pairs like /zat/-/za:d/ as compared to /stad/-/sta:t/. Although the underlying consonant is either voiced or voiceless, the surface word-final consonant for both pairs of stimuli is always voiceless ([t]). If the listener uses a surface representation with a voiceless consonant to recognize the words, there should be no difference in categorization of the vowel-length continua. The results of a vowel categorization task, however, showed a significant difference in the location of the phoneme boundaries between the two continua, suggesting that listeners' perception seems to be guided by the underlying phonological representation of words.

Whole-word training therapy in a stable surface alexic patient: It works

Abstract An experimental therapy which focuses on whole-word reading was developed for trial with a surface alexic patient whose reading deficit had been stable for several years. The patient practised reading a group of words with ambiguous vowels. A control group of words with the same ambiguous vowels were never studied but were tested at each session. The patient's reading of the studied words improved steadily over the training sessions. Performance on the control words varied considerably throughout the study. It is concluded that impaired access to orthographic entries in the lexicon can be repaired with practice, but that generalization to similar words is more difficult to achieve.

Perception of temporal and spectral information in French vowels.

Perceptual effects of orthogonal variations in temporal and spectral information differentiating French /o/ and /o/ were examined. Although both parameters contribute to acoustic differentiation of /o/ and /o/, the phonetic and phonological structure of French suggests that duration might be a less important perceptual property in French than in languages like American English. Three 10-step /kot/-/kot/ continua were synthesized by systematically varying frequencies of the first two formants of the vowel nuclei. The three continua differed in vowel duration (140, 180, and 220 msec). Two perceptual tests, identification and 5-choice category rating, were presented to three listener groups: native French, native American English who had studied French, and native American English who did not know French. For both native American English groups, spectrally ambiguous vowels were identified and rated more often as /o/ when these vowels were long and as /o/ when short, thereby showing a trading relation between temporal and spectral information. In contrast, native French listeners showed little effect of duration in either perceptual task. Despite this perceptual insensitivity to duration, acoustic measurements showed that these French subjects' productions of /o/ and /o/ consistently maintained a duration difference. Results are interpreted to support the view that perceptual integration of the acoustic properties relevant to a phonemic contrast depends not only on covariation of the properties in the production of that contrast, but also on the prominence of this covariation in the language's phonological system.

Acoustic versus phonological context effects in the perception of formant transition continua

The effect of the following vowel on listener's judgments of place of articulation in formant transition continua is perhaps the archetypal example of context sensitivity in speech perception. Is this sensitivity explicable in terms of acoustic context alone or must the judged phonological identity of the following vowel also be considered? Shammass [unpublished dissertation, University of Alberta (1985)] designed transition‐continua experiments involving ambiguous and nonambiguous vowel stimuli to examine this question in detail. Linear logistic analysis of these data confirms the systematic role of acoustic context effects, but indicates that the phonological context effects are of a highly restricted nature and, furthermore, are both task and listener dependent. A summary of this analysis is presented and the results are discussed in the context of other results from the literature (including those of Mermelstein, Massaro, Ohala, and Repp). The evidence is consistent with the hypothesis that phoneme‐sized units (rather than diphones or larger units) serve as the primary interface between acoustic cues and phonological units.

Vowel boundaries and waveform patterns

Relations between changes in the temporal fine-structure of synthetic vowel waveforms and perceptual boundaries are examined. Small changes in fundamental frequency (FO) are shown to affect temporal fine-structure and perceptual boundaries in a correlated manner. Experiments show that ambiguous vowel sounds lying on the continuum /i/ and /ɪ/ can be shifted to either category by manipulating FO and that the shifts can be predicted from the resulting changes in waveform. In an additional study, a series of pseudovowels identical in FO and formant center-frequency values but differing in temporal fine-structure are shown to bear perceptual relationships to one another similar to a normal /i/-/æ/ vowel series. The results indicate that temporal analysis of the first formant of vowel sounds can aid in delineating vowel boundaries. [Supported by NIH grant NS 03856 to CID.]

Some contextual effects on the perception of synthetic vowels.

The purpose of this investigation was to determine if the identification of vowel quality in selected vowel-like sounds is affected by the characteristics of an immediately preceding vowel. Listeners were presented recordings of pairs of vowels and instructed to identify the second of each pair as [i, I, ε, Λ] or [α] . The initial, or " affecting ", vowel in each pair was either [i, I, ε, Λ] or [α] ; two sets of affecting vowels were used—human and synthetic. All of the affected vowels (stimuli to to be judged) were generated synthetically; three were intended to be good examples of the vowels [I, ε, Λ], while the formant frequencies of the remaining four were intended to create ambiguous vowels (that is, two had formant frequencies between those of the synthetic vowels [I] and [ε] ; the others, between [ε] and [Λ]). Results indicate that some shift in vowel identification does occur as a function of an immediately preceding vowel. The effect, of moderate strength when the two vowels are similar, decreases as a function of the difference between the F1-F 2 ratios of the two vowels.