Perception of Temporal and Spectral Information in French Vowels

8/2/2019 Perception of Temporal and Spectral Information in French Vowels

1/16

Perception of Temporal andSpectral Information in FrenchVowels*Terry L. Gottfriedt and Patrice Speeter Beddort t

Perceptual effects of orthogonal variations in temporal and spectralinformation differentiating French /0/ and /;)/ were examined. Although bothparameters contribute to acoustic differentiation of /0/ and /;)/, the phoneticand phonological structure of French suggests that duration might be a lessimportant perceptual property in French than in languages like AmericanEnglish. Three lO-step /kot/-/bt/ continua were synthesized bysystematically varying frequencies of the first two formants of the vowelnuclei. The three continua differed in vowel duration (140, 180, and 220 ms).Two perceptual tests, identification and 5-choice category rating, werepresented to three listener groups: native French, native American English whohad studied French, and native American English who did not know French.For both native American English groups, spectrally ambiguous vowels wereidentified and rated more often as /0/ when these vowels were long and as /;)/when short, thereby showing a trading relation between temporal and spectralinformation. In contrast, native French listeners showed little effect ofduration in either perceptual task. Despite this perceptual insensitivity toduration, acoustic measurements showed that these French subjects' productionsof /0/ and /;)/ consistently maintained a duration difference. Results areinterpreted to support the view that perceptual integration of the acousticproperties relevant to a phonemic contrast depends not only on covariation ofthe properties in the production of that contrast, but also on the prominence ofthis covariation in the language's phonological system.

INTRODUCTIONThe temporal and spectral characteristics of vowels covary in the vowel contrastsofmany languages (Lehiste, 1970; Straka, 1959). In some ofthese languages. phoneticmeasures and phonological patterning point toward vowel length as the distinctive

property of the relevant contrasts (e.g., Czech, Hungarian, and Serbo-Croatian). Inother languages in which temporal and spectral differences co-occur, vowel qualityrather than quantity may be analyzed as the distinctive property. For example, inAmerican English, the "tense-lax" pairs, such as Ii, 11, Ire, e/ , lu. u/. and 10. AI differin both fonnant frequency and intrinsic duration: Tense vowels are more peripheraland longer than their lax counterparts (Peterson & Lehiste. 1960). These length

Haskins LaboratoriesStatus Report on Speech Research

51

SR-93/941988


2/16

52

differences are traditionally analyzed as phonologically redundant (but see Delattre.1962. for discussion).Whether or not vowel length is contrastive in a language has been shown toinfluence certain perceptual judgments by speakers of that language. For example,

Bastian and Abramson (1962) obtained steeper. more consistent labeling functionsfor Thai listeners than for American listeners on a length contrast in Thai. S imilardifferences were reported by Keating (1985) for Czech and American listeners on alength contrast phonemic in Czech.

Such findings, of course, do not mean that phonologically non-contrastive vowelduration is perceptually irrelevant. 1Wo studies (Assmann, Nearey. & Hogan. 1982;Strange, Jenkins. & Johnson. 1983) found that when th e nuclei ofnatural (Canadianand American English) vowels were truncated so that all vowels had the same length.listeners confused spectrally adjacent vowels that normally differ in length. Rakerd(1984) showed that a major factor in listeners' judgments of similarities betweennatural American English vowels was vowel duration. Indeed. under certainconditions. a duration difference alone is sufficient to specifY a vowel contrast inEnglish. Ainsworth (1972) showed that variations in durat ion of synthetic vowelsaffected th e category name given these vowels by British English listeners.Mermelstein (1978) showed that synthetic vowels that were spectrally ambiguous forle i and lrel (i.e., the frequency of the first formant fell between typical le i and lrelvalues) were labeled according to vowel duration: Longer vowels were called lrel andshorter vowels le i (see also Stevens. 1959).

There is substantial perceptual evidence. then. that vowel duration conveysinformation for vowel identification in English. Such findings might lead us toexpect vowel duration to be perceptually relevant for any language in which. likeEnglish. phonologically nondistinctive length covaries with spectral properties.Alternatively. we might hypothesize that. for a given language. the perceptualrelevance of duration for a particular vowel contras t depends not simply on thepresence of duration as an acoustic correlate of that contrast. but rather on theoverall prominence of vowel duration within the language's phonological system.

Bennett (1968) compared the perceptual role of vowel duration in English andGerman. Presented with an unfamiliar vowel contrast varying in both temporal andspectral characteristics ([w:H'lI)). British English listeners aSSigned more weight thandid German listeners to the temporal variation. ThiS perceptual difference may berelated to phonological differences between English and German. In both languages.duration and spectral properties covary in several vowel contrasts. However. i t isonly in English that vowel duration provides information for voicing of a follOWingobstruent. as well as for vowel identity.

French also differs from English in the phonological prominence of vowelduration. although the specific differences between French and English are not thesame as those between German and English. Various factors suggest that vowelduration has a relatively minor role in Parisian French (Delattre. 1959; Fry, 1968).For example. temporal and spectral cues covary in only three vowel contrasts inFrench (Delattre. 1959): la/-/a/. 101- lrel, and lo/-/'J/. The phonemic status of thef irst two of these pairs is questionable. The Iai-Ial contrast is not commonlyobserved in spoken French (Delattre. 1957), and the 10/-/rel contrast is confined totwo minimal pairs (Valdman. 1976. p. 56). The only common vowel distinction thatconsistently maintains spectral and temporal differences involves the mid backvowels 101 and 1'::>1. as in the minimal pairs paume-pomme 'palm-apple: saute-sotte'leaps-foolish,' and cote-cotte 'rib-tUnic.' These vowels are opposed only in syllablesclosed by a final consonant. (This is similar to the situation in English where the"tense-lax" vowel pairs also contrast only in closed syllables.)

Gottfried & Beddor


3/16

53

Spectrally. French 101 and I':JI consistently differ in first and second formant (F 1and F2 ) frequencies (Chollet & Malecot. 1980). Temporally. measurements byBrichler-Labaeye (1970) indicate that 101 is roughly 20-40% longer than I':JI in closedsyllables. Gottfried (1984) found that 101 was. on the average. 25% longer than I ':JI inIt/-vowel-Itl and vowel-It/ syllables produced by na tive speakers of French.(Intrinsic vowel duration differences are larger for some vowel contrasts in English.See Peterson & Lehiste. 1960; Strange. Verbrugge. Shankweiler. & Edman. 1976.)Thus spectral and temporal information simultaneously vary in certain vowelcontrasts in both French and English. However. the role of temporal information isless prominent in French in that the number of such contrasts is smaller in French(Delattre. 1959). and the frequency of the closed syllable context . where suchcontrasts occur. is less in French. Dauer (1983) reports that English closed syllablesaccount for 56% of the total in spoken language. whereas French closed syllables

account for only 26%.The present study investigated whether this difference in vowel systems hasperceptual manifestations. such that French listeners are l es s sensitive than

American English listeners to temporal differences in vowels. Some previOUSresearch points toward the pOSSibility of such perceptual differences. Gottfried (1984)examined the accuracy of identification of natural French vowels by native Frenchand by French-speaking American listeners. Analysis of th e errors by th eAmericans indicated inappropriate use of vowel duration in identifying the vowels.For example. when a natural token of French I':JI was unusually long in duration. theAmericans were more likely than the French to label that token lo/-that is . for th eAmericans. but not for the French. the temporal cue overrode the spectral cue.

This study used the trading relations paradigm (see Repp. 1982) to test the relativecontribution of temporal and spectral information to vowel identification. In thisparadigm. two acoustic parameters that naturally covary in some phonetic contrastare manipulated independently of each other. In instances where both parametersare perceptually relevant. a change in the value of one parameter can be "traded off'against an opposing change in the o ther parameter. thereby maintaining phoneticcategory identity. For example. voiceless and voiced initial stop consonants differ invoice onset time (VOT) and onset frequency of the F 1 transition. To maintainphonetic eqUivalence of intermediate stimuli. VOT must be increased when F1 onsetfrequency is lowered (Summerfield & Haggard. 1977). Trading relat ions have alsobeen found for cues Signaling place of articulation in stops (e.g. Dorman. StuddertKennedy. & Raphael. 1977). manner of articulation in consonants (e.g . Repp.Liberman. Eccardt. & Pesetsky. 1978). and many other phonetic contrasts.

The focus of our s tudy was whe ther French. and American English listenersdiffered in the extent to which. temporal information can be traded against spectralinformation in the perception of the French vowels 101 and 1:>1. American listenerswere expected to use duration. as well as spectral information. as cues to the identityof these vowels. ThiS expectationwas based on the systematic role of duration in theperception of English vowels demonstrated by the studies cited above.Given the systematic difference in the duration of 101 and I ':JI in French. Frenchlisteners might also be expected to be sensitive to temporal information. Under thishypothesis. French listeners should identify long. spectrally ambiguous vowels as101 and short. spectrally ambiguous vowels as h i. That is. vowel identity shouldshift as a function of vowel duration. This prediction is consistent with tradingrelations studies that Indicate integration ofacoustic information when the sourcesof information cospecify a phonemic contrast (Repp. 1982. 1983).Alternatively. th e restricted role of duration in the French vowel system mightlead uS to predict that French listeners would be insensitive to temporal cues. even

French Vowels


4/16

54

for a vowel distinction in which duration differences are systematically maintainedin French. The latter hypothesis claims that perceptual integration of acousticproperties relevant to a given vowel contrast depends in part on the extent to whichthese acoustic properties correlate within a phonological system.Method

StimuliThe stimuliwere five-formant synthetic versions of 101 and 1::>1 embedded in a Ik/

vowel-/tl context, corresponding to the French words cote and cotte. All stimuli weregenerated on a serial software formant synthesizer at Haskins Laboratories. Theacoustic characteristics of th e two endpoint synthetic stimuli were based on ananalysis of natural tokens of th e French words, spoken by a male native speaker ofFrench. Intermediate versions of these syllables were synthesized by manipulatingthe spectral and temporal characteristics of the vowel nuclei.The spectral manipulation was in FI and F:z. Table 1 shows the initial FI (Fli),steady-state targetFI (FIt), and final FI (Flf) for each ofthe 10 steps in the Ikot/-/btlcontinuum. Likewise, the initial (F:zi)' target (Fzt), and final (Fzf) values for Fz are

shown. Initial, target, and final Fa were 2275 Hz, 2380 Hz, and 2800 Hz, respectively,for all stimuli. The F4 was fixed at 3300 Hz and Fs at 3850 Hz. The initial burst forIkl was 20ms, followed by a 20-ms voice-onset time. The initial F I transitionwas 30ms long; the initial Fztransition was 40ms; and the initial F3 transition was 20 ms.starting 10 IDS after the FI and Fz transitions. All final transitions were 40ms. After40 IDS of silence. there was a 20-ms final burst for It!. The F0 was at 110 Hz for theinitial transitions and steady-state, and fell linearly to 85 Hz in the las t 80 ms ofvoicing.

TABLElInitial (i), target (t), andfinal (f) synthesis values for FI and Fzofthetensteplkot/-/bt/continuum.

Formant values (in Hz)StinIUlusNumber Fli Fit Flf Fzi Fzt Fj

1 315 400 260 1270 800 13402 320 415 265 1285 835 13553 325 430 270 1300 870 13704 330 445 275 1315 905 13855 335 460 280 1330 940 14006 340 475 285 1345 975 14157 345 490 290 1360 1010 14308 350 505 295 1375 1045 14459 355 520 300 1390 1080 146010 360 535 305 1405 1115 1475

Three temporal variations of each of these 10 stimuli were synthesized bymanipulating the duration of the vowel steady-state. Vowel length (includingformant transitions) was 220 ms for long vowels. 180ms for medium vowels. and 140IDS for short vowels. The long vowel length was appropriate for French cote and theshort vowel length for cotte. The ten spectral and three temporal variations provideda total of 30stimuli.

Three audio tapes were constructed using these stimuli: a familiarization set, anidentification test, and a rat ing tes t. The familiarization se t consisted of 15Gottfried & Beddor


5/16

55

randomly selected members of the stimulus set. The identification test contained 10instances of each of the 30 st imuli , for a total of 300 test items arranged in randomorder and grouped in blocks of 20. The interstimulus interval within blocks was1.5 s; the interblock interval was 5 s.

The 2-choice identification task, which provided one measure of phoneticcategorization, was supplemented by a 5-choice category rating task where theendpoint categories were for good exemplars of /0/ and /-;>/, and the middle categorieswere for vowels judged to be intermediate to the good exemplars. We speculated thatthis rating task might be more sensitive to th e effects of vowel duration than theidentification task. That is , even though vowel duration might no t influencephonemic categorization, it might affect finer judgments of categorization, for theFrench as well as the American listeners. The rating test consisted of 3 instances ofeach of the 30 stimuli, yielding 90 randomly arranged test items that were groupedinto blocks of 20. The rating test interstimulus interval within blocks was 3 s; th einterblock interval was 5 s.Subjects

Three groups of 12 listeners each participated in th e experiment: native speakersof French. native speakers of American English who had studied French (that is,"French learners"), and native American English speakers with no knowledge ofFrench (that is, "naive English"). These three groups enabled us to look at the effectof duration on vowel identification across var ious degrees of competence in aparticular phonological system: native knowledge versus second language knowledgeversus no knowledge of the vowel system within which the /0/-/-;>/ contrastfunctioned.

The native American English speakers were students at Yale University or theState University of New York at Purchase. The French learners had studied Frenchat least five years or had spent one or more years in France. Most of th e nativeFrench speakers were students or faculty at Yale or SUNY-Purchase. All subjectswere paid for their participation.ProcedureSubjects were tested in small groups of one to three in sound-attenuated rooms atHaskins Laboratories and at SUNY-Purchase. Stimuli were presented binaurallyover headphones in a I-hour session. Instructions were given in English for alll is teners. The same procedure was used for the three groups, except that, prior totesting, the naive English listeners heard a brief tape-recorded training seriesconsisting of natural-speech tokens of cote and cotte produced by a native Frenchspeaker. They were told that these were natural speech examples of th e Frenchvowels they were going to identify in the perceptual tests. Otherwise, the experimentconsisted of the familiarization set, the identification test, and the rat ing test, in thatorder.

Subjects were told they would hear computer-generated tokens of cote and cotte.The familiarization set was presented to acquaint listeners with the stimuli and thespeed of the identification test. In the identification test, subjects were instructed tolisten to each syllable and record the vowel in each syllable on printed score sheets.Native French and French learners recorded "0" or "0"; naive English subjects (whowere unfamiliar with French orthography) recorded "0" or "J." symbols that wereexplained to them during their training session.

After the identification test and a short break, listeners were informed that th eidentification test had contained several different versions of cote and cotte. bu t thatsome versions were more natural , or more native-like, than others. A rating testrequired them to categorize vowels on a 5-point scale, where 1 represented a goodexemplar of /0/ .3 represented an ambiguous vowel midway between a good /0 / and

French Vowels


6/16

56

h i, and 5 a good I'J/. Listeners circled t he ir rat ings on score sheets. Thus 10identification judgments and t hree rat ings per subject were obtained for eachstimulus.

Results and DiscussionThe pooled results of the identification test for the 12 native French speakers areshown in Figure 1. Percent 101 responses are plotted as a function of stimulusnumber (corresponding to the ten spectral changes) for t he th ree vowel durations.

The vertical line drawn through the 50% point in th e curves indicates the lo/-hlboundary for the functions. As the figure shows, the cross-over point was notdifferent for th e three durations. An analysis of variance was performed on th enumber of 101 responses for each vowel duration, summed over th e 10 spectralpatterns. For the native French speakers, the numbe r of 101 responses did notsignificantly change as a function of vowel duration, F{2,22)=0.03.Identification Native French (n= 12)

20

o

100

.-... Long Vowelstr- -6 . Medium Vowels0----0 Short Vowels

80

60

C 40Q)(j)Q..

(fJQ)(fJcoa.(fJ(j)a:

1-1_--'-_---'__.1 --1 _ - - ' - _ - - - - J . _ - - ' _ ~ ~ _ ~ I__ l--.Jo 2 3 4 5 6 7 8 9 10Stimulus Number

Figure 1. Identification of synthetic vowel series by native French listeners.Figure 2 gives the identification test results for the 12 French learners. Unlike thenative French, th e learners' performance showed an effect of vowel duration. asshown by th e diverging (rather than overlapping) identification curves for the threevowel durations and by the different 50% crossover points (indicated by the vertical

lines) for the three durations. The number of 101 responses by the French learnersincreased significantly as duration increased. F{2,22)=16.93. p


7/16

57

advanced learners. Although both learner groups showed an effect of duration. asignificant interaction of duration with group indicated that the duration effect wasstronger in the less advanced group. F(2.20)= 3.80. p


8/16

58

listeners from the overall analysis. The revised analysis revealed a significantinteraction of duration and linguistic background-that is . vowel duration affectedcategorization rat ings by th e learners. but no t the native French listeners. F(2,44)=7.75. p


9/16

59

9 10

. - . Long Vowelsts--I::>. Medium Vowels0----0 Short Vowels

3 4 5 6 7Stimulus Number

2

Identification Naive English (n== 12)SM L100

80(j)OJ(j)C0 60.(j)OJa:

C 40OJ2OJ0..

20

0l0

Figure 4. Identification of synthetic vowel series by naive English listeners.

Rating

------- . 2.5

1lal-IikeOJc"-a: 3.0cmOJ::;::

hi -like

3.5 Long MediumVowel Duration

Short

Native French

French Learners

Naive English

Figure 5. Mean rating of synthetic vowels as a function of duration by native French, French learners.and naive English.


10/16

60

1700 \600 \500 \1400 \ ~ I'JIF2 1300 1"'-(Hz)

'"200 3 ' """

,1100"'"000 ""

""00 I 101 """'"""00 450 500 550 600 650 700F1 (Hz)

Figure 6. Mean first (FI) and second (F2) formant frequencies (in Hz) for /0/ and h / spoken by fivetalkers.ACOUSTICAL ANALYSIS OF NATURAL SPEECH

MethodThe key words in the natura l speech samples were six minimal pairs differing in10 / and 1':)/: cote-cotte 'rib-tunic,' saute-sotte 'shift-foolish,' paume-pomme 'palmapple,' rauque-roc 'hoarse-rock,' notre-notre 'ours-our,' and rode-rode 'prowl-grind.'

Eight additional words (viz., belle 'beautiful,' jlamme 'flame,' peur 'fear,' seche 'dry,'rage 'rage,' sur'sure,' rotes 'say,' and Joule 'crowd') were included so that the speakerswould be less likely to be aware of th e vowel contrast of interest.

Each of t he se 20 words was embedded in two frames: a variable sentence (seeAppendix) and a fixed carr ier sentence. The variable frame provided relativelynatural and meaningful conditions for the product ion of the key words. Thesesentences were constructed to have approximately the same number of syllables,with each member of a minimal pair occupying a similar sentential position. Thefixed carrier sentence, On dit le mot souvent ('They say the word _often'), provided control over th e acoustic context of th e key words. Speakers readeach randomly-arranged l is t of sentences twice, so that there were four instances (perspeaker) of each key word, two produced in the variable frame and two in the fixedcarrier frame.

The 20 tokens (4 repetitions x 5 speakers) of each of th e 12 key words containing101 or /':)1 were digitized at a sampling rate of 10 kHz, after low-pass filtering at 4.9kHz. Measurements of vowel duration were obtained from the digitized speechsamples (using a waveform display program at Haskins Laboratories). Vowel onsetwas determined by the beginning of the Significant periodic portion of th e waveformfor syllables with initial Ip/, Ik/, and lsi. Vowel onset of syllables with initial Inland Ir I was determined from the point at which th e waveform changed itscharacteristic pattern, which usually corresponded with a noticeable increase inoverall amplitude. Vowel offset was determined by the cessation of significantperiodic pulsing for syllables ending in s top consonants . For the syllable ending inIml, offset was determined by th e change in characteristic waveform pattern and the

Gottfried & 13eddor


11/16

61

sharp reduction in overall amplitude. Thus. the vowel durations included formanttransitions as well as the relatively steady-state portion.Formant frequencies were measured for all Ik/-vowel-/tl words. using a linearpredictive coding (LPC) analysis. A Hamming window of 24 ms width was placed onthe part of the signal that had the most extreme formant excursion (as determined by

visual inspection) from the initial and final transition portions. This was usually inth e first half of the vowel and corresponded to a point in the vowel shortly (about 50ms) after the rapid increase in amplitude at vowel onset. Formant frequencymeasurements were accurate within a range of 20 Hz.Results and Discussion

Table 2 gives the mean durations for the 101 and 101 portions extracted from the sixminimal pairs. As shown by the values in the table. vowel durations varied as afunction of sentence frame and syllabic context (cf. OShaughnessy. 1981). Therewasalso conSiderable variation in duration from speaker to speake r. However.averaging across sentence frame. all five speakers produced a longer 101 than 101 ineach of the vowel contrasts. except lrodl-Ind/. 1

TABLE 2Mean duration in ms (avera.ged over two tokens from five speakers) for 101 and I':JI and ratio oflong/short vowel duration for six minimal pairs in variable and fixed sentence frame.Sentence Frame

Fixed Variable Mean

Ikotl 143 78 111Ibtl 97 73 85Ratio 1.48 1.06 130Isotl 114 78 96IS':Jtl 81 75 78

Ratio 1.41 1.05 1.24Ipoml 169 136 152Ip':Jml 121 82 102Ratio 1.40 1.65 1.50Irokl 152 161 157Ir':Jkl 107 81 94Ratio 1.43 1.99 1.67Inotrl 161 106 133In':Jtrl 138 65 101Ratio 1.17 1.62 1.31Irodl 205 164 184Ir':Jd/ 188 168 178Ratio 1.09 0.97 1.03Mean 101 157 121 139Mean h i 122 91 106Ratio 1.29 133 131


12/16

62

The effect of sentence frame on vowel duration was not consistent across syllabiccontexts. The duration difference between 101 and 1,:>1 was larger in the variablesentence frame than in the fixed sentence frame for Ipom/-/p':>m/. Irok/-/r':>k/. andInotr/-/n':>tr/. However. Ikot/-/lotl and Isot/-/s':>tl were better differentiated in thefixed frame. Most speakers in the variable sentence frame did no t maintain adifference in duration between Ikotl and 110tl or Isotl and ISJt/. 2By contrast. the spectral difference between Ikotl and 110tl was very large. MeanF 1 values for Ikotl and 110tl were 438 Hz and 541 Hz, respectively; mean F2 valueswere 1160 Hz and 1512 Hz. and mean F3 values were 2608 Hz and 2654 Hz.respectively. Figure 6 plots the mean F 1 and F2 for each ofthe five speakers for Ikotland 11ot/. Error bars represent th e standard error of the formant value for eachindividual speaker. The figure shows that all five speakers made a clear spectraldistinction between Ikotl and 11ot/.3

GENERAL DISCUSSIONThe results of the perceptual tests of native French listeners show no evidence of

perceptual integration of spectral and temporal information in the Ikotl -Ik,;;tlcontrast. These listeners were insensitive to temporal information no t only incategorizing th e vowels as 101 or h i . but also in a 5-choice category rating task.Acoustical analysis revealed that five of these same native French. who did not useduration as a perceptual cue. consistently maintained a distinction in duration (atleast in stressed pOSition) in their productions of these vowels.We believe that the explanation for this discrepancy between perception andproduction measures lies in th e phonological structure of th e French vowel system.The 10/-1,;;1 distinction is th e only common vowel contras t in French in whichduration and formant frequencies covary. We suggest. then. that our native Frenchlisteners were insensitive to duration as a cue to vowel identity in this contrastbecause duration is not a reliable cue in other French contrasts.Support for a phonological explanation of the native French data is provided bythe resul ts of th e native English lis teners . inasmuch as the same stimuli wereperceived differently by the two language groups. In con tras t to th e native French

listeners. th e native English listeners were influenced by the duration of French 101and 1,;;1 in their categorization of these vowels. The English listeners' sensitivity toduration as a cue to th e French distinction is consistent with the prominent role ofduration in the English vowel system (Ainsworth. 1972; Peterson & Lehiste. 1960;Rakerd, 1984). Our American English data are also similar to the British Englishresults reported by Bennett (1968): Both subject groups used duration in categorizingthe vowels of an unfamiliar contrast.

I f the effect of durat ion shown by the native English speakers is a function of thesystematic role of duration in differentiating spectrally adjacent vowels in English.it is no t surprising that both groups of English speakers-the naive English and theFrench learners-demonstrated this effect. The naive English listeners apparentlyused the temporal information appropriate to English vowel contrasts in labelingthe unfamiliar /0/-1,;;1 contrast. (Many American listeners said that th e vowelssounded like the English loW I-/AI contrast, as in 'coat' vs. 'cut:) Similarly, the Frenchlearners also used this information. since French 101 and h i (which differ induration as well as spectral shape) provide no phonetic impetus for the learners todiscard vowel duration as a relevant cue. Yet we do see a suggest ion of adevelopmental trend among the learners: The advanced learners were more like thenative French in their perception of these vowels than were th e less advancedlearners.

Gottfried & Beddor


13/16

63

In conclusion, our findings showed that vowel duration was perceptually relevantfor the native English listeners, but not the native French, as a cue to the Ikotl-{bt ldistinc tion. While i t is of course possible that t he French might demonstrate atrading relation under more stringent conditions (e.g., smaller spectral steps mkghlead to a greater duration effect), comparison with the nat ive English listenersindicates a relative insensitivity of the French to temporal information. We haveargued that this insensitivity is a funct ion of th e highly restricted role of vowelduration in the French vowel system, but it is possible that our perceptual data alsoreflect a change in progress involVing the French 101-1':)1 contrast. That is, it may bethat a phonological contrast for which a particular phonetic parameter is anacoustic, but no t a perceptual, correlate is an unstable contrast. I f so, then thecurrent perceptual irrelevance of temporal information may be followed at somelater time by loss of temporal differences in the production of 101 and 1':)1. Regardlessof the future of the 101-1':)1 contrast in French, ou r perceptual data lead us to thefollOWing conclusion: Perceptual integration of the acoustic properties relevant to agiven vowel contrast does not simply follow from experience with that contrast.Rather, perceptual integration depends on the extent to which th e acoustic propertiescorrelate within the broader context of a phonological system.

ACKNOWLEDGMENTThis research was supported by NIH Contract N 0 1 - H D ~ 2 9 1 0 and Grant HD-01994 to

Haskins Laboratories, the State University of New York Research Foundation (to thefirst author while at SUNY at Purchase), and NIH Postdoctoral Fellowship Grant NS07196 (to the second author). We wish to thank Dr. Bruno H. Repp and twoanonymous reviewers for their helpful comment s on earl ier versions of th emanuscript.APPENDIX

Sentences used as the variable sentence frames.1. Sa voix Hait rauque apres qu'elle avait chante.'Her voice was hoarse after she had sung.'

2. J'ai vu une belle fleur dans Ie jardin.'I saw a beautiful flower in th e garden.'

3. II Yavait une saute de vent ce matin.There was a gust ofwind this morning.'

4. Nos amis n'aiment pas notre histoire.'Our friends didn't like our story:

5. Son souffle fait trembloter la flamme de.la chandelle.'His breath makes the candle's flame tremble.'

6. II mange la potntne jaune que je lui ai donnee.'He is eating th e yellow apple that I gave him.'7. L'enfant avait peur quand il secouchait.'

'The child was afraid when he went to bed.'8. Elle a mange une cote de boeufhier sair.

'She ate a beefrib y e s t ~ r d a y evening:9. On dit qu'il rode les rues en quete de victimes.

'They say that he prowls the streets in search ofvictims.'

French Vowels


14/16

64

10. Le charpentier seche Ie chene vert.The carpenter is drying th e green oak.'

11. Elle n'est pa s assez sotte pour Ie croire.'She isn't silly enough to believe it.'

12.


15/16

65

Jones, D. (1950). The phoneme: Its nature and use. Cambridge: W. Heffner and Sons.Keating, P. (1985). Linguistic and nonlinguistic effects on the perception of vowel duration. UCLAWorking Papers in Phonetics, 60, 20-39.Lehiste, I. (1970). Suprasegmentals. Cambridge, MA: MIT Press.Mermelstein, P. (1978). On the relation between vowel and consonant identification when cued bythe same acoustic information. Perception & Psychophysics, 23, 331-336.

O'Shaughnessy, D. (1981). A study of French vowel and consonant durations. Journal of Phonetics, 9,385-406.Peterson, G. E., & Lehiste, I. (1960). Duration of syllable nuclei in English. Journal of the AcousticalSociety of America, 30, 693-703.Rakerd, B. (1984). Vowels in consonantal context are perceived more linguistically than are isolatedvowels: Evidence from an individual differences scaling study. Perception & Psychophysics, 35, 123136.Repp, B. H. (1982). Phonetic trading relations and context effects: New experimental evidence for aspeech mode o f perception. Psychological Bulletin, 92,81-110.

Repp, B. H. (1983). Trading relations among acoustic cues in speech percept ion are largely a resul tof phonetic categorization. Speech Communication, 2, 341-362. .Repp, B. H., Liberman, A. M., Eccardt, T., & Pesetsky, D. (1978). Perceptual integration of acousticcues for stop, fricative and affricate manner. Journal of Experimental Psychology: HumanPerception and Performance, 4, 621-637.

Smi th , M. R. (1983). The integration of temporal and phonetic cues to word stress. Doctoraldissertation, University of Connecticut, Storrs, CT.Stevens, K. N. (1959). Effect of duration on identification. Journal of the Acoustical Society ofAmerica, 31, 109. (Abstract)

Straka, G. (1959). Duree et timbre vocalique. Zeitschrift fur Phonetik und allgemeineSprachwissenschaft, 12, 276-300.Strange, W., Jenkins, J. J., & Johnson, T. L. (1983). Dynamic specification of coarticulated vowels.Journal of the Acoustical Society of America, 74, 695-705.Strange, W., Verbrugge, R. R., Shankweiler, D. P., & Edman, T. R. (1976). Consonant environmentspecifies vowel identity. Journal of the Acoustical Society of America, 60, 213-224.Summerfield, Q., & Haggard, M. (1977). On the dissociation of spectral and temporal cues to t hevoicing distinction in initial stop consonants. Journal of the Acoustical Society of America, 62, 435448.

Valdman, A. (1976). Introduction to French phonology and morphology. Rowley, MA: NewburyHouse.

FOOTNOTES*Language and Speech, in press.tLawrence University, Appleton, WI.

ttYale University. (Currently at the University of Michigan Program in Linguistics, Ann Arbor, MI).'Table 2 shows that the durations for Irodl were only marginally longer than Ir:Jdl in the fixed frame,and were slightly shorter than Ir:Jdl in th e var iable f rame. Spectral analysis revealed that thevowels in the Ir/-vowel-/dl context had very similar formant frequencies. The mean values of F" F;yand F3 for Irodl were 420, 947, and 2876 Hz, respectively; for Ir-:xl.1 these values were 468, 1023, and2878 Hz. Only one of the five speakers made a clear and consistent spectral distinction betweenthese vowels. The similar ity of both duration and formant frequency measures suggests that theI rod 1-Ir:Jdl distinction was not phonemic for most o f our speakers. (When asked about thi sputative minimal pair, these speakers usually responded that they did not contrast these words,despite the orthography, which represents a dif ference in vowel duration and t imbre in severalother minimal pairs.)2'fhe lack of temporal differentiation in these pairs might be due to the prosodic characteristics of thesentences in which they were embedded. Although we attempted to balance these sentences forkey word posit ion, dif ferences in stress may have attenuated differences in duration for theminimal pair Isot/- IS:Jt/. An analogous explanation of the slight duration differences betweenIkotl and I b t l in the var iable sentence frame is not possible, because the key words are in


16/16

66

unstressed position in both sentences (see Appendix). It may be, however, that in unstressedposition these words do not ordinarily differ in durat ion. Jones (1950, pp. 124-125) asserts thatintrinsically long and short French vowels differ in duration only in phrase-final position.Descriptions of American English have stated that the duration of long vowels, but not that of shortvowels, is greatly affected by stress position (see Durand, 1946, p. 29). Smith (1983) showed that theintrinsic duration difference between / i f and /I/ is smaller in unstressed sentence positions (meanratio of long/short is 1.38) than in stressed position (ratio is 1.69). Our data do not include /kot/ and/ b t / in stressed position in a variable sentence frame, bu t the vowel of /kot/ was nearly 50% longerthan that of / b t / when these words were in stressed position in the fixed sentence frame (see Table2).

'There were fairly large context effects on the spectral characteristics of /kot/ and / b t / . In general,FI was higher in the fixed context, and F2 was lower. However, the spectral differentiation between/0/ and h / was large in both conditions. Because there were only two tokens by each speaker ineach context, FI and F2 values were averaged over context conditions.

Gottfried & Beddor

Date post:	05-Apr-2018
Category:	Documents
Upload:	luis-urbina-thompson
View:	217 times
Download:	0 times

Perception of Temporal and Spectral Information in French Vowels

Documents