「→ 2750 猟
A: Where is she now?
4.2. Experiment 4.1
boundaries were determined based on the sound waves,broadband spectrograIIIs,and narrow band spectrograms. Syllable duration, defined here as the interval between the beginning of the initial consonant closure and the end of the final consonant closure,was also lneasured for each syllable.
Five native English speakers,who were all teaching English at universities in Japan,participated in this experiment.l A1l of them were free of any speech or hearing prOblem. They were all monolingual English speakers who were raised in several different English speaking countries.
bought a new Mexican glass yesterday.'' When the sentence
nucleus was placed on the first f00t(eitherの /θθ Or Meノ′″ゴθ),the duration of the second foot(the phraseわ θlrghι ′)was measured and
COIIlpared.
It is assumed that the duration of the trisyllabic foot Me」 ′″Jiθ
should be longer than the duration of the IIlonosyllabic footと /θθ
no IIlatter if the sentence nucleus is placed on Jayθθ or Mefa″ゴθ.
Therefore, if f00t duration is affected by the duration of the preceding foot in order to maintain temporal rhythln between feet, then it is hypothesized that the duration of the phrase bθ ″gh′ ′
following the word Mela′」iθ shOuld be longer than the duration of the phrase bθ ′g力ι′f01lowing the wordと/θθ.
To ensure that the sentence nucleus was placed on Jayθ θ and Me■aコゴθ
, each subject was asked to produce a response to the
sentence Who bought a new Mexican glass yesterday?"
Experimental Materials:
Who bought a new Mexican glass yesterday?
① Joyce bOught a new Mexican glass yesterday.
Who bought a new Mexican glass yesterday?
C)Melanie bought a new Mexican glass yesterday.
4.2.1.Results
According to the results shown in Table 4.l and Figure 4。 1
below, all subjects, as predicted, produced the phrase bθ ″gh′ ′ longer when it was fo1lowing the word Me■
fo1lowing the word及ソθθ. In other words,the longer the duration of the preceding foot,the lnore the duration of the fo1lowing foot is increased. This result supports the idea that from the rhythmic point of view,there is a tendency ofisochrony in English speech.
Table 4.1:Duration of the phraseわ θロノb′ ′When it was fo1lowing the nucleus
Duration of foot fo‖ owing Jayσθ Duration of foot fo‖ owing″era″ た
SutteCt l 638 ms 646 ms
Suttect 2 410 ms 524 ms
SutteCt 3 433 ms 473 ms
SutteCt 4 322 ms 346 ms
SutteCt 5 31l ms 409 ms
700 600 500 400 300 200 100 0
Duration of the phrase bο ″3カr a when it was fb‖owing the nuc:eus
――――― ―
―― 一
―― ― ―
―
￢
SutteCt l Suttect 2 SutteCt 3 SutteCt 4 SutteCt 5
Figure 4.1:Duration of the phraseわθ″g力ι′when it was fo1lowing the nucleus
□Duration of foot fo‖ owing
″」oy9e
■ Duration of foot following
″Melanie″
4。 3.Experiment 4。 2
1t was shown in Experiment 4.l that a foot is lengthened to match the duration of the preceding foot. However,the foot which was targeted in Experiment 4.l was the one which was following the nucleic foot,so the results lnight have been due to the influence of the nucleus. Therefore, I performed another experiment to deterlnine if the same result could be obtained for English feet which are unaffected by the nucleus,
In Experiment 4.1, each subject was asked to produce a
sentence in response to the question
Who bought a new Mexican
glass yesterday?" In that experiIIlent,the nucleus was supposed to be placed on eitherノθycθ or Meノ′″ゴθ.
However,in this experiment,subjects were asked to read the saIIle sentence as in Experilnent 4.1,but without using a question as a prompt. It was assumed that the nucleus was placed on either the word giass or the word/θ s′θrごay in the target sentences of this experiment because these two words are the last two feet of the target sentences.2
1fisochrOny between feet is independent of the location of the nucleus, then the duration of the phrase bθ 口ghι
a when it was
fo1lowing the word Me■arゴθ shOuld be longer than the duration of the phrase bθ ′g力′′when it was following the word j●/θθ.Experimental Materials:
① Joyceルθ口ghι a new Mexican glass yesterday.
②Melanieゎθ″
̀滋
′a new Mexican glass yesterday.
Theoretically,every utterance contains a nucleus(0'Connor&
Arnold,1973). If a subject reads an utterance in one breath,then the utterance nlust contain only one nucleus(0'Connor&Arnold, 1973).TherefOre,if there was no isochrony between feet in this experiment, then it can be interpreted that the isochrony which was observed in ExperiIIlent 4.l was caused by the nucleus relating to focus. The purpose of this experiIIlent is to investigate if isochrony between feet can be observed in utterances in which the nucleus is not on or near the target foot.
4.3.1.Results
Table 4.2 and Figure 4.2 below show the results of this
experiIIlent. There was isochrony between feet for four out of the five subjects. The duration of the phrase わθ′g力′′
when it was
fo1lowing the word Mela″ ゴθ was longer than the duration of the phrase bθ′≦ψ ι ′when it was fO1lowing the word(元
yσθ.This
ilnplies that isochrony exists between feet which can be observed irrespective of the locatiOn of the nucleus.
Table 4.2: Duration of the phrase bought a when it was not
fo1lowing the nucleus
Duration of the foot fo‖ owing Jayσ
̀
Duration of the foot fo‖ owing″θra″Jic
SutteCt l 440 ms 506 ms
SutteCt 2 450 ms 416 ms
SutteCt 3 340 ms 414 ms
SutteCt 4 316 ms 356 ms
SutteCt 5 275 ms 284 ms
Duration of the phrase
not fo‖owing
ルο曜
"オ
a when it was the nuc:eus
ms.
600 500 400 300 200 100 0
口 Duration of foot fo:lowin『
″」oyce″ ￨
■l Duration of foot fo‖ owingi
″Melanie″ ￨
SutteCt Suttect 1 2
SutteCt Suttect Suttect 3 4 5
Figure 4.2:Duration ofthe phrase bought a when it was not fo110wing the nucleus