Читать книгу Stress Variation in English - Alexander Tokar - Страница 11

Following authors such as Turk & SawuschTurk & Sawusch (1997: 25), Cho et al.Cho et al. (2007: 218), Scarborough et al.Scarborough et al. (2009: 137), PlagPlag et al. (2011), Mücke & GriceMücke & Grice (2014: 48), etc., the present monograph will use the term accent in the sense “pitchpitch accent as a phonetic realization of phrasal stress,” whereas the term stress will be used in connection with other phonetic means (mainly durationduration) that English speakers rely upon in out-of-focus environments, in which a particular syllable bears only word but no phrasal stress. Whereas word stress mainly performs a constitutive functionconstitutive function, which means that it merely “arranges syllables in words” (Sokolova et al.Sokolova et al. 1997: 138–139), phrasal stress “indicates that a constituent contains new or important information: it is focused” (Scarborough et al. 2009: 136; authors’ italics).

Regarding the phonetic realization of phrasal stress, consider, for instance, the combination working theory, which depending upon the meaning it expresses is pronounced working ˈtheory or ˈworking theory. As PoldaufPoldauf (1984: 118) observes, when phrasal stress in this combination falls upon the righthand component theory, it means “a theory that works or functions well.” In the other case, i.e., when phrasal stress is placed upon the lefthand component working, the meaning is “a theory that can be used as a point of departure for some work/a theory that is intended for work, working, or functioning.” Phonetically, however, the difference is that while in ˈworking theory the vowel of the stressed syllable of the lefthand component working has a considerably higher pitchpitch peak than the vowel of the stressed syllable of the righthand component theory, in working ˈtheory, by contrast, the stressed vowel of working and that of theory exhibit near-identical pitch peaks (Farnetani et al.Farnetani et al. 1988). In other words, from the point of view of the speaker, working ˈtheory is actually ˈworking ˈtheory, i.e., “both words are accented,” as Farnetani et al. (1988: 171) observe. Nonetheless, listeners “incorrectly” perceive the pronunciation ˈworking ˈtheory as working ˈtheory, which is most likely due to the unfulfilled expectation of the declination of pitch, normally occurring at the end of an intonation group (Farnetani et al. 1988: 170).

More recently, similar results have been obtained by KunterKunter (2011), whose dissertation is concerned with the accentuation of Noun + Noun combinations, such as budget deficit:

in right-prominent compounds, left and right elements have nearly indistinguishable pitchpitch values, while the pitch of left elements is clearly higher than that of right elements if the whole compound is left-prominent (KunterKunter 2011: 89).

In other words, when budget deficit is pronounced ˈbudget deficit, the stressed syllable of the lefthand component budget is pitchpitch-accented. When, however, budget deficit is pronounced budget ˈdeficit, both the stressed syllable of the lefthand component budget and the stressed syllable of the righthand component deficit are pitch-accented, i.e., budget deficit is actually pronounced ˈbudget ˈdeficit but perceived to have been pronounced budget ˈdeficit.

Even more recently, KöslingKösling’s (2013) dissertation has studied the accentuation of triconstituent Noun + Noun + Noun combinations. Her main finding is that while in so-called right-branchingright-branching combinations such as Boston [gang members] (= “gang members who are from Boston or commit crimes in Boston”) there is usually “a high pitchpitch on both constituent N1 and constituent N2 and a clearly lower pitch on constituent N3,” left-branchingleft-branching combinations such as [credit card] companies (“companies that issue credit cards”) are normally characterized by “a high pitch on constituent N1 but clearly lower pitches on both constituent N2 and constituent N3” (Kösling 2013: 54). In other words, Boston gang members is accented ˈBoston ˈgang members, but credit card companies is accented ˈcredit card companies.

As for word stress, it is unfortunate that many authors do not distinguish between its phonetic realizations in accented and non-accented positions. A classic case is FryFry (1958: 126), who described “the all-or-none effect” caused by changes in the fundamental frequencyfundamental frequency of a syllable, to which the impression that a particular syllable has a particular pitchpitch level is intrinsically connected. What is meant by this all-or-none effect is that in a situation when a listener is required to make a subjective judgment as to which syllable in a word bears stress, a higher pitch peak of a particular syllable is capable of overriding the effect of a reduced loudnessloudness and shorter durationduration of the same syllable. In other words, a particular syllable is likely to be perceived as stressed even when that syllable only has a higher pitch peak, but is shorter and quieter than the other syllables co-occurring in the same word. Notice that Fry’s (1958) experiments involved artificially-synthesized manipulations of the fundamental frequency, duration, and intensityintensity of the syllables in the disyllabic words subject, object, digest, contract, and permit. His subjects listened to these manipulated pronunciations and were then asked to indicate whether the words they had just heard were nouns or verbs. As the nouns subject, object, digest, contract, and permit are stressed initially, whereas in the corresponding orthographically identical verbs the stress is final, the participants’ word-class membership judgments served simultaneously as stress-location judgments.

Similar results were later obtained by Morton & JassemMorton & Jassem (1965), who studied how test persons’ stress-location judgments were affected by manipulations of the pitchpitch, durationduration, and intensityintensity levels of the syllables occurring in the synthesized pronunciations of the nonsense words *Soso, *Sisi, and *Sasa.

Variations in fundamental frequencyfundamental frequency produced far greater effects than variations in either intensityintensity or durationduration, a syllable being marked as stressed if it differed from the ‘context’ fundamental. A raised fundamental was more efficient than a lowered one. (Morton & JassemMorton & Jassem 1965: 159)

A serious problem with the treatment of pitchpitch as a cue of word stress is, however, the well-known fact that phrasal stress usually falls on the same syllable as lexical stress:

When a word bears a pitchpitch accent, that accent is usually attached to the syllable bearing lexical stress […] As a result, a single syllable can be simultaneously lexically stressed and pitch accented, for example, the syllable par in We drove to the ‘PARty. For a word in isolation, then, lexical stress is necessarily confounded with phrasal pitch accent, and vice versa […] (Scarborough et al.Scarborough et al. 2009: 137)

The same does not, however, apply to out-of-focus positionsout-of-focus positions, in which words do not constitute new or important information and thus do not receive phrasal stress.

The determination of the phonetics of ‘stress’ (whatever kind) has turned out to be notoriously difficult […] Earlier assessments put durationduration, fundamental frequencyfundamental frequency and intensityintensity central, in addition to other factors involving the relative ‘strength’ of phonemes in ‘stressed’ syllables. […] Some claimed that fundamental frequency takes the lead in all of this […] It was then shown that the big role of (changes in) fundamental frequency results from the fact that the pitchpitch properties of ‘stressed’ syllables are due to the fact that the examined words are uttered in isolation. This causes their ‘stressed’ syllable to be bearers of intonational pitch movement. When words are examined in positions where their syllables do not attract intonational pitch movement (in ‘out-of-focus’ positions) it turns out that pitch is not a major cue at all, but rather duration, spectral tilt and other effects of articulatory force (which may include a small pitch rise). (van der Hulstvan der Hulst 2012: 1508)

Indeed, much earlier than van der Hulstvan der Hulst (2012: 1508), HussHuss (1978: 86) had reported that

Stress oppositions between words like import ≠ import are audible in the nucleusnucleus of intonation contours where they are distinguished by fundamental frequencyfundamental frequency. In the post-nuclear positionpost-nuclear position the opposition is neutralized: the two stresses are identical in terms of fundamental frequency, but measurably different in terms of intensityintensity and durationduration. (HussHuss 1978: 86)

Similarly, an experimental study by Adams & MunroAdams & Munro (1978), which was concerned with word stress cues relied upon in naturally occurring connected speech, has demonstrated that “durationduration was by far the most frequently used cue” (Adams & Munro 1978: 125).

Furthermore, even as far as in-focus positions are concerned, the all-or-none role of fundamental frequencyfundamental frequency was relativized by the findings of Mcclean & TiffanyMcclean & Tiffany (1973: 283), who had discovered that pitchpitch level contrasts between stressed and unstressed syllables predominate only in initially-stressed (nonsense) disyllables—Say */ˈsasa/ please—whereas in disyllables in which stress occurs finally—Say */saˈsa/ please—the same contrast is mainly cued by durationduration.