Delivery style and listener response in the rhythmical performance of Shakespeare's sonnets.
This paper offers further empirical evidence in favor of my conception of poetic rhythm and performance as presented in my book Poetic Rhythm: Structure and Performance--An Empirical Study in Cognitive Poetics. (1) It claims that in an enjambment, for instance, the performer may convey both the verse line boundary and the run-on sentence as perceptual units, however strained, by having recourse to conflicting phonetic cues: cues of continuity and discontinuity simultaneously. In my book I provided much empirical evidence for this assumption.
I have adopted Wellek and Warren's position, who argue in their Theory of Literature (1956, Ch. 13) that in order to account for poetic rhythm, one must assume the existence of not one, but three metrical dimensions: prose rhythm, metric pattern, and performance. My recent work has been devoted to the hitherto neglected performance dimension. My position regarding delivery style can be stated with reference to the "Performance" entry of The New Princeton Encyclopedia of Poetry and Poetics. "C.S. Lewis once identified two types of performers of metrical verse: 'Minstrels' (who recite in a wooden singsong voice, letting scansion override verse) and 'Actors' (who give a flamboyantly expressive recitation, ignoring meter altogether)" (Preminger and Brogan 1993, 893). In Wellek and Warren's terms, the Minstrel subdues prose rhythm, and foregrounds the metric pattern; the Actor subdues the metric pattern in favor of the prose rhythm. My position is that there is a third, "rhythmical performance" too, in which both metric pattern and linguistic stress pattern can be accommodated, such that both are established in the listener's perception.
Some reciters of poetry adopt one or another type of solution quite randomly; but some make a deliberate choice in adopting a consistent delivery style, whether consciously or unconsciously. My attempt to compare two readings of the same text by the same actor reveals that during a long career an actor may change his aesthetic conception, perhaps unconsciously. Such a consistent change of conceptions may be detected even when looking into the minute details of the performances of one verse line. I personally believe that rhythmic complexities arising from conflicting patterns are there in order to realize them in vocal performance too. But in our cultural situation both the "actor's approach" and the "rhythmical performance" are considered legitimate. At any rate, my treatment of the issue will be mostly descriptive; but I will also explore how description may fade into evaluation.
In my 1977 book, A Perception-Oriented Theory of Metre, I suggested that when the endings of the syntactic unit and the metric unit do not coincide (that is, when syntax is run-on from one line to the other, or the regularly alternating weak and strong positions conflict with the irregularly alternating stressed and unstressed syllables), the reciter may indicate continuity and discontinuity at one and the same time by having recourse to conflicting cues. In my later empirical work I extended this principle to additional rhythmic phenomena. My own way in empirical research is to collect judgments from students, colleagues or my research associates whether the performer was successful in conveying the conflicting aspects of language and versification. And if possible, I try to compare alternative possiblities. Then I am looking for cues in the phonetic structures of the recordings, trying to find support for the intuitive judgments. In my present research I have also elicited responses through the internet medium.
In my recent research I am relying on two papers by Gerry Knowles (1991, 1992). In the first one he investigated the nature of tone-groups. He explored the external discontinuities at the tone-group boundaries. These are temporal discontinuation (pause), pitch discontinuation (a sudden change in [F.sub.0]--pitch, in plain English) and segmental discontinuation (that is, in normal speech the articulation of adjacent words is overlapping; when there is no overlap, it may count as discontinuity, even if there is no pause). Glottal stops in words beginning with a vowel, or word-final stop releases too may indicate segmental discontinuation. (2) This would be the most evasive type of discontinuity. "The important distinction that seems to be emerging is between boundaries with or without pauses" (Knowles 1991). In what follows, I shall explore how these correlates of tone-group boundaries can be exploited as conflicting cues for the perceptual accommodation of the conflicting patterns of speech and versification.
One of the most effective kinds of segmental discontinuity is the prolongation of a phoneme or of a syllable at the end of an utterance, arousing (very much like fermata in music) a sense of stability and lack of forward motion. While this is most useful in the kind of research I am engaged in, there is a big problem with this notion. There is no standard by which we can determine whether a phoneme or sequence of phonemes is longer or shorter than ought to be. Consequently, one must rely in this respect on one's intuitive judgment, or some roundabout reasoning about measurements and comparisons. In my recent work I have tried out two new methods: comparing the word in the poetic context to readings in the audio version of Merriam-Webster's Collegiate Dictionary; and instrumental manipulation of the phonetic structure of the words. In the above dictionary an audio pronunciation of the "citation form" of words is provided. Thus, the dictionary reading may serve as an objective standard (that is, without distortion of personal feelings or versification requirements), from which the artistic recital deviates. By instrumental manipulations I have inserted glottal stops or lengthened existing speech sounds in verse lines, listening to the rhythmic effect of the outcome.
A major forward-grouping agent in the rhythmic performance of divergent poetry is what Knowles called "late peaking." The peak of the pitch contour normally occurs in the middle of the syllabic crest; in some instances, however, it occurs late in the vowel, or even after it; and sometimes it occurs earlier than the middle. We have found in our corpus that late peaking generates an impetuous forward drive; in fact, the later the peaking, the more impetuous is the forward drive. This is also predicted by Gestalt theory.
Convergent and Divergent Delivery Styles
I suggested above that there is a "rhythmical performance," in which both metric pattern and linguistic stress pattern can be accommodated, such that both are established in the listener's perception. I am going to distinguish within the legitimate boundaries of "rhythmical performance" between "convergent" and "divergent delivery styles." Such convergence or divergence typically occurs between the metric and linguistic dimensions.
My work is governed by the "limited-channel-capacity hypothesis" of human information processing. Most of the available mental space may be allocated to one sequence of information processing. In order to make parallel processes possible, they must be allocated mental processing space at the expense of the "main sequence." The need to perceive the linguistic and the versification dimensions at the same time requires some cognitive manipulation for saving mental processing space. Training cannot increase "channel capacity." But the amount of information processed may be increased by efficient coding. Consider the following joke: "How do you feel, in one word?" "Good." "And in two words?" "Not good." For our purpose, the point is that "not good" could be recoded in one word, as "bad." In the reading of poetry such semantic recoding is inadmissible. Neither words, nor their order can be changed in poetry. But there is a possibility for saving mental space by grouping and clearly articulating the speech units. Mental processing space may be saved by the grouping of shorter words into a larger perceptual unit on the one hand, or by the over-articulation of phonemes and of word boundaries on the other. Consider the phrases "peace talks" and "pea stalks." By clearly articulating the word boundaries, one may save mental processing space required to infer from the context which phrase is intended. The linguistic sequence requires continuity; the over-articulation of word boundaries generates discontinuity. The resulting spare processing space is required for the simultaneous perception of regularly alternating weak and strong positions that constitute the metrical set, and of the immediately observable string of syllables where their stress pattern deviates from the versification pattern.
Gielgud vs Gielgud
As I said, this paper is focused on four delivery instances of a single verse line from a sonnet for which I had access to four recordings, two of which are by the same actor, sixteen years apart. (3) This allows me to consider significant differences between two delivery instances where certain other things are, still, equal--enabling me to set forth some of my distinctions with greater clarity and distinctness.
[TEXT NOT REPRODUCIBLE IN ASCII]
Two structural difficulties are built into this verse line, for which the reciter must find an elegant solution. The three syllables "heaven that" must be assigned, in one way or another, to two metrical positions. And the sequence "leads men" poses a difficulty for prosodists who believe in metrical and unmetrical lines. Paul Kiparsky, for instance, claims that strings of consecutive stressed syllables must end in an even-numbered (strong) position. Strings that end in a weak (odd-numbered) position render the verse line "unmetrical." The reason is that the rightmost member in the string bears the greatest stress; it may not therefore--thus the argument goes--occur in a weak position. We will consider how Gielgud and other reciters handle the two difficulties in their readings. (4)
[FIGURE 1 OMITTED]
In certain cases poets have license to assign two syllables to one position, that is, where meter requires only one syllable. Traditional metrics solves the problem by "eliding" a vowel. Generative metrists rightly claim that this solution does not work in many instances. They have discovered certain phonetically-justifiable rules that would render such verse instances "metrical". A metrical position, they say, may be actualized by two syllables under the following conditions:</p> <pre> I. Where the syllables consist of two adjoining vowels, irrespective
of word boundary, or where they are separated by a liquid or nasal or h or, from the early sixteenth century onwards, by a voiced fricative
(cf. Halle and Keyser 1966, 209; Freeman 1969, 197-98). II. "An unstressed or weakly stressed monosyllabic word may constitute a single metrical position with a preceding stressed or unstressed syllable." </pre> <p>[FIGURE 2 OMITTED]
(Halle and Keyser 1966, 212)
There is a traditional stock of bisyllabic words which frequently occur in a single position, as "even, heaven, seven, power, tower, spirit, devil, evil," etc. Some of these words cannot be pronounced as a single syllable, even though the editor may omit one of the vowels, as in "heav'n." It will be observed that all these instances conform with Condition I. And so does "many a," which most frequently occupies two positions in iambic verse, instead three.
However, they "say nothing ... about how this and similar lines are to be performed" (Halle and Keyser, 1966, 207). I claim that this license is not a mere arbitrary "allowable deviation," nor a matter of mere internalization of an abstract "rule." Certain pairs of syllables may constitute a single position precisely because they can, though need not, be performed in a certain way. The relevant phonemes: vowels, sonorants, h, and voiced fricatives, have a common feature of performance: continuity. The property shared by pairs of syllables like "heaven ... I have ... I am ... devil ... many a ... tower ... spirit" etc. is that there is no abrupt stop consonant between them and, except [TEXT NOT REPRODUCIBLE IN ASCII], the transition between vowels has an element of periodicity. The transition from one vowel to another may be smoother, without a clearly articulated boundary between the syllables; the boundary between the two vowels is fuzzy. I have suggested that in metrically complex lines there is sometimes a need for careful articulation in order for metrical positions to be perceived. In the present issue, the converse appears to be the case. If, e.g., "heaven" is emphatically articulated only at the end of the word, it will not indicate two metrical positions, although it is perceived as two syllables.
Examples from Anthony and Cleopatra and Hamlet suggest that there may be a third, altogether different kind of condition under which extrametrical syllables may occur in a line: when two syllables that are to be assigned to the same position are separated by a major syntactic boundary (usually at the caesura). In the line discussed here, the word "heaven" conforms to both Condition I and III. As we will see, a comparison of Gielgud's two readings suggests that there may be a substantial difference between them.
When I first listened to these two performances by Gielgud, I tried to get an overall intuitive impression of the difference between them, rather than analyze it. I had an unexplained impression that Gielgud 2 is much more complex, artistically more sophisticated, rhythmically more satisfying. The best way to characterize my impression of Gielgud 1 was, perhaps, by punning on the English idioms "flat-out" and "flat out." The former is usually used as an intensive, that is, a modifier that has little meaning except to intensify the meaning it modifies; the latter suggests "in a blunt and direct manner." Later, when I compared the two readings' handling of the complexities of the verse line, this intuitive contrast was amply accounted for.
If you encounter the stretch of language "To shun the heaven that leads men to this hell" in a prose utterance, it may be uttered as a single unit, or will at most be segmented into two segments, the relative clause, and what precedes it. Both readings of Shakespeare's verse are parsed into more segments. Now when you look at the wave plots and pitch plots extracted from the two readings, an immediately-perceived difference becomes conspicuous. In Gielgud 1 there are two huge pauses (after "shun" and "heaven"). In the wave plot extracted from Gielgud 2 no such pauses are visible. Discontinuation is achieved here by means other than straightforward pauses.
Measurements concerning the word "heaven" may tell us much of the story (Table 1). In Gielgud's both readings [TEXT NOT REPRODUCIBLE IN ASCII] is about 1.77 times longer than the combined duration of the preceding sounds. As I said earlier, there is no way to decide whether a given speech sound is longer or shorter than it ought to be. Recently I began to use the electronic version of The Merriam-Webster Collegiate Dictionary as a standard for comparison. In this dictionary the words are not only printed, but also spoken by either a male or a female speaker. These recordings give the "citation form" of the words, with professional clearly-articulated pronunciation. The word "heaven" is spoken by the female speaker. Here the duration of [TEXT NOT REPRODUCIBLE IN ASCII] is less than half of the combined duration of the preceding sounds. Obviously, in Gielgud's both readings the [TEXT NOT REPRODUCIBLE IN ASCII] is exceptionally lengthened. I will argue that this difference has far-reaching rhythmic and aesthetic consequences.
Subjectively, the prolongation of [TEXT NOT REPRODUCIBLE IN ASCII] is much more conspicuous in Gielgud 2 than in Gielgud 1. Indeed, it is insignificantly longer in Gielgud 2. But, as Table 1 shows, the two [TEXT NOT REPRODUCIBLE IN ASCII]s have roughly the same duration relative to the combined duration of the preceding speech sounds in the word. One important difference must be sought in the respective amplitude envelopes (6) of these [TEXT NOT REPRODUCIBLE IN ASCII]s. Consider Figure 3. In Gielgud 1, in the first three sounds of "heav'n" the amplitude curve juts out; the beginning of [TEXT NOT REPRODUCIBLE IN ASCII] still has considerable amplitude (the second "hill" of the curve), but then the sound gradually fades away. In Gielgud 2 the amplitude level is relatively even: the first syllable does not jut out; and the curve on [TEXT NOT REPRODUCIBLE IN ASCII] is higher than the bottom line, and does not fade away as in the other reading.
The prolongation of [TEXT NOT REPRODUCIBLE IN ASCII] has two quite different effects: it suggests what Gerry Knowles calls "segmental discontinuation"; and improves the articulation of the phoneme. Now the decaying amplitude in one reading lessens the effect of the articulation of [TEXT NOT REPRODUCIBLE IN ASCII], whereas its sustained amplitude in the other reading strongly reinforces it. With reference to bisyllabic occupancy of metrical positions I suggested that if "heaven" is emphatically articulated only at the end of the word, it will not indicate two metrical positions, although it is perceived as two syllables. One would expect that the need to squeeze two syllables into one metrical position requires the reciter to elide the [TEXT NOT REPRODUCIBLE IN ASCII] and shorten the [TEXT NOT REPRODUCIBLE IN ASCII] as much as possible. Gielgud does elide the [TEXT NOT REPRODUCIBLE IN ASCII], but considerably lengthens the [TEXT NOT REPRODUCIBLE IN ASCII]; this is, precisely, what the present theory predicts: over- rather than under articulation of the [TEXT NOT REPRODUCIBLE IN ASCII]. Listening to the two recordings confirms this speculation. The extrametrical syllable at the caesura is perceived as less disturbing in Gielgud 2 than in Gielgud 1. One of several reasons must be attributed to this over articulation of the word-final [TEXT NOT REPRODUCIBLE IN ASCII].
I have pointed out above another conspicuous difference between the two readings. In Gielgud 1 there is a huge 413-msec pause between "shun" and "the"; and an even longer, 503-msec pause between "heaven" and "that." Obviously, these function words are syntactically grouped forward, and the preceding pauses strongly underpin this forward-grouping. In Gielgud 2, by contrast, there is no measurable pause between these pairs of words. Notwithstanding this, one of my associates could hardly believe that there is no pause there in this reading. What is more, as Figure 2 shows, in Gielgud 2 the words "heaven" and "that" are uttered on one falling intonation contour, effectively grouping "that" backward rather than forward. It is the prolongation and overarticulation of the word-final [TEXT NOT REPRODUCIBLE IN ASCII] that bears all the burden of generating discontinuity at the caesura.
[FIGURE 3 OMITTED]
Perceptually, what happens here is quite sophisticated. We have got here conflicting cues for continuity and discontinuity. The shared intonation contour and the lack of pause groups the word "that" backward; the listener's syntactic knowledge and the sustained [TEXT NOT REPRODUCIBLE IN ASCII] indicating a rest suggest a new start after "heaven." Consequently, a caesura and a "metrical impulsion" across it are perceived at the same time. Levin (1971, 184-85) regards caesura as a metrical, not a linguistic fact. The line exerts pressure for completion upon which the caesura obtrudes. "If caesura is regarded as the syntactic pause or break, nothing is left to explain the required sense of metrical impulsion across that break" (185). I said above that in Gielgud 1, by contrast, I intuitively felt that the same syntactic juncture was thrust upon the reader "in a blunt and direct manner" and, at the same time, acted as a modifier that has little meaning except to intensify the meaning it modifies. This happens because the grouping cues cluster differently, displaying great redundancy. The beginning of a relative clause constitutes a major syntactic juncture. This is reinforced by an unusually long pause, and is further reinforced by the prolongation of the word-final [TEXT NOT REPRODUCIBLE IN ASCII]. It is this over articulated syntactic juncture that confirms a prosodic event--a caesura. In terms of our initial stylistic distinction, in Gielgud 1 the cues act in convergence, in Gielgud 2 in divergence.
Let us go back to the three conditions suggested above under which a metrical position may be actualized by two syllables. The three syllables "heaven that" must be assigned to two metrical positions. This instance happens to conform to all three conditions, leaving the performer with a margin of freedom to choose. I have claimed that the three conditions imply different kinds of vocal performance. So far I treated the three kinds of vocal performance as more or less equal. But, as I said above, in Gielgud 2 the solution sounds somehow more acceptable than in Gielgud 1. The reason seems to be that the two readings realize different conditions. Gielgud 1 separates the syllables "-en" and "that" by a long pause, so that the two may be perceived as belonging to different units, preventing conflict (Condition III). I pointed out above that in Gielgud 1 the amplitude of [TEXT NOT REPRODUCIBLE IN ASCII] gradually fades away, whereas in Gielgud 2 it is sustained (see Figure 3); owing to this, in the latter the [TEXT NOT REPRODUCIBLE IN ASCII] is more salient and therefore more effective as a cue for discontinuation. This legitimizes Gielgud 2 under Condition I. But this verse line can be legitimized under Condition II as well: "that" is an unstressed monosyllabic word with a reduced vowel. What kind of performance would this Condition require? In this reading, the word "that" is conspicuously under-articulated. As Figure 2 clearly shows, it is also emphatically grouped backward by two means: there is no measurable pause between the two words; and they are uttered on what Gerry Knowles calls an "internally defined" intonation curve. This is followed by "pitch discontinuation" (a sudden change in [F.sub.0]). The pitch on "heaven" falls from 116.667 Hz to 97.566 Hz; then, on "that", further from 87.154 to 78.750Hz. Here the pitch curve changes direction and resets to 89.271 Hz, continuing upward to 102.558 Hz; then changing direction again, it descends to 97.137 Hz.
The change of direction and the sudden leap of pitch indicate "pitch discontinuity," that is, a tone-group boundary. Thus, an exceptionally complex structure is generated: the three words "heaven that leads" are continuous, there is no measurable pause between them; "that" is grouped backward, but is part of the ensuing syntactic unit; at the same time, this stretch of words is parsed into three distinct units by powerful cues of discontinuity. Thus, the grouping of "that" becomes somewhat ambiguous. It is both preceded and followed by cues of discontinuity, and the perceiver has some freedom to group it either forward or backward, actualizing either Condition I or Condition II of bisyllabic occupancy. In either case, conspicuous discontinuity is generated, with a powerful impulse across it. However, as with Jastrow's notorious duck-rabbit, the two possibilities cannot be realized simultaneously. It is a matter of what Wittgenstein called "aspect switching." This possibility to switch from one solution to another will be more conspicuous if we realize that in The Marlowe Society's delivery instance there is only one possible solution. Here too "that" is grouped backward, but the [TEXT NOT REPRODUCIBLE IN ASCII] is relatively short. Tone-group boundary is indicated only by the resetting pitch on "leads."
[FIGURE 4 OMITTED]
Two differences between the falling intonation contours on "heaven" seem to be quite obvious, one measurable, the other perceptual. First, in Gielgud 1 the falling intonation contour of "heaven" is much longer than in Gielgud 2: it begins at 127.457 Hz (much higher than the other reading), and falls to 77.915 Hz, slightly below the bottom line of "that" in the other reading. Second, concomitantly, the falling curve in Gielgud 1 arouses a feeling of "homecoming," whereas in Gielgud 2 there is a feeling that the curve fails to reach the point of rest, demanding completion. What is more, the [TEXT NOT REPRODUCIBLE IN ASCII] is prolonged at this unsatisfactory point, generating a sense of arrest and a sense of impulsion across it at the same time. In Gielgud 1, by contrast, the "homecoming" of the falling intonation contour coincides with a major syntactic juncture, and the beginning of a longish pause. Exceptionally great stability is achieved.
A similar story can be told, mutatis mutandis, about the sequence "shun the." In Gielgud 1 there is a longish pause between them (363 msec); in Gielgud 2 they are run one into the other. In both readings [TEXT NOT REPRODUCIBLE IN ASCII] is considerably longer than the combined duration of the preceding sounds; in Gielgud 2 its relative duration is insignificantly longer than in Gielgud 1. In The Merriam-Webster Colleagiate Dictionary, by contrast, [TEXT NOT REPRODUCIBLE IN ASCII] is considerably shorter. As after "heaven," after "shun" too Gielgud 1 resorts to redundant cues: the prolongation of [TEXT NOT REPRODUCIBLE IN ASCII] reinforces discontinuity that is also signified by a longish pause; whereas in Gielgud 2 it indicates discontinuity where the two words are run one into another.
In Callow's and The Marlowe Society's readings (Figures 7-8), too, "shun" is run into "the." In the latter, intensity and the rising-falling intonation lend exceptional accent to "shun," but there is no discontinuity here. In the former, the 22-msec-longer [TEXT NOT REPRODUCIBLE IN ASCII] and the falling terminal contour generate some discontinuity, placing it, in this respect, between Gielgud 2 and The Marlowe Society.
There are good prosodic and syntactic reasons (caesura and syntactic juncture) to indicate discontinuity after "heaven," with or without a pause. After "shun" it has neither syntactic, nor prosodic justification. It seems to be gratuitous--unless it has some rhetorical or paralinguistic justification. It is, perhaps, a prosodic mannerism.
[FIGURE 5 OMITTED]
William Benson (2003, online) suggests that a single line of poetry can readily encompass one or two intonation units. One striking thing about Gielgud's two readings of this line is that he fragments the verse lines into at least three sentence fragments (sentence fragment is a word, phrase, or clause that usually has in speech the intonation of a sentence but lacks the grammatical structure usually found in the sentences of formal and especially written composition). This is a salient feature of poetry reading in general: stretches of speech are broken up into smaller than usual intonation units, so as to exert rigorous control over every syllable and metrical position. This may have emotive effects too. The pauses in Gielgud 1 may be perceived as a source of uncertainty, of lack of control. The reading in Gielgud 2 is divided into no fewer chunks, but by other means than pauses. Consequently, it is not perceived as hesitant; on the contrary rather, as a stretch of speech every bit of which is under strict control.
Toward the end of Figures 1 and 2 the pitch contour is poor or incomplete. In Gielgud 1 one clearly hears a rise of pitch on "men" after "leads." But in Figure 1 at this point the speech signal is so poor that the graph shows no trace of it. So, in Figures 5-6 I am providing some of the missing information by an alternative method of pitch extraction.
In Figure 5, the intonation contour clearly juts out on "men." Indeed, "men" bears greater stress than "leads"; but in this verse line, it is "men" that occurs in a weak position, infringing upon metric regularity. In the four readings under discussion we encounter three different ways of handling this problem. Callow grossly under stresses "men," suppressing the stress pattern required by "prose rhythm." Gielgud 1, by contrast, duly stresses "men," infringing upon the stress pattern required by meter. Between these two extremes we find Gielgud 2 and The Marlowe Society. In Gielgud 1 the rising-falling pitch contour juts out on "men," cuing a stress that is stronger than the preceding one and, indeed, metric fluency is felt to be damaged. In Gielgud 2, by contrast, the intonation contour assigned to "men" is somewhat lower than that on "leads," but has similar shape. Thus, in this reading the stress of "men" is subordinated to that of "leads," satisfying the metric requirements. At the same time, the duration of "men" in Gielgud 2 is longer than in Gielgud 1, long enough to break somehow even with the stress on "leads." Thus it compensates for the lower pitch, just enough to satisfy the syntactic demand for a stressed syllable. The salient early peak on "men" effectively groups it with "leads" (in a strong position). The same holds true, with the necessary changes, of Marlowe Society's reading. The perceptual process of "breaking even" by lengthening a weakly stressed syllable can be observed in action in my attempt to electronically lengthen "men" in Callow's reading (see below). Consider the relative durations of "men" in Table 3. In Gielgud 1 pitch contour cues stress, foregrounding its perceptual deviance; so, "men" is relatively short. In Gielgud 2 and The Marlowe Society's reading it is duration that takes care of relative stress. In Callow's reading "men" is shorter than in the other ones, and sharpens (rather than levels) the difference between the two consecutive stresses.
[FIGURE 6 OMITTED]
The present work assumes that clear-cut articulation saves mental processing space required for the simultaneous perception of the conflicting patterns of stress and meter. Rising-falling or steeply-falling intonation contours and relatively long speech sounds (as in "men" in Gielgud 2 and The Marlowe Society's reading) both contribute to such clear-cut articulation. I claim that artificial lengthening of the speech sounds of "men" in Callow's reading too (see below) improves articulation and contributes to the same effect.
[FIGURE 7 OMITTED]
Machine vs. Callow
I suggested above that the sequence "leads men" poses a difficulty for such generative metrists as Paul Kiparsky or Gilbert Youmans (1989), who claim that strings of consecutive stressed syllables must end in an even-numbered (strong) position; strings that end in a weak (odd-numbered) position render the verse line "unmetrical." The reason is that the rightmost member in the string bears the greatest stress; it may not therefore occur in a weak position. Two of the four recordings solve this problem quite elegantly. Gielgud 2 (Figure 6) and The Marlowe Society (Figure 7) clearly articulate the boundaries of the two consecutive monosyllables by assigning them, each, a rising-and-falling intonation contour. This renders both consecutive syllables stressed, though with less perceived stress on "men" than on "leads." The clear articulation of these words allows perceiving the respective metrical positions as well. In Gielgud 1 too these words are assigned similar intonation contours; but that on "men" is somewhat higher, and the word is perceived as over-stressed, disrupting to some extent the rhythmic flow. In Callow's recording we encounter serious trouble with these two words. "Leads" (in a strong position) bears strong stress cued by a rising-falling intonation contour; "men" is perceived as insufficiently stressed, in spite of the long-falling contour. Furthermore, as Figure 8 shows, the intonation contour of "leads" constitutes an exceptionally late peak. This has two perceptual corollaries: it increases the stress; and generates a forward-pushing "perceptual force," leaning on the insufficiently stressed "men" (in a weak position). As a result, the insufficient stressing of the latter is even more strongly felt. I had a methodological problem here: from the wave plot and pitch extract in Figure 8 I could not predict this rhythmic flaw. Measurements are of limited help here ("men" in this reading is considerably shorter than in the other readings or in the Dictionary). The fault was salient for my ears. When listening one after the other to the phrase "leads men" excised from the four readings, it is felt that, in three of them, "leads" exerts a "perceptual force" on "men"; whereas "men" strikes a delicate balance: it is sufficiently stressed to "withstand" this pressure, but is not stressed too much to infringe upon meter. In Callow's reading, the forward push on "leads" is amplified by late peaking; whereas the power of "men" to stand firm against the pressure is strongly reduced.
As I said above, measurements are of limited help in accounting for the rhythmic deficiency of Callow's reading. So I tried to see, by trial and error, what happens when certain elements are manipulated. Pitch, duration and amplitude (in this order of decreasing effectiveness) are variables that crucially influence perceived stress (Fry 1958). I can control only duration in a tolerably accurate manner; but in this respect, too, there are severe limitations. In such continuous consonants as [TEXT NOT REPRODUCIBLE IN ASCII], [TEXT NOT REPRODUCIBLE IN ASCII], [TEXT NOT REPRODUCIBLE IN ASCII], [TEXT NOT REPRODUCIBLE IN ASCII] I can copy a small portion and then paste it again and again into the signal, without considerably distorting speech. But when I come to do the same with vowels, pitch and vowel quality too may be affected, generating unnatural intonation contours and emotional qualities. As to "men," I had little difficulty to lengthen the [TEXT NOT REPRODUCIBLE IN ASCII] and the [TEXT NOT REPRODUCIBLE IN ASCII]; but eventually I succeeded to lengthen the vowel between them too, with minimum distortion (there is some distortion: one of my informants commented that it "provides a false emotional emphasis to the line"). The reading labeled "Callow 2" is the doctored version. By lengthening the three speech sounds of "men," its ability to "exert force" in opposition to the thrust of "leads" is boosted. This is an ostensive way to support my argument, by changing features supposed to generate the quality under discussion, and exposing the resulting quality to the listening ear.
There is a similar problem with the last two words. Presumably for rhetorical reasons, Callow assigns a strong emphatic stress to "this," disregarding the damage done to rhythm. Again, there is a late peak in "this," pressing forward, while the stress assigned to "hell" is insufficient to "withstand" this pressure. And again, the controlled lengthening of the last two speech sounds of "hell" by "copy-and-paste" caused it to stand firm against the thrust of the overstressed "this." Incidentally, the intonation contours generated in these specific instances do approximate the contours required to solve this rhythmic problem.
[FIGURE 8 OMITTED]
[FIGURE 9 OMITTED]
A research like the present one must be guided by the researcher's intuitive perceptions regarding rhythmical and unrhythmical performances. At a certain point I had to ascertain whether other people had similar perceptions to mine. I needed informed informants. But also I had to request them to try to suspend their professional knowledge, and give as spontaneous and intuitive responses as possible. I made Gielgud's readings available on a webpage (later I added Callow's genuine and doctored versions), and asked some of my colleagues to respond to them. I also sent a request to the PSYART and Coglit lists. I received all in all five fairly detailed responses, from three literary theorists, one music theorist, and a practicing psychoanalyst who had written on Sonnet 129, as well as a brief interjected preference judgment. In such circumstances it is impossible to make a quantitative study, and I will have to weight the answers from a qualitative point of view. It is difficult to compare answers to open-ended questions, when the various informants use different metaphors to describe their impressions. Moreover, as we shall see, the same informant may provide conflicting descriptions.
All informants thought that both readings are within the boundaries of "rhythmical performance," though most informants thought that the first one is nearer to the minstrel's pole, the second one to the actor's pole. One informant made a more sophisticated judgment. Speaking of "the quavering voice on 'heaven' and 'leads'" in the second reading, he remarks: "I'm not sure, but that seems to push the first more toward the minstrel pole and the latter more toward the actor pole, but then the first version overstresses 'shun,' which seems 'actorly' to me. I suppose there are some ways in which each is minstrel-like and some ways in which each is actor-like."
All informants pointed out that the two readings had different rhythmic structures. The most accurate description of this difference runs as follows: "Overall, the rhythmical character of the first version is marked by the emphasis on and pauses after 'shun' and 'heaven,' whereas the relative lack of pausing and the holding of the reverberations of final consonants so that they flow into the next word makes the second version more even and sonorant." Another respondent put it:</p> <pre> I suppose there are some ways in which each is minstrel-like and some ways in which each is actor-like. The readings ... differ on the caesura and on the degree of stress. The stresses are more even in the second (which, I suppose, is a minstrel-like trait). In the first, "shun" has a very strong stress; "hell" somewhat less;
"heaven" and "leads" less still. But they are similar in that the
same words get stressed. </pre> <p>These two passages, when juxtaposed, suggest that the respective convergent and divergent nature of the two readings is determined not so much by making the stress pattern conform to the metric pattern, but rather by their handling of segmentation. In reading 1 the pauses converge with the other cues for segmentation. In reading 2, by contrast, the "final consonants ... flow into the next word", and segmentation is generated by conflicting cues. Most informants, however, can only report a general impression, and cannot point out such evasive details. Now consider "the stresses are more even in the second." As far as the "more even" stresses occur in strong positions, as the examples enumerated above, it is, indeed, "a minstrel-like trait." But when it is the consecutive stresses "leads men" that are more even, it is a paradigmatic "divergent" trait.
Another informant uses a simile: "The first is like a lullaby that 'lulls' the listener." Oddly enough, this is the person who felt that the first reading has "ominous rhythm" (see below). I don't consider this as sheer logical contradiction. I construe it, rather, as an expression used to describe a quality in the performance for which the writer had no technical term: that she felt a "convergent" quality in the delivery style. This conjecture is supported by her comment "Reading No. 1 approximates Style 1 [i.e., the Minstrel] (although I would not call it "wooden"!) Reading No. 2 approximates Style 2 [i.e., rhythmical performance]; clearly metrical, but an elision/syncopation that compresses the phrase for a slight rhetorical emphasis"
Most informants agreed that the first reading is more emotional. They differed, however, in their evaluation of this emotionality, as well as on its relation to other elements in the performance. Two of them spoke of "ominous rhythm" and "awesome emotional tone," respectively. They probably referred to the same perceived quality. In these comments, the emotional quality was implicitly judged as a "good-making" feature of the performance. One person, who happened to listen to the two readings (a hi-tech engineer) squarely declared that he definitely preferred number 2, because the other was "full of pathos." This person agrees, then, with the others as to the presence of the emotional quality, but makes an opposite value judgment. By the way, I happen to sympathize with his judgment.
The two informants who agree on this reading's "ominous" or "awesome" quality, attribute it to different sources. One attributes the awesome emotional tone to a "resonant" voice quality and to fragmentation of the line by pauses. The other spoke of "ominous rhythm." But in the same sentence she also mentions the "rounded, velvety timbre of the first reading"; and later she uses the phrase "the more 'stilted' first version." (7) They both seem to refer, by their different terms, to fragmentation by pauses and to a surrounding aura of rich overtones. As to the source of the emotional quality, I tend to agree with the former informant. Resounding ("thundering") voice is regularly associated with emotions. Likewise, "the discontinuity of the first reading contributes to its awesome tone, because it inspires the listener with uncertainty. Concomitantly, it displays the moment-to-moment fluctuation of the living voice." So, one must distinguish between two mental acts: discerning the elements and perceived qualities of a verse line on the one hand, and, on the other, assuming a causal relationship between the two. Even if the source of the perceived qualities is displaced, the very discernment of elements and effects may still be sound.
Most informants preferred version 1. The reason for this preference is its more emotional quality. Though I was asking about "rhythmical performance," my informants could not keep it apart from the emotional element. As far as pauses are concerned, the same element is an exponent of both the rhythmic and emotional quality. But the "surrounding aura of rich overtones" (or "rounded, velvety timbre of the first reading") has everything to do with emotional qualities, and nothing with rhythmic qualities. The two informants who detected the "ominous" or "awesome" quality in reading 1 also concur in their preference for the first reading, and both continue with a "but." They seem to feel apologetic for not preferring the second reading, where the rhythmic solution is more conspicuous (e.g., "but have to admit"). In all these responses preference for emotional qualities slightly overrides the appreciation of rhythm. One informant, however, considers rhythm an outright obstacle: "I thought the rhythmic of number one was subsumed to the acting so that it might have easily gone unnoticed to my unpracticed ear. My attention was called to the rhythm of number two and it was somehow distracting and thereby detracting." This informant is the only one in the emerging sample who has no professional interest in poetic or musical rhythm. Indeed, he mentions his "unpracticed ear" with so many words. Yet, he noticed the more compelling rhythmic nature of number two; but for him rhythm is "somehow distracting and thereby detracting." Thus, on the descriptive level he seems to be in agreement with the earlier-mentioned two informants, but on the evaluative level they are diametrically opposed. The present paper, however, is about the perception of poetic rhythm, not about its evaluation; and what he perceives seems to be in harmony with my predictions.
One of the questions I asked was "Does the rhythmic structure of one or the other delivery instance interact in any way with the emotional tone of the line?" Figure 5 shows on "hell" a falling, classical terminal contour, appropriately indicating the end of an utterance (and of the poem). Two of my informants referred to this issue in almost identical terms. Both respondents discern this intonation contour, and offer similar interpretations. One suggests: "in reading 1, 'hell' marks the end of a long stepwise descent: as the lowest pitch, it functions therefore as a sonic metaphor, and the phrase mirrors the slow descent to hell that such abandonment to pleasure might occasion." The other is more cautious. She too regards it as a metaphor suggesting descent, but provides no specific details that cannot be substantiated or refuted. So, this may be regarded as an instance of what I elsewhere called "triple encodedness" (Tsur 2000; 2002). The same falling contour indicates the end of a syntactic unit (sentence), of a rhythmic unit (the line); at the same time, it is expressive of the long fall. Likewise, the pause after "heaven" indicates a (syntactic) clause ending, a (prosodic) caesura, and an emotional quality.
I have earlier mentioned a comment that the first reading is like a lullaby that "lulls" the listener, and thus ironically points up the "heaven that leads to hell," whereas the compressed rhythm and rising tones of the second reading is all nervous "shunning" with no heavenly contrast. Here we encounter one of the widespread critical fallacies. The writer uses the term "lullaby" to designate a rhythmic quality, and then is carried away by her own expression, regarding it as ironical to the "heaven that leads to hell." Had she used the term "obtrusive rhythm" instead "lullaby," no irony would have arisen. On the other hand, I am wondering how much of the description "the compressed rhythm and rising tones of the second reading is all nervous 'shunning'" refers, in fact, to what I was referring to in my analysis of the cues for continuity and discontinuity in the second reading. At this point, I should mention again a comment by another respondent: "The discontinuity of the first reading contributes to its awesome tone, because it inspires the listener with uncertainty. Concomitantly, it displays the moment-to-moment fluctuation of the living voice." In this respect, interaction appears to be genuine.
Another question I asked was "Can you point out specific problems in the verse line, and the vocal manipulations by which Gielgud solved or failed to solve them?" In my analysis I had pointed out two difficulties in this verse line: the three syllables "heaven that" must be squeezed into two metrical positions, and the succession of stressed syllables "leads men" ends in a weak position. Only one respondent referred to both issues: the two readings "are similar in that the same words get stressed, and in both cases 'heaven' is shortened to a monosyllable. The result is that in both cases he makes the line a little more regular. Of course, he does not treat 'men to' as an iamb (which would be the sing-song version); in that sense, both are more actor-like". There is one more reference to each one of these issues, by different informants: "an elision/syncopation that compresses the phrase for a slight rhetorical emphasis"; and "The problem I see in the line is the falling of the main stress on 'men' on a W position. Metrically, this is balanced by 'leads' on the sixth position, which keeps it metrical according to the Halle-Keyser system." As to elision, the computer makes it quite clear that in Gielgud's and The Marlowe Society's readings there is no schwa in "heaven," whereas in the dictionary and in Callow's reading there is one. These three references to the two issues identify the problems and vaguely hint at the kind of solution, but I hoped also for acknowledging the differences in handling them. But that would, perhaps, be too much to ask. The reference to the Halle-Keyser theory points up the problem. According to the original version of that theory, ending the phrase "leads men" in a weak position would render the verse line unmetrical (just as in Kiparsky's theory). Therefore, the authors later revised the theory, so as to legitimise such instances, and reduce the enormous number of unmetrical lines that would result. I claim that it is such performances as Gielgud 2 or The Marlowe Society's that render such verse lines acceptable to the ear. What you hear did not change with the change of theory; it is only the abstract rule that has changed. (8) The present suggestion is that it is the performance that makes all the difference. If "men" receives its due linguistic stress, the verse line sounds unrhythmical; if it is unstressed, the iambic singsong is preserved, but the rhythm becomes childish. I claim that in the second reading Gielgud drastically lowers pitch, but compensates for this by lengthening "men." To have one's cake and eat it. In his first reading pitch juts out on "men" (see Figures 5-6).
The only response I received on the Callow manipulations was from this informant. "The emphasis on 'men' ... is balanced by that given to 'leads' (in strong sixth position). The second version, in emphasizing 'men' on the devil's note, for me throws off the rhythm (and provides a false emotional emphasis to the line)." Here, too, she implicitly invokes the Halle-Keyser theory ("is balanced by that given to 'leads'") to legitimize the deviant stress. To my great dismay, she perceives the first (genuine) reading as more rhythmical. I still feel that in Callow's genuine reading "men" is under stressed, and just like in Gielgud's second reading, here too the lengthening of "men" strikes a precarious balance between stress and unstressed. I plead, however, guilty of "providing a false emotional emphasis", which I could not avoid under the circumstances. (9)
[FIGURE 10 OMITTED]
One informant gave his overall impression in an informal tone: "I didn't care for the caesura in the first version, but I didn't care for the quavering voice on 'heaven' and 'leads' in the second." I thought this overall impression may support some of my subtler analyses. He later expressly confirmed this. Here is what I wrote to him:</p> <pre>
When you say "I didn't care for the caesura in the first version,"
I think you confirm my impression that the caesura is over- articulated by a syntactic juncture + long-falling intonation contour + a long pause + the exceptionally long [TEXT NOT REPRODUCIBLE IN ASCII]. I feel, therefore, that it is thrust on the reader "in a blunt and direct manner." When you say you "didn't care for the quavering voice on 'heaven' and 'leads' in the second," you draw attention to something I haven't noticed. In this reading there is no measurable pause at all, the function word "that" is grouped backward to "heaven" by intonation; so, it is the prolonged [TEXT NOT REPRODUCIBLE IN ASCII] that carries all the burden of discontinuity. In absolute terms it is only slightly longer than the [TEXT NOT REPRODUCIBLE IN ASCII] in the first reading; relative to the duration of the whole word, however, it is even insignificantly shorter. But in the first reading the amplitude gradually decays, whereas in the second it is sustained. Owing to this, the [TEXT NOT REPRODUCIBLE IN ASCII] in the second reading is more salient (and therefore more effective in generating discontinuation). Now you have pointed out that the [TEXT NOT REPRODUCIBLE IN ASCII] is quavering too; this, in my opinion, increases its salience. At the same time it conveys an emotional quality which may be felt to be overdone, and therefore offensive". Again, this may be regarded as an instance of "triple- encodedness. </pre> <p>Finally, in my first book on meter (Tsur 1977, 22) I made my distinction between three delivery styles as follows:</p> <pre> As for performance, we may observe, roughly, three "delivery styles." At one extreme we find the singsong of children chanting their nursery-rhymes and the mechanical scansion of "freshmen," suppressing "prose-rhythm"; at the other extreme, we find the prosaic rendering of readers "insensitive to rhythm," and of actors who wish to make poetic drama sound more "natural," suppressing meter. In the middle, there is a complex rhythmical performance that avoids the suppression of either stress pattern or metric pattern. It may be possible to do justice to the conflicting patterns by accommodating them in a third, superimposed pattern of performance." </pre> <p>In my 1998 book I preferred to quote the Princeton Encyclopedia (Preminger and Brogan 1993) which, in turn, invokes C. S. Lewis. That was also the definition of "rhythmical performance" which I presented to my informants. Now I had to realize that though the two formulations refer to the same distinction, had I resorted to my first formulation, I may have obtained quite different results. The point is that in my original formulation I give a more purely rhythmical definition. Lewis's terms introduce the noisy elements "wooden" and "flamboyantly expressive." These elements had a negative and a positive effect on the results. On the one hand, they diverted the respondents' attention from the rhythmic structure to other aspects. On the other hand, they prompted comments on the emotional impact of the readings, and indications of the "triple-encodedness" of phonetic cues.
Some Methodological Conclusions
Analytic philosophy offers a meta-language required for handling such a welter of responses. Analytic philosophers as Beardsley (1958) and Margolis (1962) distinguish between descriptive, interpretive and evaluative statements. Sibley (1962) distinguishes aesthetic qualities in a work of art (referred to by "aesthetic concepts"), and nonaesthetic features that may "count toward" or "against" some aesthetic quality. Zemach (1976) points out that "aesthetic concepts" have both a descriptive and an evaluative component; the greater the share of the descriptive component, the smaller is that of the evaluative component, and vice versa. According to Beardsley, this evaluative component can be supported by reference to three General Canons of Evaluation: unity, complexity and some intense human quality. Another key expression would be "perceiving as" in the Wittgensteinean tradition. One may detect a resonant voice quality in a delivery instance, which may be perceived as an "emotional quality"; this, in turn, other things being equal, may be judged as a good-making feature of the aesthetic event, under the Canon of intense human quality. In describing this emotional quality, one may resort to aesthetic concepts that have more specific contents, such as "ominous" or "awesome." Other things, however, need not be equal. The term "full of pathos" too refers to some (perhaps the same) emotional quality, but has little specific descriptive contents and much evaluative contents. This term suggests that the aesthetic event does have an intense human quality, but little complexity or unity. According to Morris Weitz (1962), statements in aesthetic discourse are not factual statements, but crucial recommendations what to look for in aesthetic objects, and how to look at it.
English meter baffles scholarship. Renaissance poets who laid its foundations thought they were doing one thing but were, actually, doing another. All researchers are in agreement on one point: that this practice had admirable results. Prosodic research in the past four centuries attempted to discover the principles governing it. But all would-be explanations ran into compelling counterexamples. Most notably, Milton and Shelley, the poets most praised for their musicality, violated all the proposed rules. Wellek and Warren attempted to find method in this madness by breaking down the rhythmic process into three dimensions: prose rhythm, meter, and performance. In my recent work I have been engaged in an instrumental investigation of "rhythmical performance," of which I conceive as of a problem-solving activity. Conflicting patterns of prose rhythm and meter must be accommodated in a pattern of performance. I soon had to realize that "rhythmical performance" is not a unitary phenomenon: there are different legitimate delivery styles. In this article I set out to propound the linguistic, metric, phonetic and cognitive elements involved in the rhythmical performance of poetry, suggesting that in different delivery instances these elements may constitute different clusters, yielding different delivery styles. I started with C. S. Lewis's distinction between the minstrel and the actor, suggesting that between them one may locate "rhythmical performance." Within rhythmical performance I distinguished "Convergent" and "Divergent" delivery styles. The performance Gielgud 1 tends toward the former, Gielgud 2 toward the latter. In this article I have offered basic elements and general principles rather than a systematic exposition of specific delivery styles.
At a certain point I could not avoid ascertaining whether other listeners perceive the same aesthetic qualities as I do in the aesthetic events discussed. I had to look for informants whose professional training enabled them to understand that they were requested to respond to certain aspects of the aesthetic event. This involved me in another difficulty: such informants may be guided by their theoretical knowledge. So, I made efforts to prevent their responses to be affected either by their own theories or the experimenter's presumed expectations. I tried to elicit from my respondents responses as spontaneous as possible. The result was that only very few persons responded to my request; and those who responded provided responses that overlapped in some respects, but were at variance in others. In the last section of my article I tried to sort out what possible perceptions did those responses report, and what kinds of metalinguistic statements they were.
The aesthetic event of poetry recital is very complex. The responses I received suggest that in some instances my informants responded to different aspects of the same event. The event may be consistently and quite thoroughly described by the tools offered here. One must, however, realize that in a poll like the one reported here all the respondents will respond only to a small subset of aspects. Thus, even widely different responses may be consistent with one another. In the preceding section I made an attempt to point out how the different responses relate to the same underlying comprehensive description.
This procedure taught me the limitations of what I was doing. I cannot claim that my perceptions, which guided my investigation, are the correct ones, or shared by most people. Nor were the perceptions reported by any of my respondents. But this does not mean that "anything goes." Some of my respondents, at least, did draw attention to some significant aspect of the aesthetic event, whether an aesthetic quality, or some set of features that generate it, or an evaluation of the resulting effect. The most I can do in these circumstances is to make a crucial recommendation as to what to look for in the rhythmical performance of poetry and how to look at it.
(1) Consequently, the present theoretical part of this paper is heavily drawing upon that book. The sound files for this article are available online: http://www.tau.ac.il/~tsurxx/Gielgud_Experiment/Listener_Response.html
(2) Glottal stop is the speech sound inserted before "aim" when saying "I said 'an aim,' not 'a name'." Stop release is the movement of one or more vocal organs in quitting the position for a speech sound. It is perceived as a "click." In the dictionary recording of "bright" there is an exceptionally salient stop release after [TEXT NOT REPRODUCIBLE IN ASCII]; in "brighten" there is none. Listen online to these two words, and then the isolated stop release.
(3) I am going to explore four readings of this verse line: by The Marlowe Society, Simon Callow, and two by Sir John Gielgud. Both of Gielgud's recordings are by Caedmon, in 1963 (the reading labelled "Gielgud 2") and 1979 ("Gielgud 1"). The discrepancy between the ordinal numbers and the order of recordings reflect my well-reasoned but mistaken assumption that the more complex aesthetic conception must be the later one.
(4) Readers who find it difficult to work through the following technical analyses may skip to the treatment of listener response in "Flesh-and-Blood Listeners."
(5) The lower window presents the wave plot display which shows a plot of the wave amplitude (in volts) on the vertical axis, as a function of time (in milliseconds) on the horizontal axis. The upper window presents a fundamental frequency plot, which displays time on the horizontal axis and the estimated glottal frequency (F0 = pitch) in Hz on the vertical axis.
(6) "Amplitude" is the acoustic correlate of what appears to consciousness as "loudness," that is, the magnitude of the auditory sensation produced.
(7) I asked her what did she mean by "stilted." The dictionary defines "stilted" as "1. lacking fluency in being halting or unnatural in flow 2. pompous or unduly formal." She answered that she had meant only the first meaning, and in an objective, descriptive, non-evaluative sense. "Reading 1 may indeed be considered as referencing definition 2, but I was assessing only the 'rhythmical' aspects of the reading." The person who thought it was full of pathos would have agreed, presumably to the second meaning as well.
(8) This appears to be a conspicuous instance in which the informant's response reflects her theoretical knowledge rather than what she hears.
(9) Fortunately enough, there was a delay between the completion of this article and its publication, during which I had access to the speech analyser "Praat." This application offers features for manipulating pitch and duration, yielding relatively natural results. This may provide less "false emotional emphasis" for the manipulated words in Callow's reading. I have manipulated only the duration, not the pitch of the words "men" and "hell," so as to leave all other things equal. Notwithstanding this, the change of speed did affect the voice quality of these words to some extent. Listen online to a version produced on a combination of two speech analyzers, Praat and SoundScope (Callow 3).
(10) Listen to them online.
Beardsley, Monroe C. 1958. Aesthetics: Problems in the Philosophy of Criticism. New York and Burlingame: Harcourt & Brace.
Freeman, Donald C. 1969. "Metrical Position Constituency and Generative Metrics." Language and Style 2: 195-206.
Fry, D. B. 1958. "Experiments in the Perception of Speech". Language and Speech 1: 126-51.
Halle, Morris, and Samuel Jay Keyser. 1966. "Chaucer and the Study of Prosody." College English 28: 187-219.
______. 1971. English Stress: Its Growth and Its Role in Verse. New York: Harper and Row.
Knowles, Gerry. 1991. "Prosodic Labelling: The Problem of Tone Group Boundaries." In English Computer Corpora. Selected Papers and Research Guide. Topics in English Linguistics. Vol. 3, ed. Stig Johannson and Anna-Brita Stenstrom. Berlin: Mouton de Gruyter.
Knowles, Gerry. 1992. "Pitch Contours and Tones in the Lancaster/IBM Spoken English Corpus." In New Directions in English Language Corpora-Methodology, Results, Software Developments, ed. Gerhard Leitner. Berlin: Mouton de Gruyter.
Levin, Samuel R. 1971. "The Conventions of Poetry." In Literary Style: A Sumposium, ed. Seymour Chatman. London and New York: Oxford University Press.
Margolis, Joseph. 1962. "The Logic of Interpretation." In Philosophy Looks at the Arts: Contemporary Readings in Aesthetics, ed. Joseph Margolis. New York: Scribner.
Preminger, Alex, and T. V. F. Brogan. 1993. The New Princeton Encyclopedia of Poetry and Poetics. Princeton: Princeton University Press.
Sibley, Frank 1962. "Aesthetic Qualities." In Philosophy Looks at the Arts: Contemporary Readings in Aesthetics, ed. Joseph Margolis. New York: Scribner.
Tsur, Reuven. 1977. A Perception-Oriented Theory of Metre. Tel Aviv: The Porter Israeli Institute for Poetics and Semiotics.
______. 1998. Poetic Rhythm: Structure and Performance--An Empirical Study in Cognitive Poetics. Bern: Peter Lang.
______. 2000a. "The Performance of Enjambments, Perceived Effects, and Experimental Manipulations." PSYART: A Hyperlink Journal for the Psychological Study of the Arts. http://www.clas.ufl.edu/ipsa/journal/articles/psyart2000/tsur04.htm
______. 2000b. "Phonetic Cues and Dramatic Function--Artistic Recitation of Metered Speech." Assaph--Studies in the Theatre: 173-96.
______. 2002a. "Aspects of CognitivePoetics." In Cognitive Stylistics--Language and Cognition in Text Analysis, ed. Elena Semino and Jonathan Culpeper. Amsterdam/Philadelphia: John Benjamins Publishing Company.
______. 2002b. "Phonetic Cues and Dramatic Function--Artistic Recitation of Metered Speech" (Expanded version). PSYART: A Hyperlink Journal for the Psychological Study of the Arts. http://www.clas.ufl.edu/ipsa/journal/2002/tsur05.htm
Weitz, Morris. 1962. "The Role of Theory in Aesthetics." In Philosophy Looks at the Arts, ed. J. Margolis. New York: Scribner.
Wellek, Rene, and Austin Warren. 1956. Theory of Literature. New York: Harcourt, Brace & Co.
Wittgenstein, Ludwig. 1976. Philosophical Investigations. Trans. G. E. M. Anscombe. Oxford: Blackwell.
Youmans, Gilbert. 1989. "Milton's Metre." In Phonetics and Phonology, Volume 1: Rhythm and Meter. New York: Academic Press.
Zemach, Eddy. 1976. Aesthetics. Tel Aviv: Literature, Meaning, Culture (in Hebrew).
Callow, Simon. Reading Shakespeare's Sonnets. Hodder Headline AudioBooks HH 185.
Gielgud, Sir John. Reading Shakespeare: Ages of Man. HarperCollins AudioBooks HCA 53.
Gielgud, Sir John. Reading Shakespeare: Sonnets. HarperCollinsAudioBooks HCA 67.
The Marlowe Society, and Professional Players. Reading Shakespeare: The Sonnets. Argo ZPR 254.
Reuven Tsur is professor emeritus of Hebrew literature at Tel Aviv University. His most recent books in English include On the Shore of Nothingness--A Study in Cognitive Poetics (2003) and Poetic Rhythm: Structure and Performance--An Empirical Study in Cognitive Poetics (1998).
[TEXT NOT REPRODUCIBLE IN ASCII] Dictionary 44 Gielgud 1 100 Gielgud 2 142 Marlowe 111 Callow 53 [TEXT NOT REPRODUCIBLE IN ASCII] Dictionary 106 Gielgud 1 89 Gielgud 2 77 Marlowe 81 Callow 78 [TEXT NOT REPRODUCIBLE IN ASCII] Dictionary 55 Gielgud 1 80 Gielgud 2 59 Marlowe 78 Callow 62 [TEXT NOT REPRODUCIBLE IN ASCII] Dictionary 95 Gielgud 1 Gielgud 2 Marlowe Callow 83 [TEXT NOT REPRODUCIBLE IN ASCII] Dictionary 137 Gielgud 1 478 Gielgud 2 492 Marlowe 203 Callow 133 [TEXT NOT REPRODUCIBLE IN ASCII] ratio Dictionary 300 0.456 Gielgud 1 269 1.776 Gielgud 2 278 1.769 Marlowe 197 1.030 Callow 276 0.481 Table 1 Durations of the speech sounds of "heaven" in milliseconds (msec). Gielgud and The Marlowe Society elide the [TEXT NOT REPRODUCIBLE IN ASCII] the Dictionary and Callow do not. In Gielgud's readings [TEXT NOT REPRODUCIBLE IN ASCII] is about 1.77 times longer than the combined duration of the preceding sounds. In the Dictionary and in Callow's reading the combined duration of the preceding sounds is over twice as long as that of [TEXT NOT REPRODUCIBLE IN ASCII] ln The Marlowe Society's reading [TEXT NOT REPRODUCIBLE IN ASCII] is insignificantly longer than the combined duration of the preceding sounds. [TEXT NOT REPRODUCIBLE IN ASCII] Dictionary 183 Gielgud 1 184 Gielgud 2 197 Marlowe 262 Callow 281 [TEXT NOT REPRODUCIBLE IN ASCII] Dictionary 177 Gielgud 1 174 Gielgud 2 109 Marlowe 147 Callow 131 [TEXT NOT REPRODUCIBLE IN ASCII] Dictionary 241 Gielgud 1 444 Gielgud 2 400 Marlowe 278 Callow 300 [TEXT NOT REPRODUCIBLE IN ASCII] Dictionary 360 Gielgud 1 358 Gielgud 2 306 Marlowe 409 Callow 412 Table 2 Durations of the speech sounds of "shun" in milliseconds (msecs). In the dictionary and in Callow's and The Marlowe Society's readings [TEXT NOT REPRODUCIBLE IN ASCII] is shorter than the combined duration of the preceding sounds: in Gielgud's two readings it is longer. [TEXT NOT REPRODUCIBLE IN ASCII] Dictionary 44 Gielgud 1 98 Gielgud 2 100 Marlowe 88 Callow 88 [TEXT NOT REPRODUCIBLE IN ASCII] Dictionary 278 Gielgud 1 150 Gielgud 2 175 Marlowe 181 Callow 128 [TEXT NOT REPRODUCIBLE IN ASCII] Dictionary 186 Gielgud 1 216 Gielgud 2 234 Marlowe 241 Callow 188 [TEXT NOT REPRODUCIBLE IN ASCII] Dictionary 322 Gielgud 1 248 Gielgud 2 275 Marlowe 269 Callow 216 [TEXT NOT REPRODUCIBLE IN ASCII] Dictionary 508 Gielgud 1 464 Gielgud 2 509 Marlowe 510 Callow 404 Table 3 Surprisingly, "men" is longish in the Dictionary. In Callow's reading it is the shortest of the four. In Gielgud 2 and The Marlowe Society's reading it is considerably longer than in Gielgud 1. compensating for the relatively low pitch of their intonation contour.
|Printer friendly Cite/link Email Feedback|
|Date:||Jan 1, 2006|
|Previous Article:||"The dark backward and abysm of time": the Tempest and memory.|
|Next Article:||Race, cognition, and emotion: Shakespeare on film.|