Studia Metrica et Poetica sisu 4_1.indd


Metre, rhythm and emotion in poetry
A cognitive Approach

Reuven Tsur*1 

Abstract: This essay integrates what I have written on the contribution of meter and 
rhythm to emotional qualities in poetry, opposing them to emotional contents. I dis-
tinguish between “meaning-oriented” approaches and “perceived effects” approaches, 
adopting the latter; and adopt a qualitative (rather than quantitative) method of 
research. Providing a simplified list of structural elements of emotion, I explore 
structural resemblances between rhythmic patterns and emotions. I investigate such 
issues as convergent and divergent poetic styles, convergent and divergent delivery 
styles, hypnotic poetry, the contribution of meter and rhythm to a “dignified quality”; 
and the rhythmic performance and emotional effect of stress maxima in weak posi-
tions. Finally, I locate my work between impressionist criticism on the one hand, and 
meaning-oriented criticism on the other. 

Keywords: poetic rhythm; emotional qualities; “meaning-oriented” criticism; “percei-
ved-effects” criticism; qualitative approach; quantitative approach; cognitive poetics

Introductory

This essay is an attempt to integrate what I have written during the past forty-
five years on the contribution of meter and rhythm to emotion in poetry, that 
is, emotional qualities rather than emotional contents, and locate it among 
other approaches. 

Impressionist criticism devotes much attention to emotions in poetry, but 
fails to relate them to the texts. Analytic criticism excels in the verbal analysis 
of the texts, but sometimes has difficulties to relate them to human qualities. 
Cognitive poetics explores emotions as described by cognitive psycholo-
gists, and offers plausible hypotheses (Margolis 1962) to justify systematically 
the attribution of emotional qualities to poetic texts. One must distinguish 
between emotional contents and emotional qualities in a poem. The former 
are closer to thought processes, the latter to perception. “Chilling and killing 
my Annabel Lee” may be thought to be a sad event, but perceived as playful or 

* Author’s address: Reuven Tsur, Tel Aviv University, Tel Aviv 69978, Israel; email:  
tsurxx@post.tau.ac.il.

Studia Metrica et Poetica 4.1, 2017, 7–40

doi: doi.org/10.12697/smp.2017.4.1.01



8 Reuven Tsur

as suggesting simplified mastery of reality. Perception, as opposed to think-
ing and remembering, occurs in a situation defined here and now, while the 
stimulus is present. In other words, perception is characterized by immediacy; 
thought and memory may be directed to more remote things. One becomes 
aware of the contents of this verse line by focusing on the extralinguistic world 
to which it refers; one becomes aware of its emotional quality by immedi-
ate attention to the interaction of its contents, syntactic structure and sound 
patterns. Sentences like “My sister is sad” and “This music is sad” are used 
in different senses (Bouwsma 1961). In the former, “sad” refers to a mental 
process of a flesh-and-blood human. In the latter, it does not refer to a mental 
process of the music; nor does it refer to the emotions the music arouses in the 
listener. One may be perfectly consistent when saying “This sad music inspired 
me with great happiness”. “Sad” means here that the speaker has detected some 
structural resemblance between the music and an emotion (Hepburn 1968). 
This is its perceived effect, its aesthetic quality. 

Approaches to the Versification–Emotion Relationship

When discussing approaches to the contribution of versification to emotion 
in poetry, one must make two consecutive distinctions. First, one must dis-
tinguish between approaches that rely on ad hoc principles and those that 
are governed by explicit, consistently applicable principles. Within the latter, 
one may distinguish between “meaning-oriented” approaches and “perceived 
effects” approaches. The former tends to displace attention from the sound 
of words to their meaning; the latter attends to resemblances between the 
sound structure and the structure of some emotion. Methodologically, one 
may distinguish between exploratory-qualitative method and confirmatory-
quantitative method. The latter counts the number of subjects who claim 
to have perceived a certain emotion in a text; the former offers “plausible” 
hypotheses based on psychological and linguistic research to account for such 
perceptions, even of a single participant. Cognitive poetics is an instance of 
the former, the statistical methods adopted by literary studies from the social 
sciences are instances of the latter. As I have demonstrated elsewhere (Tsur 
2012: 271–274), rigorous confirmatory-quantitative studies in reader response 
that are not preceded by proper exploratory-qualitative analysis are prone to 
predetermine the results of the research. Another direction in confirmatory-
quantitative studies is large-scale statistical investigation of certain stylistic 
devices in literary texts. The present study adopts the “perceived effects” 



9Metre, rhythm and emotion in poetry

approach, in the exploratory-qualitative mode, within a theoretical framework 
of cognitive poetics. In the rest of this section I will present several issues as 
discussed by the ad hoc approach, a statistical meaning-oriented approach, and 
an analysis that oscillates on the verge of the perceived-effects and meaning-
oriented approaches, and then suggest how the perceived-effects approach 
would treat them. 

Traditionally, emotions belong to the domain of impressionistic criticism. 
The critic is concerned with his own emotional responses rather than the 
structure of the text. Analytic criticism, by contrast, displays great concern for 
the structure of the text and the underlying principles. In the age of analytic 
criticism impressionism is frequently disguised by linguistic analysis; in such 
cases, however, the critic cannot point at some consistently applicable principle 
that would justify the attribution of some emotional quality to a text structure. 
Consider the following passage from Pope:

1. Our plentous Streams a various Race supply; 
 The bright-ey’d Perch with Fins of Tyrian Dye, 
 The silver Eel, in shining volume roll’d,
 The yellow Carp, in Scales bedrop’d with Gold, 
 Swift Trouts, diversify’d with Crimson Stains, 
 And Pikes, the Tyrants of the wat’ry Plains.
    (“Windsor Forest”, 141–146)

John A. Jones, in his book on Pope’s couplet art, quotes this passage, and makes 
the following comment on its third line:

Because the participle “roll’d” is the rhyme word, the verb quality of “rolling” is 
emphasized rather than adjectival or substantive quality. “Shining volumes” is 
more effective coming before the rhyme “roll’d” than it would be after it, for it 
is the climactic rolling or writhing that is highlighted. We do not always think 
of volumes as round but here it means “coils”; and when “roll’d” describes “vol-
umes”, the eelish quality is heightened, as the reader can easily imagine, even if 
he has never landed an eel. (Jones 1969: 74–75)

Doubtless, ‘the climactic rolling or writhing” of a landed eel has strong emo-
tive connotations. But, unfortunately, Pope used the past participle “rolled” 
and not the present participle “rolling”. To smoothen the transition, Jones 
mentions only “participle”, without specifying “present” or “past”. Or, consider 
the first sentence of this quotation from Jones. The word because suggests a 
logical, causal relationship between its two clauses. But is there? To justify 



10 Reuven Tsur

such a statement, there must be some unmentioned generalization that can be 
consistently maintained as, for instance, “When a past participle occurs in the 
rhyme word, its verb quality is emphasized rather than adjectival or substan-
tive quality”. I am not aware of any such valid generalization. Nor am I aware 
of any principle that would justify shifting the meaning of the passive “rolled” in a 
context of “shining volume” from ‘being formed into a mass by turning over and 
over’ into active “climactic rolling or writhing”. “Volume” suggests precisely 
such a mass. Rather than “climactic rolling”, it is the anticlimactic “rolled” that 
is manipulated into the rhyme. 

In fact, all the available grammatical and stylistic evidence suggests that the 
adjectival quality is emphasized in this epithet. Jones’s interpretation, however, 
crucially seems to depend on the past participle’s active “verb quality”. Since, 
however, everybody feels that rhymes do something important to words, and 
so little is known about what they do to them, Jones quite safely resorts to the 
rhyme word to enlist it in the service of the “verb quality” construal of the par-
ticiple. As a matter of fact, a straightforward adjective would do the job equally 
well, if it were possible to construe it as suggesting activity rather than a pas-
sive state. When words with certain meaning components are systematically 
manipulated into the rhyme, one can, perhaps, make a case for its significance; 
but even that cannot justify any generalization of this kind. Whenever the critic 
claims that there is some interaction between sound and meaning, he must 
make explicit the principles on which he is relying. Briefly, even if one could 
make out a case for a “climactic rolling or writhing” construal on metaphoric 
grounds, it would have nothing to do with versification.

Actually, such syntactic inversion is not confined to this couplet in Pope’s 
poetry: he quite frequently resorts to this stylistic device, manipulating the 
verbal element into the rhyme, with the complement preceding it, in precisely 
the first line of the couplet. Consider, for instance, the following couplet: 

2. Sole judge of Truth, in endless Error hurl’d: 
 The glory, jest and riddle of the world.
   (Pope: “An Essay on Man”, II. 17–18)

The antithesis leaves little room for doubt that Error in the first line is a word 
of key importance. Although hurl’d contributes such components to the image 
as helplessness, passive endurance and inconstancy, its decisive component, 
physical transfer, has little relevance to the thought. As a matter of fact, its 
main function is syntactic: to attribute somehow “error” to Man (‘victim to 
endless error’ would do as well). Hurl’d constitutes a virtuoso rhyme with 
world, but this has been achieved at the double price of an “inelegant” syntactic 



11Metre, rhythm and emotion in poetry

inversion and manipulating the word of key importance out of the rhyme. Is 
it possible that a great master of poetic technique like Pope should be guilty 
of such incompetence? And how could we explain, then, that precisely this 
“incompetent” line constitutes one of the most famous couplets on which 
Pope’s reputation as a major poet rests? The “perceived effects” approach, by 
contrast, assumes that the syntactic inversion and the manipulation of the key 
word out of the rhyme serve one common effect (not necessarily meaning). 
Likewise, semantically, “rolled” in Excerpt 1 serves to reinforce “volumes”; 
syntactically it is required to complete the attribution of “shining volumes” to 
“eels”. As the casual collection of couplets in my book (2008: 120–121) may 
suggest, such inversions, manipulating words of relatively little importance 
into the rhyme, are not uncommon in Pope. In all but one of those couplets 
the inversion occurs at the end of the first line. I shall argue that this is quite 
significant.1 Briefly, while Jones’s discussion presupposes an ad hoc principle, 
the “perceived effects” approach invokes here an explicit principle formulated 
by gestalt psychologists and art critics: 

The rule that governs the process is evident. The effect depends on the degree of 
simplicity of the whole as compared with the degree of simplicity of the parts. 
Greater simplicity of the whole makes for greater unity. The simpler the parts, the 
more clearly they tend to stand out as independent entities (Arnheim 1967: 61). 

“On the whole, however, the simplicity of any part must be modified or weak-
ened sufficiently to make the part dependent on, and therefore integrated with, 
its context” (Arnheim 1967: 65). By inverting the phrase and by manipulating 
the word of lesser importance into the rhyme, Pope weakens the simplic-
ity of the first line and increases its dependence on the couplet, enhancing 
the punchline quality of the second line. Briefly: it is perfectly true that the 
syntactic inversion manipulates “rolled” into the rhyme; the only question is, 
how does this affect the couplet’s perceived quality, and by what principle one 
may account for this. Rhyme cannot change the meaning of the word; but the 
manipulation of a word of lesser importance into the rhyme may affect the 
perceived quality of the rhyme. Persons who read poetry for mainly its mean-
ing will not notice such quality changes. 

Meaning-oriented approaches tend to displace the source of emotional 
qualities from the sound patterns of poetry to the meaning of the words. An 

1 In the first 50 couplets of “An Essay on Man” such line-final inversions occur in twelve first 
lines and in two second lines. 



12 Reuven Tsur

illuminating case can be found in Benjamin Hrushovski’s 1980 paper “The 
Meaning of Sound Patterns in Poetry: An Interaction View”, at the very meet-
ing point of the meaning-oriented and perceived-effects approaches. While 
I agree with much of what he says in that paper, I wish he went one step fur-
ther. Here I will quote his analysis of one of his examples and consider how 
metric structure may or may not interfere with the sound-expressive process. 
Traditional poetics has important things to say about how tones, moods and 
emotions are abstracted from the meaning of the words; but how are they 
abstracted from the speech sounds? My 1987 book can be described as an 
inquiry into possible sources of the “tone” or “emotional quality” generated by 
speech sounds and the way these traits are grasped as parallel to an abstraction 
from the meaning of the words (tone, mood, emotion, etc.). One important 
aspect of the issue is that sounds are what I call “double-edged”; that is, they 
may be “expressive” of vastly different, or even opposite, qualities. Thus, for 
instance, the sibilants /s/ and /š/ may have a hushing quality in one context, and 
a harsh, noisy quality (to varying degrees) in some other. Thus, for instance, 
Hrushovski quotes Poe’s line

3. And the silken, sad, uncertain rustling of each purple curtain

where the sibilants may be onomatopoetic, imitating the noises; or they may 
reinforce – or be expressive of – a quiet mood in Shakespeare’s sonnet:

4. When to the sessions of sweet, silent thought 
 I summon up remembrance of things past,
 I sigh the lack of many a thing I sought
 And with old woes new wail my dear time’s waste.

Hrushovski argues that the source of the difference is that the two excerpts 
exemplify two different types of logical relationship between sound and mean-
ing. The “perceived effects” approach would rely here on the assumption that 
sounds are bundles of features, on the acoustic, phonetic and phonological 
levels, and that the various features may have different expressive potentials. 
The claim I have elaborated in my 1987 book is that in different contexts, dif-
ferent potentials of the various features of the same sounds may be realized. 
Thus, for instance, the sibilants [s] and [š] may have at some level of descrip-
tion features with noisy potentials as well as features with hushing potentials. 
To spell this out: speech sounds are typically encoded, that is, the acoustic 
information that transmits it is restructured into an abstract phonetic category, 
and excluded from consciousness. Some speech sounds are more thoroughly, 



13Metre, rhythm and emotion in poetry

some less thoroughly encoded. If you ask “which one is acoustically higher, 
[ba], [da], or [ga]”, most people will answer that they don’t know what you are 
talking about. If, however, you ask which one is higher [s] or [š], most people 
will easily answer that [s] is higher. In certain circumstances, responsiveness 
to abstract categories goes with rigidity; responsiveness to rich elusive sensory 
information with flexibility. The double-edgedness of the sibilants is related 
to different aspects of the same noises. The tender or hushing quality of [s, 
š] may have to do with their feature [+CONTINUOUS] and with their being 
of the few consonants that are less thoroughly restructured in the course of 
phonetic decoding of the signal, enabling the perceiver to attend to some 
rich, inarticulate sensory information; tenderness is typically correlated with 
openness to rich sensory information. That is why mothers are wont to sound a 
prolonged [š] sound to hush their crying babies. The “noisy” quality of these 
sibilants springs from the aperiodic nature of the very same sensory informa-
tion, a stream of irregular noises. In Poe’s line, the contents realize the noisy 
potential, in Shakespeare’s quatrain the hushing potential. In Wittgenstein’s 
terms, readers are “aspect-switching” from one excerpt to the other. On the 
other hand, speech sounds may definitely lack certain other potentials, as a 
rewriting exercise performed by Hrushovski (1980: 44) may suggest:

Now, if this is the case, would not any sound pattern do? Let us try to “rewrite” 
the Shakespearean lines using words similar in content:

5. When to the crux of crucial quiet thought
 I crave and call remembrance of things past

We have already created a very similar network of sounds, this time based on the 
repetition of K, strengthened by the cluster K + R (involving the original word 
“remembrance” too). Nevertheless, it seems that this sound pattern cannot pos-
sibly express silence, though “quiet thought” starts with K as “silent thought” 
starts with S. It is plausible that a reader will impute to this text something strong 
and harsh, reinforced by the sound pattern. The pivotal word may become “crux”, 
though its counterpart “sessions” was subordinated to “sweet silent thought”. One 
may generalize that, in a part of a text in which a sound pattern coexists with a 
number of semantic elements, the sound pattern may contribute to shifting the 
center of gravity from one direction of meaning to another.

Had the speech sounds no expressive potential of their own, the network 
of sounds based on /k/ would have readily assumed the emotional quality 
of quietness, which it doesn’t. Here the sound tends to confer upon the text 



14 Reuven Tsur

“something strong and harsh”, and “may contribute to shifting the center of 
gravity from one direction of meaning to another” (e. g., from quiet to crux). 
But Hrushovski does not go into the phonetic-acoustic source of this potential. 
According to the present conception, voiceless plosives are perceived as hard, 
because they are abrupt and thoroughly recoded: none of the rich precatego-
rial sensory information reaches awareness. 

This putative shifting of the center of gravity became possible only through 
a regularization of metre in the transcription: metric deviation, the two suc-
cessive, alliterating stressed syllables in “sweet silent” foreground these words 
and focus attention on their meanings. So, let us amend the transcription to

6. When to the quorum of kind, quiet thought

Now notice this: even though we are prevented now from shifting the center of 
gravity to some other “pivotal word”, the /k/ sound retains its hard and strong 
quality, and by no means becomes expressive of some “kind, quiet” atmos-
phere originating in the meaning of the words. The sound pattern becomes 
either “neutral”, or improper to the emotion expressed. In view of the present 
analysis, one can account for the harsh qualities of the [k] and the hushing 
qualities of the sibilants without having recourse to the meaning of the alleged 
“pivotal words”, by the oppositions abrupt vs. continuous and more vs. 
less encoded.

As this exercise may suggest, speech sounds do have (sometimes conflict-
ing) emotional potentials of their own and one may not ascribe to them just 
any quality suggested by the meaning of the text. The meaning of the words 
may activate one or the other potential, but cannot generate them. Metric 
structure can focus attention on the emotional connotations of certain words; 
but, typically, it generates perceived effects through its gestalt qualities (as 
demonstrated throughout this essay). 

Highly relevant to the present inquiry, Maria Kraxenberger’s recent PhD 
dissertation “On Sound–Emotion Associations in Poetry” (2017) explores the 
associations of a variety of prosodic aspects with emotions in poetry. Our 
approaches differ at least in two respects: her work is confirmatory-quantita-
tive; and we differ in the “object of imitation” too, to use Aristotle’s term. She 
relies on “acoustic profiles that are dependent on suprasegmental parameters” 
that characterize people as they speak. I, by contrast, focus on the resemblance 
between the structure of texts and the structure of emotions. In this, I rely 
mainly on gestalt qualities and precategorial auditory information behind the 
phonetic categories. In discussing vocal performances, Kraxenberger relies on 
the following parameters: 



15Metre, rhythm and emotion in poetry

we tested the extent to which the suprasegmental features that are known to 
influence joyful and sad prosody in general are actually characteristic of how 
native-speaking participants recite poetry. To do this, we analyzed the acoustic 
features of the emotional prosody of nonprofessional native German speakers’ 
recitations of  joyful and sad German poems. In line with previous research, we 
focused on mean pitch, mean intensity, and articulation rate. 

I said that this approach represents people as they speak, because joyful people 
typically display greater pitch span, greater span of intensity in speech, and 
speak faster than sad people. The fact that she relies on the mean rather than 
the span of pitch, intensity, and articulation rate obscures, of course, the issues 
to some extent. As we shall see, I on the other hand, argue that the structure of 
e.g. divergent poetry resembles in significant respects the divergent streams of 
information in emotions. I also treat, following Henry J. Todd, Milton’s phrase 
“supréme Kíng” as displaying some structural resemblance between the con-
secutive heavy stresses and the outward manifestation of dignity in humans 
(not necessarily as they speak),, when performed as a spondee. 

Kraxenberger, Maria “On Sound–Emotion Associations in Poetry” (2017). 
Unpublished Freien Universität Berlin dissertation.

At this point, I want to mention Marina Tarlinskaja’s work in metrics. Her 
approach is diametrically opposed to mine in many respects, among others, it 
is strongly meaning-oriented and confirmatory-quantitative. Perceived effects 
are not within the scope of her research. It is based “not so much on insight and 
intuition as on wide quantitative analyses of observable facts”. In spite of this, 
our findings are considerably overlapping, but with quite different implications.  
I will present these overlaps and differences below, in the context of my discus-
sion of stress valleys.2 

2 The Lakoff school of “cognitive poetics” is essentially meaning-oriented, and has very lit-
tle interest in the sound patterns and perceived qualities of poetry. One of the rare studies in 
this respect, Eve Sweetser’s (2006) paper on versification in Cyrano de Bergerac, is extremely 
meaning-oriented. In my (2008a) detailed counterproposal (reprinted in Tsur 2008b), I point 
out the perceived effects she misses owing to her unduly rapid shift of attention from the sound 
patterns to meaning. 



16 Reuven Tsur

Emotion and Emotional Qualities

At the beginning let us give in a nutshell a simplified list of some structural 
elements of emotion. Any element in the description of emotions may “typi-
cally count toward” more than one aesthetic (or emotional) quality, sometimes 
toward incompatible qualities (Sibley 1962). Emotions involve situation 
appraisal through thought processes, plus undifferentiated energy. They con-
sist of streams of information that display a tendency toward an object judged 
suitable or away from an object judged unsuitable. Emotions involve devia-
tion from normal levels of energy; e. g., low energy level may count toward 
such emotions as sadness or calm; high energy level may count toward such 
emotions as gladness or anger. Emotions are active in the background, with-
out preempting everything else. Weak gestalts are associated with emotional 
qualities, strong gestalts with rational or witty qualities. A special combina-
tion of weak and strong gestalts tends to yield “hypnotic poetry” (see below). 
Psychologists distinguish “convergent thinking” and “divergent thinking abili-
ties” (for logical and creative thinking, respectively) (Guilford 1970 [1959]). 
Emotions are more divergent than creative thinking. Some altered states of 
consciousness as mindfulness meditation are highly correlated with divergent 
thinking (Berkovich-Ohana et al. 2016). There is no fixed point on this spec-
trum where one category turns into another (Duffy 1968 [1941]). Language 
and logic are typically related to the left hemisphere of the brain, emotions 
to the right hemisphere. Our aim is to create critical tools that have sufficient 
descriptive contents to describe the resulting structures in a text. 

This paper will explore the contribution of metre and rhythm to emotional 
qualities in poetry. Emotional contents will concern us only as part of a more 
complex configuration. The assumption is that the sound patterns contribute 
to poetry something that cannot be reduced to meaning, only to perceptual 
qualities. They generate some general psychological atmosphere, that can be 
individuated by semantics in a withdrawn, and receptive attitude, or an asser-
tive, outward directed and actively organising attitude, respectively, in a wide 
variety of emotions.

Convergent and Divergent Poetry 

Poetic rhythm has three dimensions: versification, language, performance 
(Wellek & Warren 1956). Versification and language may act in convergence or 
divergence. When they conflict, a rhythmical performance may accommodate 



17Metre, rhythm and emotion in poetry

them such that both can be perceived in one vocal actualisation. I conceive 
of a rhythmical performance as a problem-solving activity. In performance, 
we must assume “triple-encodedness”: the same noises may have simultane-
ously three different functions: a falling intonation contour, for instance, may 
indicate the end of a syntactic unit (sentence), of a rhythmic unit (the line); at 
the same time, it may be expressive of a long fall or, alternatively, of coming 
to a rest. 

In a convergent style, phrase endings and line endings tend to converge; 
stressed syllables tend to occupy metrical strong positions, unstressed sylla-
bles – weak positions; alliterations tend to coincide with stressed syllables in 
strong positions; stanza forms tend to be symmetrical. In a divergent style, all 
these tend to diverge, and stanza forms tend to be asymmetrical. Up to a theo-
retically undefined point, divergent style tends to yield increasingly emotional 
poetry, as in Milton and Shelley; beyond that point, it tends to yield witty 
poetry, as in Donne’s Satires. A convergent style, in certain circumstances may 
yield witty poetry, as in Alexander Pope, or simplified mastery of reality, as in 
nursery rhymes and folk poetry. Leonard B. Meyer accounts for the association 
of weak and strong Gestalts in music with emotional and intellectual qualities 
respectively, as follows. “Because good shape [and we could add “convergent 
structures”] is intelligible in this sense, it creates a psychological atmosphere 
of certainty, security, and patent purpose, in which the listener feels a sense of 
control and power as well as a sense of specific tendency and definite direc-
tion” (Meyer 1956: 160).

A combination of regular metre with irregular or asymmetrical stanzas 
and/or frequent enjambments (and some additional characteristics, see below) 
may yield hypnotic poetry, as in Coleridge’s “Kubla Khan”, or in Edgar Allan 
Poe’s “The Raven” and “Ulalume”; in “Annabel Lee” the beginning suggests 
a childish attitude, simplified mastery of reality, but the last stanza becomes 
hypnotic poetry (prosodic structure remains the same, the difference is in the 
situation appraisal). 

Structuralist phonetics conceives of speech sounds as bundles of distinctive 
features (Jakobson [et al.] 1952). According to one of the Gestalt rules of per-
ception, interaction between gestalt-free elements (colour in visual perception 
and overtones in music) is increased within, but inhibited across strong gestalt 
boundaries; but it is boosted across weak Gestalt boundaries. This may apply 
to distinctive features too. In convergent and divergent poetry, alliteration 
assumes different characters: the alliterating speech sounds are more scat-
tered in divergent than in convergent poetry; in convergent style, they tend to 
reinforce meter, in divergent style they tend to blur meter; at the same time, 
the distinctive features are perceived as more closely packed in convergent 



18 Reuven Tsur

alliteration, more diffuse in divergent alliteration, interacting across the weak 
Gestalt boundaries.

Consider two excerpts from Poe and Milton:

7. That the wind came out of the cloud by night,
 Chilling and killing my Annabel Lee.
   Poe: “Annabel Lee” 

8. Of Man’s first disobedience, and the fruit
 Of that forbidden tree whose mortal taste
 Brought death into the World, and all our woe,
 With loss of Eden, till one greater Man
 Restore us, and regain the blissful seat,
 Sing, Heavenly Muse,
   Milton: Paradise Lost

Where is alliteration more intensive? More likely than not, we will get the 
answer that “chilling and killing” is more intensive. But consider this. In 
Milton’s passage, there is an unusually great number of intensive alliteration 
patterns. F-r-t in first and fruit; s-t in first, taste, restore, seat; l-s in loss and 
bliss; Eden is wholly included in -bidden which, in turn, is wholly included 
in -bedience. Now notice this too. In the Poe excerpt, stressed syllables occur 
in all and only strong positions, unstressed syllables in all and only weak 
positions; there is one exception: “came” in a weak position, but it too is sub-
ordinated to “out”. The alliteration pattern consists of two symmetrical whole 
words that comprise identical speech sounds (all but one), in the same order. 
The alliteration patterns reinforce metric regularity. In Milton’s excerpt, by 
contrast, stressed syllables occur in mainly, but not all and not only strong 
positions; unstressed syllables occur in mainly, but not all and not only weak 
positions. Consequently, Poe’s metric shape is convergent, much stronger than 
Milton’s. Conversely, Milton’s verse is highly divergent, has much weaker met-
ric gestalts. What is more, Milton’s passage displays an exceptionally complex 
enjambment, blurring the line boundaries. It begins with a preposition (“of ”) 
predicting a verb, which occurs six lines later (Sing). All the time, the reader 
must remember he has a loose end, while following the proceeding of the 
text. In addition, we have such minor enjambments, as “and the fruit/Of ”, 
or “mortal taste/Brought”, or “till one greater Man/Restore us”. Consider a 
rewriting of Excerpt 8 as “Heavenly Muse, Sing /Of Man’s first disobedience 
and the fruit”, etc. It has a more straightforward, single-minded psychological 
atmosphere. Consequently, the actual Milton passage has exceptionally weak 



19Metre, rhythm and emotion in poetry

gestalts at all levels. This boosts the interaction of the precategorial phonetic 
and acoustic features of the alliteration patterns across all those boundaries. 
Briefly, in Poe’s text, alliteration is focused on two compact words, perceived 
as playful punning rather than diffuse musical texture. Milton’s alliterations, 
on the contrary, are backgrounded as rich, diffuse, but fused, precategorial, 
“musical” texture. Consequently, Milton’s “great argument” assumes an intense 
emotional quality (Tsur 1978). 

Convergent and Divergent Delivery Style

In poetic structure I have distinguished “divergent” and “convergent” styles. 
Pope is the archetypal convergent poet, usually perceived as witty; Milton and 
Shelley are the archetypal divergent poets, usually perceived as emotional. In 
my later work I found that in performance too there is a divergent and a con-
vergent delivery style. The performer has considerable freedom in choosing 
his delivery style. Sometimes the same actor performs the same verse line in 
a convergent style on one occasion, and in a divergent style on another. I will 
present my categories of “convergent” and “divergent” delivery styles by way 
of scrutinising two recordings by Sir John Gielgud, sixteen years apart, of the 
last line of Shakespeare’s Sonnet 128 (Tsur 2006): 

 9. To shún the héaven that léads mén to this héll.
    w      s     w       s         w      s       w     s    w    s

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 ana-
lyse it. I had an unexplained impression that Gielgud 2 is much more complex, 
artistically more sophisticated, rhythmically more satisfying. The best way 
to characterise 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 inten-
sive, 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 seg-
ments, 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 



20 Reuven Tsur

and pitch plots extracted from the two readings, an immediately-perceived 
difference becomes conspicuous (Figures 1–2). 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” [listen3]. In the wave plot extracted from Gielgud 
2, by contrast, no such pauses are visible. Discontinuation is achieved here by 
means other than straightforward pauses [listen]. Notwithstanding this, one of 
my associates could hardly believe that there is no pause there in this reading. 

Figure 1. Wave plot and pitch extract of “To shun the heaven that leads men to this 
hell” Gielgud 14

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” back-
ward rather than forward. It is the prolongation and overarticulation of the 
word-final /n/ that bears all the burden of generating discontinuity at the 

3 Sound files are available in the online version of this paper. Ed.
4 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. 



21Metre, rhythm and emotion in poetry

caesura. 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 group the word “that” backward; the listener’s 
syntactic knowledge and the sustained /n/ indicating a lack of progression 
suggest a new start after “heaven”. Consequently, a caesura and a “metrical 
impulsion” across it are perceived at the same time. The line exerts pressure 
for completion upon which the caesura obtrudes. 

Figure 2. Wave plot and pitch extract of “To shun the heaven that leads men to this 
hell” Gielgud 2

I said above that in Gielgud 1 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 mean-
ing (or quality) it modifies. This happens because the grouping cues cluster 
differently from Gielgud 2, 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 /n/. It is this overarticulated syntactic juncture that confirms 
a prosodic event – a caesura. In other words, in Gielgud 1 the cues act in 
convergence, in Gielgud 2 in divergence. Two differences between the falling 



22 Reuven Tsur

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 
in 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 /n/ 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, and a 
longer-than-usual word-final /n/. Exceptionally great stability is achieved. In 
Gielgud’s readings /n/ is about 1.77 times longer than the combined dura-
tion of the preceding sounds h+ε+v. By comparison, in the Audio Edition of 
Merriam-Webster’s Collegiate Dictionary, and in Simon Callow’s reading of the 
same line the combined duration of the preceding sounds is over twice as long 
as that of /n/. In The Marlowe Society’s reading /n/ is insignificantly longer 
than the combined duration of the preceding sounds. A similar story can be 
told, mutatis mutandis, about the sequence “shun the”. In Gielgud 1 there is a 
longish pause between them (413 msec); in Gielgud 2 they are run one into the 
other. In both readings /n/ 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 Merriam-Webster’s Collegiate Dictionary, by con-
trast, /n/ is considerably shorter. As after “heaven”, after “shun” too Gielgud 
1 resorts to redundant cues: the prolongation of /n/ 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. 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 nei-
ther syntactic, nor prosodic justification. It seems to be gratuitous – unless 
it has some rhetorical or paralinguistic justification. I have quoted Meyer on 
the psychological atmosphere inspired by convergent structures in music and 
poetry (cf. Meyer 1956: 160). In Gielgud 1 some elements of that atmosphere 
are present, the combination of cues for discontinuity thrusts upon the listener 
a psychological atmosphere of patent purpose as well as a sense of specific 
tendency and definite direction; but coupled with the voice quality, it inspires 
a psychological atmosphere of insecurity and anxiety.

I made an attempt to assess how listeners respond to Gielgud’s two read-
ings. Though I was asking about “rhythmical performance”, my informants 
could not keep it apart from the emotional element. As far as pauses are 



23Metre, rhythm and emotion in poetry

concerned, the same element may be an exponent of both the rhythmic and 
emotional quality. It is the latter that may justify the huge pause after “shun” 
too in Gielgud 1. Most informants agreed that the first reading is more emo-
tional. 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 prob-
ably referred to the same perceived quality, though attributed it to different 
aspects (rhythm, tone). In these comments, the emotional quality was implic-
itly judged as a “good-making” feature of the performance. One respondent 
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. In terms of the foregoing discussion, 
in Gielgud 1 the intensive human quality is not balanced by complexity. As 
we have seen, the effect of divergent cues may be subtler; that of convergent 
cues – more robust. To conclude, then, in Gielgud 1 the combination of cues 
for discontinuity is thrust upon the reader “in a blunt and direct manner”, act-
ing as a modifier that has little meaning except to intensify any perceptual or 
emotional qualities present, in this case anxiety inspired by the voice quality.

Hypnotic Poetry

Edward D. Snyder (1930) put forward the fruitful idea that “certain poems 
have a peculiar trance-inducing technique; that they owe their mysterious 
‘spell’ to a magic no more incomprehensible than that of hypnotism; that by 
intensifying the listeners’ suggestibility they permit experiences where – for 
better or for worse – the poet holds sway over the listeners’ conscious and 
subconscious mind” (Snyder 1930: 38). In an effort to isolate the personal-
ity variable “absorption”, devised to predict the hypnotisability of a person, 
Tellegen (1981: 222) propounded the dichotomy “experiential” and “instru-
mental set”, the former being “a state of receptivity or openness to undergo 
whatever experiential events, sensory or imaginal, that may occur, with a ten-
dency to dwell on, rather than go beyond, the experiences themselves and 
the objects they represent”; in the latter, input from receptors is not used 
to enhance experiencing but to make the discrimination needed for guiding 
instrumental acts and evaluating achievements. Such altered states of con-
sciousness as hypnosis or meditation strongly presuppose an experiential set. 



24 Reuven Tsur

One of the distinguishing marks of hypnotic poetry is its “obtrusive 
rhythm”. The poems mentioned above as hypnotic are of the most metrically 
regular poems in English. The regular rhythms do not generate here rational-
ist poetry as in Pope, for instance. Such obtrusive rhythm may amplify our 
involuntary physiological and psychological processes and, by the same token, 
it has been suggested, give “false security to the Platonic Censor in us” – so 
that the reader may feel freed to attend to ambiguities, or the irruption of the 
irrational, in the other layers of the poem. Regular, predictable meter gives 
security. Sometimes it turns “false” owing to the ambiguities and the irrup-
tion of the irrational in the poem, as in the afore-mentioned poems by Poe 
and Coleridge. But sometimes the source of insecurity is in the versification 
itself: unpredictable rhyme patterns, unpredictable line length, irregular stanza 
forms, etc. Poe is the grand master of highly predictable rhyme words, but 
irregular stanza forms. In hypnotic poetry, as in “The Raven” or “Kubla Khan”, 
alliteration may reinforce both security and insecurity. Alliterations like “On 
the pallid bust of Pallas” and “of a demon that is dreaming” reinforce security; 
the diverging and interlacing sound patterns that blur each other inspire inse-
curity. Jakobson pointed out such interlacing alliterations in the poem, e. g.,

“raven” contiguous to the bleak refrain word “never” appears once more as an 
embodied mirror image of this “never”: /n.v.r/ – /r.v.n/. Salient paronomasias 

interconnect both emblems of the everlasting despair, first “the Raven, never flit-
ting” at the beginning of the very last stanza, and second, in its very last lines the 

“shadow that lies floating on the floor” and “shall be lifted – nevermore”: /nεvәr 
flítíŋ/ – /flótíŋ/ ... /flɔr/ ... /líftәd nεvәr/. (Jakobson 1960: 372)

Jakobson quotes these alliterations as instances of the poetic function in general; 
I am quoting them as characteristic of emotional poetry, and in combination 
with more than usually regular meter as characteristic of hypnotic poetry. 

In “Kubla Khan” we have such alliterations as “Kubla Khan”, “measureless 
to man”, “sunless sea”, “woman wailing”, “mingled measure” where the allitera-
tions fall on stressed syllables in strong positions. At the same time, we have 
got such alliterations as “Xanadu did”, “dome decree”, where the alliterations 
occur in adjacent positions and blur meter. In Poe, obstinately repeated stock 
phrases and rhyme words, verse lines repeated with minute changes suggest 
“a tendency to dwell on, rather than go beyond, the experiences themselves 
and the objects they represent”. The enormous energies in “Kubla Khan” and 
the drastically reduced energy levels in “Ulalume” generate different modes 
of experiencing the irruption of the irrational (in “Ulalume”, the obstinate 
repetitions also suggest a psychological atmosphere of lack of progression). 



25Metre, rhythm and emotion in poetry

Glicksohn, Tsur and Goodblatt (1991) tested the responses of high- and 
low-absorption participants to a Hebrew hypnotic poem. The former evalu-
ated it as interesting, tense, pleasant and complex; the latter, as boring, relaxed, 
unpleasant and simple. The different qualities perceived by high- and low-
absorption participants in this poem are hypothesized as being the result of 
the application of different cognitive strategies in organizing the poem as a 
whole. The latter tend to foreground those elements in the poem that induce 
“certainty, security, patent purpose” and minimize those that inspire awe and 
anxiety (ambiguities and the irruption of the irrational); the former tend to 
balance them against each other. Glicksohn et al. assume, therefore, that the 
different cognitive strategies used in organising the poem in question as a 
whole involve mainly the factors that determine whether the rhythm is per-
ceived as one of false or genuine security. This may explain, why some readers 
regard Poe’s poetry as the quintessence of poetry, whereas others as rather 
mediocre, minor poetry. 

Let us have a look at the first stanza of “Ulalume”. 

10. The skies they were ashen and sober; 
 The leaves they were crispèd and sere – 
 The leaves they were withering and sere; 
 It was night in the lonesome October 
 Of my most immemorial year; 
 It was hard by the dim lake of Auber, 
 In the misty mid region of Weir – 
 It was down by the dank tarn of Auber, 
 In the ghoul-haunted woodland of Weir. 

The verbal structure of this stanza displays a conspicuous tendency to dwell on, 
rather than go beyond, the experiences themselves, and also conveys consider-
able minute, precategorial information. None of the finite verbs suggest action, 
all are forms of the copula be, with descriptive phrases as predicates. The reader 
must also dwell on the obstinately-repeated rhyme words sere, Auber, Weir at 
unpredictable points of the stanza, suggesting lack of progression. So does the 
almost verbatim repetition of line 2. By the same token, this repetition upsets 
the symmetry of the stanza, inspiring uncertainty. In the second and third 
line there are three near-synonyms: crisped, sere, and withering. Lines 2 and 3 
are literally identical, except for the substitution of the adjective withering for 
its near-synonym crisped. This generates a minimal semantic pair, based on 
the opposition crisped ~ withering, the latter emphasizing the process of los-
ing moisture and vitality, the former emphasizing the resulting texture. These 



26 Reuven Tsur

synonyms emphasize different features in sere. Thus, the two tokens of sere 
have, and do not have, identical meanings, at one and the same time, generat-
ing a diffuse, precategorial texture. A similar, though less elusive, story can 
be told of the other parallel phrases in the lines with repeated rhyme words: 
dim lake ~ dank tarn, misty mid region ~ ghoul-haunted woodland, that also 
include an element of amplification. On the phonetic level, Excerpt 10 contains 
an unusual number of liquids ([l, r]), nasals ([m, n, ŋ]) and semivowels ([w, 
y]). These speech categories share two features: they are periodic and relatively 
unencoded. In periodic speech sounds the same wave form is repeated, induc-
ing a relaxed attitude (there will be no surprises). Speech sounds are encoded: 
we perceive the unitary speech categories rather than the acoustic information 
that transmits them. In liquids, nasals and semivowels some of the acous-
tic information is perceptible in the background, contributing to the diffuse, 
precategorial background texture. Thus, both on the semantic and phonetic 
levels “chaotic over-differentiation” is generated, enhancing “the falseness” of 
security, and openness to the subtleties of experience.

Consider the same stanza with the repetitions-plus-variations omitted. It 
suggests a much more straightforward, goal-oriented psychological atmosphere: 

11. The skies they were ashen and sober; 
 The leaves they were crispèd and sere – 
 It was night in the lonesome October 
 Of my most immemorial year; 
 It was hard by the dim lake of Auber, 
 In the misty mid region of Weir – 

Note that the first four lines of Excerpt 11 constitute a closed and stable sym-
metrical a-b-a-b quatrain. The addition of two more lines restructures the 
perception of the stanza, and generates a fluid ab-ab-ab structure that displays 
a psychological atmosphere of definite direction and patent purpose, going 
beyond, rather than dwelling on, the experiences themselves. 

A poem becomes hypnotic, then, when greater than usual regularity is 
coupled with greater than usual irregularity, ambiguity and irruption of the 
irrational, inspiring awe and anxiety – as well as substantial deviation from 
normal energy level, either increasing or decreasing it, with a tendency to dwell 
on, rather than go beyond, the experiences themselves. 



27Metre, rhythm and emotion in poetry

Metre, Rhythm, Dignity

I will introduce this issue via a controversy regarding the first line of the fol-
lowing excerpt from Milton: 

12. And sat as Princes, whom the supreme King
 Exalted to such power, and gave to rule,

Milton’s 1809 editor, Henry J. Todd comments on the first line of this excerpt: 
“I conceive that Milton also intended the last foot of the following verse to 
be a spondee, as more dignified and impressive than the accentuation, not 
uncommon indeed in our old poetry, of súpreme on the first syllable” (Todd 
1970: 199). One hundred eighty years later, Gilbert Youmans takes issue with 
him: “Todd is almost certainly mistaken about Milton’s intention in this case. 
[…] Otherwise, these lines would violate one of Kiparsky’s categorical rules for 
Milton” (Youmans 1989: 349–350). The issue at stake here is whether rhythm 
ought to be regularized or not. 

The two agree about the structure of the verse line; they disagree about its 
performance. And they support their respective positions by reasons of dif-
ferent kinds. Youmans invokes metricality judgments, Todd – poetic effects. 
Psychologically, “the rhythmical performance” of a poem is a perceptual solu-
tion to a perceptual problem: when the linguistic and versification patterns 
conflict, they are accommodated in a pattern of performance, such that both 
are perceptible simultaneously. Youmans and Todd seek their solutions to this 
line within a convergent and a divergent delivery style, respectively.

Where several delivery styles are possible, one reason for choosing one 
rather than another may be aesthetic. Beardsley (1958: 465–469) speaks of 
three “general canons” of evaluation: unity, complexity, and some intense 
human quality. The greater the complexity of an aesthetic object, or the 
greater its unity, other things being equal, the better it will be considered. If 
the reciter “accentuates súpreme on the first syllable”, straightforward unity 
may be greatly enhanced, but at the expense of complexity. If, however, the 
reciter strongly emphasizes both -preme and King, he may offer, in certain 
circumstances, a “rhythmical performance”, so that both unity and complexity 
are enhanced, relative to each other.5 The “certain circumstances” include the 

5 Arnheim (1967: 49) speaks of “simplicity” and “complexity”, and observes: “In an absolute 
sense, a thing is simple when it consists of a small number of structural features. In a relative sense, 
a thing has simplicity when it organizes complex material with the smallest possible number of 
structural features”. This holds true, mutatis mutandis, for unity and complexity as well.



28 Reuven Tsur

over-articulation of -preme, to free mental processing space for the perception 
of the versification pattern conflicting with it. The present assumption is that 
listeners decode the consecutive heavy stresses using their knowledge of the 
phonological and versification features whose interaction produced their into-
nation signals; by the same token, they perceive the metric Gestalt as blurred. 

Finally, when Todd defends his reading as a spondee on the grounds that 
it is “more dignified and impressive”, he invokes the general canon of “intense 
human quality”. What may “dignified” mean with reference to metric struc-
tures? It suggests that the listener may detect some structural resemblance 
between the consecutive heavy stresses and the outward manifestation of dig-
nity in humans, such as weightiness, reserve of manner, and clearly articulated 
gestures. “Weightiness” in a context of a disyllable with its stressed syllable 
in a weak position, followed by another stressed syllable, suggests “having 
considerable mass”; in a context of dignified human behaviour it suggests “of 
much importance or consequence”. As to “reserve”, a stressed syllable in a weak 
position (followed by another stressed syllable) “holds back” the rhythmic 
movement of the line, whereas a dignified person “holds back”, “controls” his 
responses, the expression of his emotions or thoughts. As to “clearly articulated 
gestures” in poetry, I have suggested that the rhythmical performance of such 
constructs as a disyllable with its second, stressed, syllable in a weak position 
requires exceptionally clear articulation. 

I do not mean to imply that this is the permanent perceptual quality of 
consecutive stresses. As I have suggested, one poetic structure may typically 
count toward a wide variety of sometimes conflicting perceived effects. Such 
expressions as Milton’s “supreme King” and Pope’s line “And ten low Words oft 
creep in one dull Line” exploit different potentials of slow movement induced 
by successive stresses. Shakespeare spelled out the nature of slowness of a 
desirable kind:

13. They that have power to hurt and will do none,
 That do not do the thing they most do show,
 Who, moving others, are themselves as stone,
 Unmoved, cold, and to temptation slow,
 They rightly do inherit heaven’s graces ... (Sonnet 94)

We must clear one more obstacle on the way of this solution. Wimsatt and 
Beardsley (1958: 598) write: “The notion of an accentual spondee (or ‘level’ 
foot) in English appears to be illusory, for the reason that it is impossible to 
pronounce any two successive stresses in English without some rise or fall of 
stress – and some rise or fall of stress is all that is needed for a metrical ictus”. 



29Metre, rhythm and emotion in poetry

If they are right, Todd’s analysis is fallacious. But they are wrong. With refer-
ence to the line “To a green thought in a green shade” they write: “whatever we 
do with the two pairs of syllables, it remains absolutely certain that ‘thought’ 
is stronger than ‘green’ and that ‘shade’ is stronger than ‘green’” (595). This 
absolute certainty is not as well founded as it could be. Thus, for instance, in 
his investigation of eleven recorded readings of Shakespeare’s Sonnet XVIII, 
Chatman (1965) played excised segments of two syllables each to a panel of 
twenty-one professors of English, asking them to make judgments on their 
relative stresses. The professors voted unanimously that, e. g., in four out of 
eleven performances, Rough and winds were performed with level stresses, in 
six winds was performed as slightly more prominent, whereas in one perfor-
mance Rough was slightly more prominent. This I take as evidence that “the 
notion of an accentual spondee (or ‘level’ foot) in English” is not illusory.

There is another way to generate a dignified quality by metre in poetry, 
pointed out by Aristotle in his Rhetoric: 

Of the various rhythms, the heroic has dignity, but lacks the tones of the spoken 
language. The iambic is the very language of ordinary people, so that in com-
mon talk iambic lines occur oftener than any others: but in a speech we need 
dignity and the power of taking the hearer out of his ordinary self. The trochee 
is too much akin to wild dancing: we can see this in tetrameter verse, which is 
one of the trochaic rhythms. (Aristotle 3.8, 1408b)

Aristotle had, then, considerable interest in the perceived effects of poetic 
rhythm. I will argue that his assertions about the iambic are doubtful; but 
about the rest, I tend to agree with him. Generations of critics during the past 
two millennia repeated Aristotle’s assessment of the iambic with reference 
to many languages. Roger Fowler (1966) was the first one to point out that 
this assertion fails to conform with the facts. “This, paradoxically, may help 
to explain why the iambic measure is felt to be suited to English: not because 
its pattern corresponds to the prose rhythms of language, for it does not; but 
because it necessitates a constant syncopation of prose rhythm against its own 
rhythm, inviting poets to be metrically complex, not to jog along with simple 
regularity” (1966: 99). Moreover, as the great Hungarian poet, Miklós Radnóti 
(1943) pointed out in the Afterword to his volume of poetry translations, even 
Hungarian poets consider the iambic as the most natural metre (even though 
in Hungarian, stress falls invariably on the first syllable of the word). 

In my chapter on the iambic and the trochaic in my book (Tsur 1977; 
reprinted in Tsur, 2017) I quote H. Woodrow, who found in his tick-tack 
experiments, back in the beginning of the Twentieth Century, that in a series 



30 Reuven Tsur

of tick-tacks, “with equal temporal spacing, a regularly recurring, relatively 
greater intensity exerts a group-beginning effect, and a regularly recurring, 
relatively greater duration a group-ending effect” (Woodrow 1951: 1233). 
“Intensity has a group-beginning effect: duration, a group-ending effect: pitch, 
neither a group-ending nor a group-beginning effect” (Woodrow 1911: 77). In 
his doctoral dissertation, Curt Rice (1992) replicated Woodrow’s experiments. 
“The technological resources for conducting this research are dramatically 
more sophisticated than those which Woodrow had available”. At variance 
with Woodrow, Rice (1992: 198) showed that variations in pitch do lead to a 
significant shift towards iambic groupings. Coming back to verbal rhythms, 
spoken language consists of syllables of varying duration. Schramm (1935), 
Fry (1958), and Chatman (1965) demonstrated that the acoustic cue to lin-
guistic stress is a complex of pitch, duration, and loudness, in that order of 
decreasing effectiveness. Usually, at least two of these three components are 
present. If pitch differences are irrelevant to grouping direction and duration 
differences are more effective in stress perception than amplitude differences, 
end-accented meters should be more natural in poetry in a variety of lan-
guages. If variations in pitch also lead to a significant shift towards iambic 
groupings, it should reinforce this effect. 

Aristotle had in mind a metric system based on duration differences. In 
English and in many other modern Western languages there is a prevalent 
intuition that in syllabotonic metre too, the iambic is more natural than the tro-
chaic (since Fry found that duration is a more effective cue for stress perception 
than intensity, this too is explained by Woodrow’s findings). In a completely dif-
ferent cultural tradition and different metric systems, in Mediaeval Arabic and 
Hebrew poetry in Spain, based on the alternation of longer and shorter units, 
there is a huge number of metres available to poets. I have found that, never-
theless, Mediaeval Hebrew poetry is dominated by one end-accented metre 
(Tsur, Bentov 1996); Golston and Riad (online) found that Arabic poetry too 
is dominated by end-weighted metres. I hypothesized that this is the result of 
some natural selection, to conform with the constraints of the cognitive system. 

In this context, however, there is a disconcerting issue. The natural selection 
suggested here works well in some poetic traditions, but other traditions force 
rigid limitations upon poets and prevent them from preferring the metrical 
patterns that have a good fit to the natural constraints of cognitive economy. 
For instance, the venerable Greek and Roman epic tradition is dominated by 
the dactylic hexameter (“the heroic rhythm”). The dactylic foot is the least 
natural of the most frequent feet; namely, the iambic, the trochee, the ana-
pest and the dactyl. The trochaic and the dactylic feet are relatively unnatural 
because the strong position (longer syllable) precedes the weak one(s). The 



31Metre, rhythm and emotion in poetry

dactylic is less natural than the trochaic, because two weak positions lean back 
upon one strong position. This would conform with Aristotle’s description of 
these metres. Why should, then, precisely that rhythm prevail and multiply 
that has a poor fit to the constraints of the cognitive system, that is, is the least 
natural of the rhythms? I found the solution to this riddle only recently, when 
I re-read, for a different purpose, Aristotle’s paragraph.

Aristotle claims that “of the various rhythms, the heroic has dignity” and 
“the power of taking the hearer out of his ordinary self ”. In other words, the 
heroic action does not demand the most natural rhythms, but those that take 
the hearer out of his ordinary self – i. e., precisely the least natural metre. This 
is how the most “dignified” rhythm is generated, so as to support the intense 
human quality of the epic. The “heroic rhythm” (the dactylic hexameter) is 
the most dignified one because of its deviation from the commonplace “tones 
of the spoken language”. Poor fit may have its own expressive value. One may 
even assume that poets sometimes deliberately seek out the conventions that 
have a poor fit, for effects that the Structuralists would call “marked”. Now 
notice this too: In what Aristotle calls “the heroic rhythm”, the dactylic hexam-
eter, in most parts of the verse line a spondee may be substituted for a dactylic 
foot. In other words, the heroic rhythm consists of two kinds of feet that, for 
different reasons, may suggest a dignified quality. 

Stress Maxima in Weak Position

One of the unique features of my work in prosody is my handling of stress 
maxima in weak position. I will try to explain this in a nutshell, and then point 
out its emotive implications. 

Morris Halle and Samuel Jay Keyser (1971) proposed in their generative 
theory an exceptionally parsimonious criterion for distinguishing all “metri-
cal” lines from all “unmetrical” ones, assuming that this criterion is somehow 
internalized by poets and readers. The iambic pentameter line consists of 
an abstract pattern of regularly alternating weak and strong positions, upon 
which the sequences of linguistic stresses are “mapped”. The “stress maxi-
mum” is a theoretical construct, defined as “a stressed syllable between two 
unstressed ones, within the same line and the same syntactic constituent”. 
The phrase “a gárden” contains a stress maximum; “a bíg gárden” contains no 
stress maximum, because neither of the two stressed syllables occurs between 
two unstressed ones – they “neutralize” each other. All mappings are “allow-
able”, except one: a stress maximum in a weak position, that renders the line 
unmetrical, as in the following lines: 



32 Reuven Tsur

               7
14. How many bards gild the lapses of time!
 (Keats, “How many bards gild the lapses of time”)
                  7
15. And made him bow to the gods of his wives (PR, 171)
             7
16. Burnt after them to the bottomless pit
 Milton, Paradise Lost, VI. 865–866
      7
17. With them from bliss to the bottomless deep (PR, I. 361)
     7
18. And whelm on them to the bottomless void 
 (Shelley, Prometheus Unbound, III. i. 76)
          7
19. Buffet and scoffe, scorge, and crucifie mee (Donne, Holy Sonnet 11. 2)

I have found that the distribution of the violating stress maxima is far from 
random. Over sixty instances of stress maxima in weak positions have been 
recorded in iambic pentameter lines, in major English poetry. Two thirds of 
them occurred in the seventh position (out of four positions available for 
violation). This suggests that the distribution is not random, and that poets 
like Milton, Shelley and Keats assumed that their readers can perform such 
lines rhythmically. I predicted that experienced readers will tend to agree that 
a verse line with a stress maximum in the seventh position can be performed 
rhythmically. The solution to the perceptual problem will arise not with refer-
ence to the isolated stress maximum, but within a larger group. Such groupings 
are the performance patterns available to the reciter (Halle and Keyser declare 
that they have nothing to say about how such lines should be performed). If 
a reader is requested to read the line rhythmically, so as to preserve the stress 
pattern of the words and as much of the metre as possible, he is likely to group 
together, emphatically, the last four syllables and segregate the group from the 
preceding context. In other words, he will foreground in his performance a 
unitary perceptual group called “stress valley”. A stress valley is a cluster of 
four syllables; in terms of Gestalt theory it has a closed symmetrical (that is, 
“good”) shape: two stressed syllables embrace two unstressed ones, as in “Píty 
the wórld” or “bóttomless pít”. Such a perceptual organization may save mental 
processing space, needed for the perception of the conflicting linguistic and 
versification patterns simultaneously. A stress valley that begins in the seventh 
position ends in the tenth position, imposing upon the line an exceptionally 
strong closure. I have found that experienced readers of poetry did indeed tend 



33Metre, rhythm and emotion in poetry

to foreground such a stress valley; they were surprised to discover that rather 
than playing down the violating stress, they all tended to overemphasize it.

In what follows, I will give first an example of what I don’t mean by “the 
expressive potential of stress valley”, and then briefly suggest my conception 
of it, when initiated by a stress maximum in the seventh position. Consider 
Excerpt 14 above. The first syllable of lapses in this line constitutes a stress 
maximum in a weak position; according to the Halle-Keyser theory, this ren-
ders the line unmetrical. Halle and Keyser attempt to “save” this line as a 
“metrical lapse”, a kind of onomatopoeia, a metric joke. “The line is literally 
what it speaks of figuratively, a ‘lapse of time’” (1971: 171). According to the 
conception of performance outlined here, confirmed by a great number of 
readings, such configurations as the four syllables of lapses of time can be 
performed rhythmically, by grouping them together. This renders the line 
acceptable, and the ad hoc explanation of “metric joke” becomes superfluous. 
Keats and Halle and Keyser use “lapses” in different senses and the “metric 
joke” construal of the deviation is far-fetched. 

My solution to the problem relies on Leonard B. Meyer’s discussion of emo-
tion and meaning in music (1956). Emotion or affect is aroused, Meyer says, 
when a tendency to respond is arrested or inhibited (14). The pleasantness of 
an emotion seems to lie not so much in the fact of resolution itself as in the 
belief of resolution – the knowledge, whether true or false, that there will be 
a resolution (19). That is why we can know whether an emotion is pleasant or 
unpleasant before it is actually over (that is, before apprehension is dispelled, 
or stability is achieved). “The sensation of falling through space, uncondi-
tioned by any belief or knowledge as to the ultimate outcome, will, for instance, 
arouse highly unpleasant emotions. Yet a similar fall experienced as a para-
chute jump in an amusement park may, because of our belief in the presence 
of control and in the nature of the resolution, prove most pleasurable” (20). 
Now a stress valley beginning with a stressed syllable in a weak position and 
ending in a strong position at precisely the most stable points of the pentam-
eter line (in the tenth or the fourth positions), may initiate an experiencing of 
such a “pleasurable apprehension”. Loss of control is experienced during three 
syllables; but one also knows that control and stability will be regained at the 
fourth unit, where the stress pattern and the metric pattern have a coinciding 
downbeat, and the metre becomes fresh and new. In the verse lines in which 
the first syllable of “bottomless” occurs in the seventh position, anxiety is much 
stronger, there is a greater danger that the line will return to chaos, than when 
a stress valley begins in the first position, because it involves a stress maximum 
in a weak position; thus, what applies to a stress valley beginning in the first 
position is much more forcefully valid here. 



34 Reuven Tsur

How does a listener know, before the stress valley is completed, that there will 
be a coinciding downbeat? Speech processing is governed, at any given time, by 
immediate memory and anticipation. To give just one example of anticipation, 
[d] is a voiced stop; its enunciation consists of a gesture articulating a [t] and 
the activation of the vocal folds. This requires the prolongation of the preceding 
vowel, to foreground the contrast between a (voiced) vowel and a voiced stop. 
When a vowel is prolonged, the listener anticipates a voiced stop and hears a [d] 
even if actually a [t] is articulated. Likewise, Cooper and Meyer observe that “in 
musical performance, the placing of some extra accent may affect the grouping 
of sounds. […] [A]n accent on a weak beat, presents the group as end-stressed” 
(1960: 8). When a reciter overemphasizes (rather than downplays) a stress maxi-
mum in the seventh (weak) position, he expects an exceptionally strong stress 
in the tenth (strong) position. And so does the listener. 

The stress valley has no semantic meaning, only perceptual dynamics. But 
this dynamics may generate a combinational potential, that is, a potential 
to combine with semantic elements that are relevant in one way or other to 
that dynamics. “Downward movement, falling”, for instance, has an element 
of loss of control; but in the present instances, this loss of control may be 
pleasurable, because the reader believes that the stress valley and the pairs of 
iambic feet will have a coinciding downbeat, where stability and control will 
be regained – even though the “pit”, the “void”, the “deep” are “bottomless”; 
that is, control can be regained in the stress valley only, not in the endless fall. 
It makes us perceive the loss of control but, by the same token, it gives false 
security, to face it. 

Applying, then, Leonard B. Meyer’s phrases to stress maxima in the sev-
enth position, this infringement inspires the reader with “awe, apprehension, 
and anxiety” that the utterance may escape back to chaos, “arousing pow-
erful desires for, and expectations of, clarification and improvement”. These 
desires and expectations are fulfilled, precisely, in the last position of the line, 
generating a strong feeling of closure. The line becomes well shaped and, para-
doxically, at the same time, near-chaotic. Hence its strong emotional impact. 
The feeling of uncertainty, of “anxiety”, as it were, is felt in the active violation 
of metre by an unsupported stressed syllable followed by unstressed syllables 
of such polysyllables as crúcifie or bóttomless, where they must lean back on 
that broken reed of a stress maximum in the seventh position, or one must 
wait until regularity is reasserted again, in the tenth position.

In Excerpts 16, 17, 18 above, bót- is a stress maximum in a weak position, 
adding momentum to the feeling of “anxiety”. Considering the relative scarcity of 
stress maxima in weak positions, these examples – whether independently arrived 
at, or imitated by Shelley from Milton – cannot be dismissed as insignificant. 



35Metre, rhythm and emotion in poetry

They are the more remarkable since the poets could easily have avoided the stress 
maximum in a weak position – by using “to bottomless abyss”, for example. 

I have emphasized above that the pattern illustrated here may appear to 
have an iconic impact, because what we have is not exactly iconic; it is more 
accurately handled in terms of “individuation of a general psychological 
atmosphere”. The “falling” analogue is not part of the metric pattern; it is 
merely a metaphor to suggest a peculiar psychological atmosphere of inse-
curity and anxiety, which are perceptual qualities of the metric shape under 
discussion. This metric shape is associated with a feeling of momentum and 
craving for stability that may combine with any content and lend impetus to 
such expressions as “crucifie mee”, or almost any other instance quoted here. 
When it combines with “bottomless pit” and its near-synonyms, the feeling of 
insecurity and anxiety and the craving for stability may combine with the con-
tent of the line, reinforcing each other. In Halle and Keyser’s handling of “the 
lapses of time” the parallel between metric structure and contents crucially 
depends on the words by which we describe the metric figure: if we describe 
it as “metric lapse”, the parallel exists; if we describe it as “metric deviance”, it 
does not exist. Moreover, it also depends on assigning to the word “lapses” a 
meaning that is different from the one suggested by the context. My discussion 
of “the bottomless pit”, by contrast, concerns an atmosphere of anxiety gener-
ated by the loss of control caused by the extreme metric deviation, “arousing 
powerful desires for, and expectations of, clarification and improvement”. This 
is a possible source of a sense of “momentum” seeking “focal stability”. 

Tarlinskaja and the Stress Valley

I have said that Marina Tarlinskaja’s work in metrics is diametrically opposed 
to mine in many respects; among others, it is strongly meaning-oriented 
and confirmatory-quantitative. Perceived effects are not within the scope of 
her research. It is based “not so much on insight and intuition as on wide 
quantitative analyses of observable facts”. In spite of this, our findings are con-
siderably overlapping, but with quite different implications. In her 1987 paper 
“Rhythm and Meaning: ‘Rhythmical Figures’ in English Iambic Pentameter, 
Their Grammar, and Their Links with Semantics” she touches on issues which 
I treat under the heading of “stress valley” (but do not involve a stress maxi-
mum in a weak position). 

Consider the first two examples from a group of lines quoted by Tarlinskaja 
(1987: 20), in my metric notation: “To stand in thy affairs, fáll by thy síde” (Sh. W         . S        W    S



36 Reuven Tsur

Son. 151.12); “fáll like amázing thunder on the casque”(Sh. R2 1.3.81), and so 
forth. I have marked in these lines the linguistic stresses and the underlying 
weak and strong positions in a stretch of four syllables. In the terminology of 
the present study these are “stress valleys” beginning in the first or seventh 
(weak) position and ending in the fourth or tenth (strong) position. While in 
my work the group of four constitutes the unit of analysis, Tarlinskaja refers 
only to the first two syllables of these groups as a metric figure, labelling them 
as WS, that is, a stressed syllable in a weak position and an unstressed sylla-
ble in a strong position. It would appear that Tarlinskaja was interested here 
only in the deviating portion of the line. Rhythmical figures may stretch for 
her over as far as 5 and more syllables; but in these instances she focuses on 
the two deviating syllables only. By contrast, I was interested in the percep-
tual unit in which the deviance is accommodated. She is concerned with the 
observable facts, not with how such figures should be rendered acceptable in 
an iambic context. This difference is in perfect harmony with her statistical, 
and my cognitive approach. Tarlinskaja found by statistical means what are 
the meanings typically associated with this figure. One of them is, she says, 
“motion downward”. All the examples she quotes in this group contain the 
word “fall” or “fell” in a weak position. But “motion” and “falling” is just one 
of the most frequent semantic areas coupled with this rhythmical figure; the 
verb is frequent, but not obligatory; other parts of speech may appear, but 
the context may still be “motion”, and the verb of motion may be absent, 
or placed outside the figure. Tarlinskaja offers the following explanation for 
this association of “downward motion, falling” with this metrical figure: “The 
semantic component ‘falling’, so frequent in the figure WS-1, probably has 
something to do with the accentual-syllabic structure of the figure itself, which 
usually begins a verse line or a phrase within the line: it is a stressed syllable 
followed by two unstressed ones; possibly this arrangement does produce a 
physical impression of a heavy weight falling down”. It will be remembered 
that with reference to “bóttomless pít” and its near synonyms in Milton and 
Shelley I argued that the metric figure with a stressed syllable followed by two 
unstressed ones, beginning in a weak position, does not contain the falling 
analogue, but displays a general sense of loss of control; the fourth, stressed 
syllable in a strong position, in turn, yields (false) security that enables to face 
the anxieties associated with endless fall. The word “fall” does not occur in 
these lines, but the violent downward motion is there. It is rather the adjective 
“bottomless” that occupies the critical metrical positions, reinforced by the 
metric figure’s sense of loss of control.

The main difference between Tarlinskaja’s and my own approach seems to be, 
in this respect, that while she conjectures some hidden link between the metric 

W        S         W      S

W      S         W            S



37Metre, rhythm and emotion in poetry

figure and the specific meaning “vigorous fall”, I discern a general perceptual 
dynamics of “loss of control” that can be individuated by association with a vari-
ety of meanings. The present conception assumes that no amount of “observable 
facts” in the text may suggest the nature of that dynamics. For this, one must 
assume the immediate responses of a human perceiver. It is also obvious that the 
answer should be sought in the structure of signs that have no predetermined 
semantic meaning. Consequently, Leonard B. Meyer’s analysis of emotion and 
meaning in music (1956) as discussed above may be the clue to this issue. 

Tarlinskaja quotes 17 lines that contain such figures, with “fall” or “fell” in 
the critical weak position, by poets ranging from Shakespeare, through Pope, 
Swift, Byron, Shelley, Wordsworth, to Arnold. 13 of the resulting stress valleys 
end in the fourth position, 4 in the tenth. One of the latter does not belong 
to this discussion, because it occurs in an iambic hexameter line, beginning 
immediately after the caesura. In another group, with the verbs “fly”, “run”, 
“rush” (that is, indicating vehement movement away) Tarlinskaja quotes 8 
lines, by Shakespeare and Shelley; in 5 of them the stress valley ends in the 
fourth position, in 2 in the tenth position, and only in 1 in the eighth position. 
Thus, out of 24 pentameter lines quoted in these two groups, only in one line 
(by Shelley) the stress valley ends in a position other than fourth or tenth (that 
is, at the unmarked caesura and the line ending). 

To sum up. I have propounded a qualitative “perceived effects” approach 
to the contribution of meter and rhythm to emotion in poetry. I have briefly 
compared it with John A. Jones’s disguised impressionistic approach, and to 
three thoroughly systematic works: Marina Tarlinskaja’s large-scale statistical 
“meaning-oriented” analysis, Benjamin Hrushovski’s structuralist work focus-
ing on four types of logical relationship between sound and meaning that, 
nevertheless, implicitly acknowledges that speech sounds may have inherent 
(sometimes conflicting) perceived effects; and to Maria Kraxenberg’s quantita-
tive study of sound–emotion associations. One of my main assumptions was 
that sound effects cannot be reduced to meanings: they can only generate some 
general psychological atmosphere that may be individuated by meaning into 
a variety of more specific emotional qualities. While emotional meanings are 
subject to understanding, emotional qualities are subject to immediate per-
ception. The emotional quality of a poem consists in a structural resemblance 
between some emotion and a poetic text. We have explored the contribution 
of meter and rhythm to this resemblance. 



38 Reuven Tsur

References

Aristotle. The Rhetoric of Aristotle, tr. W. Rhys Roberts. http://libertyonline.hypermall.
com/Aristotle/Rhetoric/Rhetoric.html

Arnheim, Rudolf 1967. Art and Visual Perception. London: Faber. 

Beardsley, Monroe C. 1958. Aesthetics: Problems in the Philosophy of Criticism. New 
York, Burlingame: Harcourt, Brace & World. 

Berkovich-Ohana, Aviva; Glicksohn, Joseph; Ben-Soussan, Tal Dotan; Goldstein, 
Abraham 2016. Creativity is Enhanced by Long-Term Mindfulness Training and 
is Negatively Correlated with Trait Default-Mode-Related Low-Gamma Inter-
Hemispheric Connectivity. In: Mindfulness 8(3), 717–727.

Bouwsma, Oets Kolk 1961. The Expression Theory of Art. In: Philipson, Morris (ed.), 
Aesthetics Today. Cleveland & New York: Meridian Books, 145–168. 

Chatman, Seymour 1965. A Theory of Meter. The Hague: Mouton. 

Duffy, Elizabeth 1968 [1941]. An Explanation of Emotional Phenomena without the 
Use of the Concept Emotion. In: Arnold, Magda B. (ed.), The Nature of Emotion. 
Harmondsworth: Penguin, 129–140. 

Fowler, Roger 1966. ‘Prose Rhythm’ and Metre. In: Fowler, Roger (ed.), Essays on Style 
and Language. London: Routledge & Kegan Paul. 

Fry, Dennis B. 1958. Experiments in the Perception of Speech. In: Language and Speech 
1, 126–151. 

Glicksohn, Joseph; Tsur, Reuven; Gootblatt, Ch. 1991. Absorption and Trance-
Inductive Poetry. In: Empirical Studies of the Arts 9(2), 115–122. 

Golston, Chris; Riad, Tomas. The Phonolog y of Classical Arabic Meter.  
http://zimmer.csufresno.edu/~chrisg/index_files/ArabicMeter.pdf

Guilford, Joy Paul 1970 [1959]. Traits of Creativity. In: Vernon, Philip Ewart (ed.), 
Creativity. Harmondsworth: Penguin, 167–188.

Halle, Morris; Keyser, Samuel Jay 1971. English Stress: Its Growth and Its Role in Verse. 
New York: Harper and Row.

Hepburn, Ronald W. 1968. Emotion and Emotional Qualities. In: Osborne, Harold 
(ed.), Aesthetics in the Modern World. London: Thames and Hudson, 81–93. 

Hrushovski, Benjamin 1980. The Meaning of Sound Patterns in Poetry: An Interaction 
View. In: Poetics Today 2(1a), 39–56.



39Metre, rhythm and emotion in poetry

Jakobson, Roman 1960. Closing Statement: Linguistics and Poetics. In: Sebeok, 
Thomas A. (ed.), Style in Language. Cambridge, Mass.: The MIT Press, 350–377.

Jakobson, Roman; Fant, C. Gunnar M.; Halle, Morris 1952. Preliminaries to Speech 
Analysis: the Distinctive Features and their Correlates. Cambridge, MA: The MIT 
Press.

Margolis, Joseph 1962. The Logic of Interpretation. In: Margolis, Joseph (ed.), 
Philosophy Looks at the Arts: Contemporary Readings in Aesthetics. New York: 
Charles Scribner’s Sons, 108–118. 

Meyer, Leonard B. 1956. Emotion and Meaning in Music. Chicago: Chicago University 
Press.

Radnóti, Miklós 1943. Orpheus Nyomában. Budapest: Pharos.

Schramm, Wilbur L. 1935. Approaches to the Science of English Verse. Iowa City: Iowa 
University. 

Sibley, Frank 1962. Aesthetic Qualities. In: Margolis, Joseph (ed.), Philosophy Looks at 
the Arts: Contemporary Readings in Aesthetics. New York: Scribner, 63–88. 

Snyder, Edward D. 1930. Hypnotic Poetry: A Study of Trance-inducing Techniques 
in Certain Poems and its Literary Significance. Philadelphia: University of 
Pennsylvania Press. 

Sweetser, Eve 2006. Whose Rhyme is whose Reason?: Sound and Sense in Cyrano de 
Bergerac. In: Language and Literature 15(1), 29–54.

Tarlinskaja, Marina 1987. Rhythm and Meaning: “Rhythmical Figures” in English 
Iambic Pentameter, Their Grammar, and Their Links with Semantics. In: Style 
21(1), 1–35.

Tellegen, Auke 1981. Practicing the Two Disciplines for Relaxation and Enlightenment: 
Comment on ‘Role of the Feedback Signal in Electromyograph biofeedback: 
The Relevance of Attention’ by Qualls and Sheehan. In: Journal of Experimental 
Psychology: General 110(2), 217–226. 

Todd, Henry J. 1970 [1809]. The poetical works of John Milton, with notes of various 
authors. London: Law and Gilbert.

Tsur, Reuven 1978. Emotions, Emotional Qualities and Poetry. In: Psychocultural 
Review 2, 165–180.

Tsur, Reuven 2006. Delivery Style and Listener Response in The Rhythmical 
Performance of Shakespeare’s Sonnets. In: Hogan, Patrick Colm; Pandit, Lalita 
(eds.), Cognitive Shakespeare: Criticism and Theory in the Age of Neuroscience 
(College Literature 33(1)), 170–196. 



40 Reuven Tsur

Tsur, Reuven 2008a. Comparing Approaches to Versification Style in Cyrano de 
Bergerac. In: Cognitive Semiotics 2, 146–168 

Tsur, Reuven 2008b. Toward a Theory of Cognitive Poetics. Second, expanded and 
updated edition. Sussex Academic Press: Brighton and Portland.

Tsur, Reuven 2017. Poetic Conventions as Cognitive Fossils. Oxford: Oxford University 
Press.

Tsur, Reuven; Bentov, Yehosheva 1996. The Rhythmic and Strophic Organization of 
Mediaeval Hebrew Poetry (A Cognitive Approach). In: Empirical Studies of the 
Arts 14, 183–206.

Wellek, René; Warren, Austin 1956. Theory of Literature. New York: Harcourt, Brace 
& Co. 

Wimsatt, William K.; Beardsley, Monroe C. 1959. The Concept of Meter: An Exercise 
in Abstraction. In: PMLA 74(5), 585–598. 

Woodrow, Herbert 1911. The Role of Pitch in Rhythm. In: Psychological Review 18(1), 
54–77.

Woodrow, Herbert 1951. Time Perception. In: Stevens, Stanley S. (ed.), Handbook of 
Experimental Psychology. New York: Wiley, 1234–1236.

Youmans, Gilbert 1989. Milton’s metre. In: Phonetics and Phonology 1: Rhythm and 
Meter. New York: Academic Press, 341–379.