Professional Documents
Culture Documents
Series Editors:
Alexander Bergs, University of Osnabrück
Margaret H. Freeman, Myrifield Institute for Cognition and the Arts
Peter Schneck, University of Osnabrück
Achim Stephan, University of Osnabrück
Advisory Board:
Mark Bruhn, Regis University Denver, CO, USA
Peer Bundgard, Aarhus University, Denmark
Michael Burke, University College Roosevelt Middelburg, the Netherlands
Wallace Chafe, University of California Santa Barbara, USA
Barbara Dancygier, University of British Columbia Vancouver, Canada
Frank Jäkel, Universität Osnabrück, Germany
Winfried Menninghaus, Freie Universität Berlin, Germany
Keith Oatley, University of Toronto, Canada
Jan Slaby, Freie Universität Berlin, Germany
Reuven Tsur, Tel Aviv University, Israel
Mark Turner, Case Western Reserve University Cleveland, OH, USA
Simone Winko, Georg-August-Universität Göttingen, Germany
Dahlia Zaidel, University of California Los Angeles, USA
Cognitive Approaches to Early Modern Spanish Literature
Isabel Jaén and Julien Jacques Simon
Cognitive Literary Science: Dialogues between Literature and Cognition
Edited by Michael Burke and Emily T. Troscianko
A Prehistory of Cognitive Poetics: Neoclassicism and the Novel
Karin Kukkonen
Expressive Minds and Artistic Creations
Szilvia Csábi
Poetic Conventions as Cognitive Fossils
Reuven Tsur
iii
Poetic Conventions
as Cognitive Fossils
Reuven Tsur
1
iv
1
Oxford University Press is a department of the University of Oxford. It furthers
the University’s objective of excellence in research, scholarship, and education
by publishing worldwide. Oxford is a registered trade mark of Oxford University
Press in the UK and certain other countries.
9 8 7 6 5 4 3 2 1
Paperback printed by WebCom, Inc., Canada
Hardback printed by Bridgeport National Bindery, Inc., United States of America
v
CONTENTS
Preface vii
Preface by the Author xi
Acknowledgments xvii
About the Companion Website xix
References 261
Index 271
vi
vi
P R E FA C E
In developing his pioneering theory of cognitive poetics over the past half
century, Reuven Tsur has embraced an astounding range of research and
approaches, including Gestalt psychology, Russian formalism, analytic phi-
losophy, stylistics, and theories of prosody. Even more remarkable is the
scope of the poetic forms, genres, and strategies to which he has applied his
theoretical and conceptual methodologies. They include approaches both
to basic themes and to aspects of rhyme, genre, archetypal patterns, period
style, and sound symbolism. In his books and articles Tsur has applied his the-
ories to English, French, German, Hungarian, Italian, Spanish, and Hebrew
poetry, ranging from the Bible through medieval to Renaissance, classic to
Romantic, and modern to contemporary poetry. Chapters in his monumen-
tal work Toward a Theory of Cognitive Poetics, now in its second expanded and
revised edition, cover issues that include the intricacies of poetic rhythm,
the complexities of poetic metaphor, meditative poetry and altered states
of consciousness, translation theory, the implied critic’s decision style, and
the thorny question of critical competence. Poetic Conventions as Cognitive
Fossils draws its illustrations from Tsur’s research into the poetries of Middle
and Modern English; Old, Middle, and Modern French; German; Hungarian;
and Medieval and Modern Hebrew. And it draws its intellectual foundations
from literary criticism, linguistic theory, phonology, cognitive psychology,
and anthropology, to name only the most prominent of its resources.
Tsur defines cognitive poetics as a theoretical methodology that explains
how poetic language and literary form are shaped and constrained by
human cognitive processes. In his contribution to Brône and Vandaele’s
book Cognitive Poetics: Goals, Gains, and Gaps, Tsur begins with an anecdote
about human information processing:
There was an old joke in Soviet Russia about a guard at a factory gate who at the
end of every day saw a worker walking out with a wheelbarrow of straw. Every
vi
day he thoroughly searched the contents of the wheelbarrow, but never found
anything but straw. One day, he asked the worker: “What do you gain by taking
home all that straw?” [The worker responded:] “The wheelbarrows.” (2009, 23)
Tsur’s point in telling the joke was to introduce the basic gestalt rules of fig-
ure and ground. But the joke certainly also resonates with Tsur’s own life
and work in finding unconventional and innovative ways to accomplish one’s
objectives. Tsur has always been smuggling out wheelbarrows, while the rest
of us were examining the straw. Following are just some of the many theo-
retical and methodological wheelbarrows he has brought out over the years.
Noam Chomsky aimed to develop a rule system for language that would
generate all and only grammatical sentences. In an article titled “Chaucer
and the Study of Prosody” published in 1966, Morris Halle and Jay Keyser
set out to do something similar for English poetry, to develop a rule sys-
tem that would generate all and only metrical lines. Having noticed that
the occurrences of many unmetrical lines in English poetry, according to
their theory, shared certain regularities, Tsur instead developed a theory
of rhythmical performance that focuses on explaining poetic effects, cul-
minating in the publication of Poetic Rhythm: Structure and Performance
(1998). Similarly, whereas linguistic research, in analyzing poetic meter,
applied phonological stress rules to distinguish rhythmic grouping in iam-
bic and trochaic feet, Tsur focused instead on their versification patterns.
His findings explain more precisely the tendency in English poetry toward
iambic groupings. His perception-oriented theory of meter is based on
empirical investigation using the latest technological resources.
Tsur’s ability to place linguistic studies within the larger perspective of
research in other disciplines and his unwavering focus on the nature of the
poetic text distinguish his work from other cognitive research. Cognitive
metaphor studies, for example, inspired by the work of George Lakoff,
Mark Johnson, Gilles Fauconnier, and Mark Turner, have tended, with
some notable exceptions, to focus on the metaphorical underpinnings of
conventional cognitive processes. For that very reason, such studies have
not been designed to illuminate the qualities that distinguish poetry from
prose or to explain how poets can use the same metaphor with different
effects. In contrast, Tsur provides convincing explanations for the differ-
ences among poetic metaphors that depend on such aspects as their discor-
dant versus concordant elements, low versus high differentiation, and split
versus integrated focus and the literary effects of synaesthetic transfer.
In Poetic Conventions as Cognitive Fossils, Tsur draws upon much of his
past research to show how conventional poetic styles originate from cog-
nitive rather than cultural principles. Whereas traditional scholarship
[ viii ] Preface
ix
Alexander Bergs
Margaret H. Freeman
Peter Schneck
Achim Stephan
P r e fac e [ ix ]
x
xi
P R E FA C E B Y T H E A U T H O R
One of the most exciting minutes in one’s research record is when one
discovers that a wide variety of one’s writings have grown together into
one coherent whole. That is what I experienced when I wrote my Toward a
Theory of Cognitive Poetics (Tsur 2008a): I discovered that all my work over
the decades had grown together into a coherent, comprehensive theory of
cognitive poetics. I had a similar experience when, in writing this book, a
subset of my work coalesced into a coherent subfield of cognitive poetics
that conceives of poetic conventions as cognitive fossils shaped and con-
strained by cognitive constraints, rather than accounting for them through
an infinite regress of ever-earlier influences.
SYNOPSIS OF THIS BOOK
This book explores the questions of where conventions come from, how
they work, how they could be understood by their first audience, and how
they could become conventions. Chapter 1 presents a “constraints-seeking”
or “cognitive-fossils” approach to conventions as against the prevalent
“migration,” “influence-hunting,” or “culture-begets-culture” approach.
Two of its main assumptions are derived from Roy D’Andrade and
Anton Ehrenzweig. According to the former, cultural programs (in this
case poetic inventions) may become conventions when in the course of
repeated social transmission they take on forms that are consistent with
the natural capacities and constraints of the human brain. According to
the latter, in order to avoid the more “dangerous” expressive elements of
an artistic work, we tend to neutralize expressive features by formalizing
them—we turn them into characteristics of style, into harmless ornament.
The culture-begets-culture approach, by contrast, accounts for conventions
by pointing at earlier occurrences in other cultures, assuming that the
xi
earlier occurrence influenced the later one, merely transferring the mys-
tery from one place to another.
I use the term fossils in a weaker and a stronger sense. In the weaker
sense, for example, such verbal devices as interjections, exclamations,
vocatives, and repetitions have emotive force owing to their relation to
processing in the right hemisphere of the brain. In Elizabethan drama,
interjections followed by repeated vocatives became a hackneyed conven-
tion that affects the audience’s beliefs about the character’s emotion rather
than a perception thereof. In the stronger sense, in the ballad “Edward”
the same conventions turn into a rigid formula that is perceived as “ballad
style” rather than expressive devices: they affect the reader’s beliefs about
the genre rather than about the speakers’ emotions.
The opposition between the constraints-seeking and influence-hunting
approaches may be illustrated through their handling of caesurae.
According to the cognitive approach, the rhythmic processing of poetry is
constrained by the limited channel capacity of the cognitive system. This
has shaped a wide range of versification conventions in a number of lan-
guages and versification systems. Thus, in the English iambic pentameter
line, the caesura may occur after the fourth, fifth, or sixth position (out of
ten), but the vast majority occur after the fourth. The influence-hunting
approach attributes this distribution to the influence of French poetry,
where the majority of caesurae occur after the fourth position in deca-
syllabic lines. The cognitive-constraints approach, by contrast, assumes
that owing to the limited channel capacity of immediate memory, when
there are two parallel segments of unequal length, the longer coming last
is felt to be the well-shaped order in both English and French poetry,
and in many other languages, as well as in music. The influence-hunting
approach cannot explain why French poetry places the caesura after the
fourth position or why English poets have adopted this practice in a dif-
ferent versification system—it merely transfers the mystery from one
place to another.
In c hapter 2 I elaborate on the implications of Roy D’Andrade’s concep-
tion that in the course of repeated social transmission cultural programs
take on forms that are consistent with the natural capacities and con-
straints of the human brain. I explore two crucial cognitive constraints and
their literary implications: (1) The cognitive system has a limited channel
capacity and therefore demands great parsimony; and (2) language is typi-
cally logical and conceptual, whereas some poetry at least is supposed to
convey emotional or mystical qualities that are nonconceptual and illogi-
cal. As to repeated social transmission, I quote Bartlett, who staged a con-
trolled experiment with an Eskimo story that in the course of repeated
P r e fac e b y t h e Au t h or [ xiii ]
xvi
P r e fac e b y t h e Au t h or [ xv ]
xvi
ACKNOWLEDGMENTS
www.oup.com/us/tsurpoeticconventions/
Oxford has created a companion website that houses sound files accompa-
nying Chapters 7 and 8. The reader is encouraged to consult this resource.
x
xxi
CHAPTER 1
Tongs are made with (other) tongs, but who made the original [pair of] tongs? It was,
perforce, a Heavenly creation!
—Babylonian Talmud, Tractate Pesachim, 54a1
COGNITIVE FOSSILS
I cannot doubt that language owes its origin to the imitation and modification,
aided by signs and gestures, of various natural sounds, the voices of other ani-
mals, and man’s own distinctive cries. . . . [W]e may conclude from a widely
spread analogy that this power would have been especially exerted during the
courtship of the sexes, serving to express various emotions, as love, jealousy,
triumph, and serving as a challenge to their rivals. The imitation by articulate
sounds of musical cries might have given rise to words expressive of various
complex emotions. (Darwin 1871: 56)
To give just one continuous example of the weaker and the stronger
senses to be elaborated later, one of the main problems facing poets is
that they have to convey information that is nonlogical and nonconceptual
through language, which is logical and conceptual. One of their preferred
tools for doing this is to transfer some of the linguistic information process-
ing to the right, emotional hemisphere of the brain. One possible device
for accomplishing this is to have recourse to interjections and vocatives,
arousing right-hemisphere information-processing activities. When this
practice became a poetic convention in much Elizabethan drama, it came
to affect the audience’s beliefs concerning the characters’ emotions rather
than stimulating direct perception of the emotional experiences. When this
convention was overused, it was parodied by Shakespeare in the Pyramus
and Thisbe interlude in A Midsummer Night’s Dream (see c hapter 2). Finally,
the vocatives and interjections became a rigid poetic formula in the ballad
“Edward,” in the stronger sense of the term cognitive fossils (see c hapter 4).
It would appear that such fossilizing processes are ubiquitous in human
experience and take place outside art and language too, in the social
domain. I propose to adduce here an illuminating, nonliterary example of
fossilization concerning customs that originated in solutions to very real
problems posed by the constraints of certain historical circumstances.
Those customs became detached in the course of repeated social transmis-
sion through the centuries from the causes that had prompted them, until
their practitioners had no idea why they would observe them at all.
Recently I met a young woman, originally from Spain, whose grand-
mother had observed a custom of lighting candles in the cupboard. It’s in
memory of the dead, she used to tell her. Her grandfather would go to the
butcher, buy pork meat, go home, and throw it out into the garbage. When
asked why he did this, he said that that was what his father had taught
him to do. Her mother was an inquisitive soul who eventually discovered
that they were descendants of the Marranos—Jews living in medieval
Spain who converted to Christianity but continued to practice Judaism
in secret. Those “New Christians” whose chimney was observed not to
smoke on Shabbat were subjected to the Inquisition. Obviously, they also
had to buy pork meat to demonstrate that they didn’t practice Judaism.
They preserved these customs over the centuries, even when there were
no persecutions anymore, even when they no longer remembered their
Jewish origins—when those practices had become mere fossils. This fam-
ily has settled now in Israel and converted to Judaism, and they are happy
orthodox Jews.
I am going to present three main points in the thesis of this book through
three passages quoted from different sources. The first one is taken from
The past twenty years of cultural studies have focused on the ways in which cul-
ture begets culture, ignoring the cognitive capacities that are the preconditions
for cultural change. We investigate the complex relations between the mind, the
world that both determines it and is determined by it, and the cultural forms
that arise from this interaction and feed back into it.2
“Culture begets culture” suggests infinite regress, just as tongs are made
with tongs. But who made the original cultural program? Or who made the
original (pair of) tongs?
The second passage is by the cognitive anthropologist Roy D’Andrade:
I will argue that “cultural programs” have solid cognitive foundations and
are shaped and constrained by the natural capacities and limitations of the
human brain, which results in certain significant regularities.3 I will also
argue that the “culture-begets-culture” paradigm cannot explain some of
the most significant aspects of cultural processes that are easily accounted
for by the “cognitive-constraints” paradigm. I will attempt to explore the
ways in which an infinite variety of cultural forms can arise in cultural pro-
grams constrained and shaped by the same cognitive capacities.
The third passage is derived from Ehrenzweig: “In music as in the visual
arts, the secondary elaboration succeeds in transforming a chaotic primi-
tive form into a rigid and ornamental style” (1953: 118). I will argue, then,
following Ehrenzweig, that there is a cognitive (or depth-psychological?)
mechanism of “secondary elaboration” that transforms chaotic forms
and stylistic devices that have “too much” emotive power into rigid orna-
ments, into distinctive characteristics of some style. The vocatives repeated
2. I assume that by “cultural studies” the paragraph means simply “the study of cul-
ture,” and not “a field of theoretically, politically, and empirically engaged cultural
analysis.”
3. In this sense, cultural programs include poetic conventions.
throughout “Edward,” for instance, lose their emotional force and may be
regarded as distinctive characteristics of the ballad style. Nevertheless, I do
not deny the validity of the “culture-begets-culture” paradigm; I simply argue
that it should not usurp the legitimate place of the “cognitive-constraints”
paradigm.
This book is devoted to the question of how cognitive processes shape
and constrain cultural and literary forms. It assumes that the generation
of cultural forms has to do with the deployment of devices that adapt to
the individual’s physical and social environment, whereas the response
to poetry, a literary form, involves adaptation devices turned toward aes-
thetic ends. The ensuing argument contrasts the view that “culture begets
culture” with the claim that the generation of culture is governed by adap-
tation devices exploited for cultural and aesthetic ends.
The theoretical choice between these two approaches is, to a very large
extent, a matter of prevalent intellectual and academic trends; these trends,
in turn, are no mere freaks of intellectual history but are related to more
general attitudes. According to Morris Weitz (1962), the role of theory in
aesthetics is to make a “crucial recommendation” as to what to look for
and how to look at it in works of art. In the present book I give two ways
of looking at cultural history in general and literary history in particular.
I suggest that the acceptance or rejection of one point of view or the other
is determined not only by prevalent intellectual trends but also by the crit-
ic’s cognitive complexity. The theoretical foundations for this conception
in terms of “perception,” “information processing,” and “personality style”
have been described in detail elsewhere (Tsur 1975, 1979, 2006c: 11–77).
The verb and the notion of “influence” are an illuminating case. I will
not review or document the controversies concerning this notion in the
twentieth century or the resulting crisis in comparative literature. Rather,
I will make only two brief comments: this verb entails a semantic-syntactic
anomaly, and the conflicting approaches discussed here are related to wider
issues—namely, to conflicting worldviews. Specifically, the use of the verb
influence is anomalous in this context and suggests that its subject is the
agent and its object is the recipient. When we say “A influenced B,” the sug-
gestion is that A is active, whereas B is passive. The “older” culture forces,
so to speak, the “younger” culture to accept some of its tenets. This is, in
fact, an authoritarian conception of cultural processes: namely, that the ear-
lier the culture, the greater its authority. This view denies B the possibility
of autonomous (let alone creative) action. In fact, the opposite is true: In
intercultural contacts B is active; it is B that selects the cultural elements it
needs from neighboring cultures to achieve its exigencies and yearnings.
Underlying such cultural exigencies and yearnings are tensions, pressures,
and wants generated within the system, whether cultural or cognitive. In this
sense, cultural change is motivated by a tendency to resolve such tensions,
pressures, and wants by accommodating them in a wider structure. This is a
self-realization conception of cultural interactions. In this type of view, it is not
enough to realize that in cultural interactions the recipient is active; one must
also realize that such activities arise from certain problem-solving needs. This
approach is shared by cognitive poetics, on the one hand, and cultural studies
as implemented by Itamar Even-Zohar (1978) and his followers, on the other.
The influence-hunting approach assumes that when the same conven-
tions occur in different places at different times, there must have been
an influence; the scholar’s task is to map how this influence proceeded in
space and time. The cognitive-constraints approach allows for such a pos-
sibility but also that the same convention can be reinvented several times,
independently, in several places, for similar reasons.4 As early as 1890, Sir
James Frazer observed in his monumental comparative studies The Golden
Bough (1922) and Folk-Lore in the Old Testament (1919) that “many of these
Even humans could have been “invented” more than once over the course of evolution:
A controversial question has been whether H[omo] erectus evolved into H. sapi-
ens, or a separate line that became H. sapiens displaced H. erectus worldwide as
it did the Neanderthals in Europe. Two main hypotheses have been proposed.
One hypothesis argues for a multiregional evolution of humans whereby H. erec-
tus, H. neanderthalensis, and other populations semi-independently became
the modern races of H. sapiens, with some gene flow between them. The other
hypothesis contends that all modern humans derive from a small group of com-
mon ancestors that lived in Africa. These individuals evolved into the modern
races of humans during their emigrations from Africa and displaced all other
hominid lines. (Bownds 1999: 99)
In either case, the various kinds of hominids sprang from different branches of the
evolutionary tree.
can also provide some indications on the “insights and intuitions” of poets
dead for hundreds of years. As to “how those attributions acquired the
status of widespread conventions,” let us consider the cognitive principles
that govern the placement of caesurae.
Such verse lines as the iambic tetrameter, pentameter, and hexameter
are divided by a caesura into two roughly equal segments. Classical prosody
requires a caesura at midline; in my view, this convention reflects certain per-
ceptual needs (see Tsur 1977a: 66–82, 1992b: 134–139, 1998: 113–139; see
chapter 2). Iambic lines, whose number of positions is divisible by four, can
be divided into segments of equal length and equal structure. The tetrameter
and the hexameter can be divided into 4 + 4 and 6 + 6, respectively, with each
segment beginning with a weak position and ending with a strong position.5
In these meters there is a very active caesura at the exact middle, even if it
is not marked by punctuation and even if it is overridden by the absence of
a word boundary. The iambic pentameter, by contrast, can be divided into
5 + 5 positions, but in this case the first segment begins and ends with a weak
position, whereas the second segment begins and ends with a strong posi-
tion. Alternatively, there can be two segments of unequal length, containing
4 + 6 or 6 + 4 positions, where each segment begins with a weak position and
ends with a strong position. So a caesura is more flexible in the pentameter
than in the tetrameter and hexameter, where it is rigidly fixed. As a result,
in the tetrameter, short rigid units follow fast upon each other, giving rise to
certain perceptual qualities especially when coupled with the trochaic meter.
Aristotle, for instance, perceived a quality in them that is “too much akin to
the comic dance” as opposed to the “characteristic rhythm of people as they
talk” (1932). I also claim that in the syllabotonic system in English, Hebrew,
and Hungarian poetry, 4 + 6 lines are overwhelmingly more frequent than
6 + 4 lines.6 The same asymmetry is found in French syllabic verse. I con-
clude, therefore, that the 4 + 6 line contains the unmarked caesura and the
6 + 4 line contains the marked caesura (see Tsur 1977a, 1992b; Tsur 1998 also
provides the cognitive reasons for this conclusion).
A classic instance of the “culture-begets-culture” approach to cultural
processing is to be found in the oft-heard claim, for example, with respect to
5. I use the terminology of Halle and Keyser (1966, 1971), who define meter as an
abstract pattern of regularly alternating weak and strong positions actualized in vari-
ous systems by strings of alternating stressed and unstressed syllables, or long and
short syllables.
6. In Hungarian iambic pentameter lines, the majority of the caesurae occur after
position V; but regarding the 4 + 6 or 6 + 4 division, the former division is the
dominant one.
poetic meter, that “the hemistich form 4 + 6 of the English iambic pentam-
eter is derived from the pattern of the French decasyllable” (Anonymous
reviewer, personal communication). But a far stronger argument can be
found for the “cognitive-constraints” approach in the seemingly unrelated
findings of an overwhelming tendency to place the caesura after the fourth
rather than the sixth position of decasyllabic lines in languages as remote
from each other as English, French, Hungarian, and Hebrew and in such
different versification systems as syllabic and syllabotonic meters. This set
of sufficiently unrelated findings clearly calls for a cognitive rather than an
influence-based explanation. It is much easier to show that this was the
case in French poetry and in sixteenth-century English poetry than that
the latter was derived from the former.
A number of important questions follow from these research findings.
Why does the same pattern recur in Hebrew and Hungarian poetry? Why
did the French poets, in the first place, prefer the 4 + 6 division so emphati-
cally rather than a random distribution? And why did this French practice
become such a widespread convention, in widely different versification sys-
tems? In English, French, Hungarian, and Hebrew I have not found a single
poet who uses a majority of verse lines with a segmentation of 6 + 4. The
culture-begets-culture approach merely transfers these mysteries from one
place to another.
There is also an interesting phenomenon in these languages: namely,
that in all these poets (with the notable exception of Shelley) the propor-
tion of “marked” caesurae increases with the proportion of marked forms
in other prosodic respects. In what follows I will claim that the marked
nature and unmarked nature of caesurae are determined by cognitive fac-
tors and are independent variables. But they may be associated with a vari-
ety of other independent variables, marked or unmarked, on the prosodic,
semantic, syntactic, and thematic levels; interact with them; and contrib-
ute to the generation of a wide variety of “regional qualities”7 that may
have great aesthetic significance. The culture-begets-culture approach, by
contrast, maintains that, as an anonymous reviewer from the Russian tra-
dition of verse studies commented on an earlier paper of mine,
the hemistich form 4 + 6 of the English iambic pentameter is derived from the
pattern of the French decasyllable; it is characteristic of periods where there
7. Regional quality is a Gestaltist term taken from Beardsley’s Aesthetics. It refers
to a perceptual property of a whole that is not a property of its parts (Beardsley
1958: 83–88).
was a more rigorous iambic canon in English literature. In the later works of
Shakespeare, in the verse of Romantic and Post-romantic poets the “hemis-
tich boundary” moves to the right, positions 6/7. What makes the 4 + 6 form
unmarked? The fact that it appears in a more rigorous variant of the English
iambic pentameter?
This and my approach represent two opposite lines of thought. In one line
of thought, the caesura derives its cognitive and aesthetic nature from the
other neighboring elements; in the other, certain cognitive factors deter-
mine the structural and aesthetic potential of the caesura, which, in turn,
contributes to the overall effect. Thus, in certain poetic styles the caesura
tends to move to the right (the later Shakespeare, Milton), back to the left
(Pope), and then back to the right (some Romantics). This is not a random
process but, rather, is governed by certain pervasive cognitive and aes-
thetic principles.
The exception in the context of the relative frequency of caesurae is
Milton, where there is no significant difference in the number of caesurae
after positions IV, V, and VI. This forces the conclusion that the unmarked
form is the caesura after position IV. Had a similar distribution occurred
in Pope, we would have to conclude that caesurae after positions IV and
VI are equally unmarked. These different conclusions relate to a distinc-
tion I make elsewhere (e.g., Tsur 2008a: 100–104) between convergent
and divergent poetry. In Milton’s divergent poetry, contrasts are typi-
cally blurred (for instance, by stressed syllables in weak positions and
unstressed ones in strong positions and by divergent patterns of allitera-
tion). Statistically, the difference in deviant stresses between Milton and
Pope is negligible: twenty-four deviations per one hundred lines, that is,
per one thousand positions, which hardly suffices to account for the intu-
ition that Milton is the prototypical deviant poet and Pope is the prototypi-
cal regular poet. However, the divergent effect of Milton’s poetry cannot
be accounted for merely by the number of deviating stresses but, rather, by
his recourse to the marked options in a wide range of versification devices,
of which stressed syllables in weak positions is only one and a marked
caesura appears to be another. He has recourse not infrequently to such
marked forms as strings of stresses that end in weak positions, disyllab-
ics with their stressed syllables in weak positions, and last but not least,
stress maxima in weak positions.8 In Paradise Lost, book I, fifteen out of
twenty compounds have their first (strongest) stress in a weak position.
8. A stress maximum is a syllable that bears lexical stress between two syllables that
do not. “A gárden” contains a stress maximum; “a stóne gárden” does not.
All these marked forms are virtually nonexistent in Pope. So we should not
be surprised that with respect to the caesura as well, Milton resorts to the
marked form more frequently than most other poets.
As to the relative frequency of unmarked and marked caesurae, I have
concluded that this is an instance of a wider linguistic principle, which
Jakobson subsumes under the poetic function:
“Why do you always say Joan and Marjory, yet never Marjory and Joan? Do
you prefer Joan to her twin sister?” “Not at all, it just sounds smoother.” In a
sequence of two coordinate names, as far as no rank problems interfere, the
precedence of the shorter name suits the speaker, unaccountably for him, as a
well-ordered shape of the message. (1960: 356–357)
This principle seems to hold for the “precedence of the shorter hemistich”
as well. This, in turn, intimates an even wider linguistic principle, which
has to do with the limitations of short-term memory that constrain many
speech-processing and rhythm-processing activities. Paraphrasing Cooper
and Ross (1975: 92), members that are easier to process (in this case, which
are shorter) tend to occupy the first place(s), enabling the listener to han-
dle the preliminary processing of this member while new information is
still being presented by the speaker. Geoffrey Leech formulated this as fol-
lows: “There is a general tendency for the weight of syntactic structure to
occur later rather than earlier in the sentence, so as to avoid strain on a per-
son’s short-term memory in the course of constructing and interpreting
sentences” (1974: 197). This principle can also be formulated as “the item
hardest to process comes last” and can account for a wide range of phonetic
preferences, including for such expressions as sing-song and ding-dong to
song-sing and dong-ding, of trick or treat to treat or trick, of walkie-talkie to
talkie-walkie, and so on (Cooper and Ross 1975; Tsur 1992a: 84–86).
In the traditional view, we are used to these expressions, and this is why
their reversal sounds odd. Here, I claim that there are certain general prin-
ciples (all derived from the limitations of short-term memory) that govern
these preferences in a variety of languages, in unforeseen combinations
as well. Consider, for instance, the title of Joseph Haim Brenner’s Hebrew
novel Shkhol vǝkhishalon (Bereavement and Failure). Shkhol is shorter
than kishalon, and, accordingly, it precedes it. Hillel Halkin, in his English
translation, reversed the nouns: Failure and Bereavement, establishing the
well-ordered shape. The following anecdote illustrates how intuition works
in such cases. A few years ago I met an Israeli poet in the corridor of a pub-
lishing house. “What are you doing here?” he asked. “Publishing a vol-
ume of poetry translations.” “What’s the title?” “The Cuckoo and the Flute.”
“It doesn’t sound good. I don’t know why, but The Flute and the Cuckoo would
sound better. It’s your job to find the explanation.” The director general of
the publishing house, himself a professor of literature, called out from his
office: “Joan and Marjory!”
Far from being used to this order of the two nouns, “the precedence
of the shorter name suits the speaker, unaccountably for him, as a
well-ordered shape of the message.” These two principles are certainly not
mutually exclusive. Rather, the existence of “conventional” pairs of nouns
is an additional illustration of the central claim of this book that rigid for-
mulae reflect the freezing of cognitive processes into “style.”
Cooper and Meyer (1960: 61) found that this principle prevails out-
side verbal communication, in music. Likewise, according to Woodrow’s
experiments, which date to the 1920s (Chatman 1965: 26–27; Meyer
1956: 106–107; Tsur 1977b: 88–89; and chapter 9 below), nonlinguistic
sound stimuli are perceptually grouped into end-accented groups when
there are differences of duration and into beginning-accented groups when
there are differences of amplitude. Differences of pitch do not affect the
direction of grouping.9 Among a variety of explanations, Cooper and Meyer
(1960: 61) referred to this finding to account for their own discovery that
musical sequences tend to divide into a shorter followed by a longer musi-
cal phrase.
Prosodists have pointed out for years that the longest verse line that has
no obligatory caesura is ten syllables long. The received view until recently
was that it is difficult to pronounce a longer verse line in one breath. The
truth is that even twenty syllables can be pronounced in one breath. In
fact, there are two different perceptual reasons that account for a caesura at
midline. One is the limitation on short-term memory, which, according to
George Miller (1956), is the “magical number” of 7 ± 2 monosyllabic words.
The other one is the Gestalt rule that the simpler the parts, the more they
tend to stand out as independent entities, such as, for instance, in sym-
metric wholes that have two equal parts. Iambic lines of eight and twelve
syllables can be divided into two segments of equal length and equal struc-
ture. Therefore the tetrameter, though well within the 7 ± 2 limit, forcefully
demands a break after the fourth syllable, whereas the hexameter both
exceeds the limit and divides into two segments of equal length and equal
structure.
To sum up this part of the controversy, the “culture-begets-culture” view
cannot explain how cultural models arise or why these models, once cre-
ated, become models to be imitated. It can merely refer us to ever-earlier
occurrences of the model. The cognitive approach, by contrast, can account
for the need for the caesura in the first place and the need to prefer the 4 +
6 segmentation to the 6 + 4 segmentation in decasyllabic lines, irrespective
of language and versification system. Furthermore, it can be shown that
this cognitive dynamic applies to additional linguistic phenomena, as well
as to nonverbal communication in music and mechanical beats. Even if it
is assumed that “the hemistich form 4 + 6 of the English iambic pentam-
eter is derived from the pattern of the French decasyllable,” we can now,
at least, give a reasonable answer to the questions of why a French poet
would want to segment a verse line into 4 + 6 and why an English poet in
the syllabotonic versification system would ever want to adopt “the hemis-
tich form 4 + 6” from the French syllabic system, rather than be satisfied
with a random distribution of caesurae. The compelling reason is that the
English poet intuitively feels that this is the well-ordered shape of the line,
irrespective of what French poets do.
deviations from this style; to use George Klein’s term (see 1970: 134–136),
we tend to “level out” the differences between unique works. In such cases,
an explication of a poem may be of decisive importance.
The characteristics of ballad style, for instance, are typically defined as
repetition (often merely decorative) and ellipsis (often quite easily com-
pleted). When we ask, “What does this repetition or that ellipsis contribute
to the ballad?” we are often handed the diagnosis that “this is a character-
istic of ballad style” or that “it contributes to the balladic atmosphere.”
Whenever there is a threefold repetition, one can expect to be told that “it
contributes to the popular-ballad style.” I by no means deny that repeti-
tion does contribute to the balladic character of the ballad. I would even
admit that these characteristics add a certain charm, a naïveté, to the bal-
lad. Nevertheless, this naïveté is sometimes the consequence of repressing
some utterly nonnaïve and expressive feature of the poem.
This principle can be applied to poetry and, among other poems, to the
popular ballad “Edward.” Let us take a fleeting look at its first stanza:
10. “ ‘Why does your sword so drip with blood, / Edward, Edward? / Why does your
sword so drip with blood? / And why so sad are ye, O?’ / ‘O, I have killed my hawk so
good, / Mother, mother: / O I have killed my hawk so good: / And I had no more but he,
O’ ” (http://www.poemhunter.com/poem/edward-edward-a-scottish-ballad/).
The main objection against the Victorian interior is that it is “vulgar,” by which is
meant that it allows a too easy, childish gratification in the gaudy and “cheap.” . . .
In design Victorian “atrocities” have acquired the distinction of being “amusing”—
which must be taken to mean that they have receded far enough into the back-
ground no longer to be dangerous. (Gombrich 1954: 20)
Figure 1.1 “Leveling” and “sharpening”: when a stimulus pattern shows a slight devia-
tion from some symmetrical pattern (as in a and d), subjects tend to restore symmetry (as
in b and e), or exaggerate the asymmetry (as in c and f), so as to avoid ambiguity. (After
Arnheim 1957.)
The very disgust we feel at the “cheap,” the “gaudy,” the “sloppy,” proves our
strong emotional involvement. Nor is the nature of this involvement hard to
guess. We react as if we resisted seduction. (Gombrich 1954: 20)
Ehrenzweig would certainly agree with the second paragraph; but he puts
forward an argument that is rather different from what is said in the first
paragraph:
Until recently, Victorian fashions were held almost in disgust. To us, the
Victorian drawing-room appeared to be crowded with useless furniture and
bizarre ornament. But in the recent sentimental revival of a newly discovered
Victorian “style” the grandeur and purity of the great Victorian age seemed sud-
denly to be revealed. It was as though there emerged a noble outline previously
hidden in the chaos of preposterous detail. (1953: 77)
CHAPTER 2
1. Note that in one sense the two approaches differ with respect to what they take for
granted: the influence-hunting approach takes for granted the existence of a full-blown
convention as its starting point, whereas the cognitive-fossils theory takes for granted
the existence of responses acquired for adaptation purposes.
whose solution was “love.” I start with the Freudian notion of an ambiva-
lent attitude toward the beloved. I do not claim that all love is ambivalent
or that there are no love poems that are not ambivalent. But some forms of
love are certainly ambivalent. Or to take one of the greatest literary mas-
terpieces, according to Ernest Jones (1949), Freud’s colleague and biogra-
pher, Hamlet’s inexplicable behavior is related in complex ways to splitting
the father figure into a chivalrous loving father and a “smiling damned
villain” (King Claudius). A straightforward wicked stepmother seems to
be less threatening than an ambivalent emotion. Such a response is thus
“untaught and unlearned.”
At this stage, if Freud is right in his supposition, the psychological pro-
cess is fossilized into a thematic convention. In poetry, ambivalent attitudes
can be sharpened into paradoxical statements that are less threatening
than ambivalent emotions but still hard enough to take. A catalog of para-
doxes may mitigate their disquieting effect. Here, the same psychological
process is fossilized into a structural convention. At this point, an additional
question arises: Why is it so obvious to Provençal riddle-makers that a cat-
alog of contradictions suggests “love”? Obviously, it is a convention. But
then we are back to the question of where conventions come from and, in
this case, to the answer that they arise from sharpening ambivalent atti-
tudes into contradictions.
According to the cognitive-constraints conception, we need not assume
an infinite recursion of influences. It is easy to grasp why the first poet
invented some verbal device, why the first audience could immediately
appreciate it, and how it could become a convention (through repeated
social transmission). In the next sections I briefly explore two crucial cog-
nitive constraints and their literary implications: (1) The cognitive system
has a limited channel capacity and therefore demands great parsimony;
(2) language is typically logical and conceptual, whereas some poetry at
least is supposed to convey emotional or mystical qualities that are non-
conceptual and illogical.
PARSIMONY
S o m e I m p l i c at i o n s of D ’A n dr a de ’ s A s s u m p t i o n s [ 21 ]
2
symmetrical halves, that is, with an identical length and an identical struc-
ture. They can be divided into segments of 4 + 4 and 6 + 6 syllables, respec-
tively. Since the iambic meter consists of regularly alternating weak and
strong positions, both segments in both meters begin with a weak position
and end with a strong position. Since the two segments of the line possess
similar features, they tend to stand out as relatively independent entities,
enforcing a break in the precise middle. Since the hexameter is longer than
ten syllables, it has an additional reason to enforce a caesura (the magical
number 7 ± 2).
But the “prevailing conditions” do not allow for such simplicity in all
iambic lines. As suggested above, their psychological organization will be
only as good as the prevailing conditions allow. An iambic pentameter line
can be divided into 5 + 5 syllables, but in that case the two segments will
have different structures: the first segment will begin and end with a weak
position, and the second, with a strong position. It is possible to obtain seg-
ments that begin with a weak position and end with a strong position if we
divide the line into segments of unequal length, 4 + 6 or 6 + 4. Obviously,
we cannot have two segments of equal length and equal structure, so they
stand out less clearly, and the whole line is more integrated. Furthermore,
there is no single point of caesura but an area of caesura after the fourth,
fifth, or sixth position.
S o m e I m p l i c at i o n s of D ’A n dr a de ’ s A s s u m p t i o n s [ 23 ]
42
Put this way, the principle applies to parallel simple phrases as well. As
Jakobson pointed out, “Joan and Marjory” sounds somehow better than
“Marjory and Joan.” Outside linguistic structures, according to Cooper
and Meyer (1960: 61), where parallel musical phrases occur, the longest
comes last. This is due to the constraints of the cognitive system. In all
these instances, the longest segment comes last so as to avoid jamming
short-term memory—which has a limited channel capacity—while infor-
mation is still coming in.
On the foot level as well there is a tendency to prefer feet in which the
longest constituent comes last. Tenth-century Hebrew poets in Spain
imported a complicated metrical system from contemporary Arab poets
based on short and long syllables and adjusted it to the constraints of
Hebrew. This was a deliberate, conscious act. In this system there is an
enormous number of possible meters. However, in a very short time, for
mysterious reasons, two of these emerged as the most frequent meters
(I claim that the reasons are not so mysterious, since they can be accounted
for by Gestalt theory). Of these two meters, one occurs several times more
frequently than the other. The longest-comes-last principle can easily
explain this pattern of occurrence as well. Both meters are based on feet
that consist of four units: a schwa mobile and three full vowels.2 The schwa
is considerably shorter than the full vowels. In the more frequent meter,
the schwa occurs in the first half of the foot, at the very beginning, so that
the longer part comes last. In the less frequent meter it occurs in the sec-
ond half, in the second to last position, so that the longer part comes first
(as would be in English “about painting” [əbaʊt peɪntɪŋ] and “this cutlery”
[ðɪs kʌtləri], respectively—the issue at stake being not the sequence of
stresses but the placement of the reduced vowel ə among the full vowels
and diphthongs). There is no trace of this preference in poets’ explicit poet-
ics; it is all intuitive (Tsur and Bentov 1996). There was a similar develop-
ment, mutatis mutandis, with the overwhelming dominance of the iambic
meter (where the longer syllable of the foot also comes last) in a number of
Western languages.
At this point I must acknowledge three difficulties and attempt to
account for them within this model. First, in Hungarian iambic pentameter
lines, when we compare caesurae after the fourth and sixth positions, the
former do, indeed, considerably outnumber the latter; however, the most
frequent placement of caesurae (unlike in English and Hebrew poetry)
is after the fifth position. This seems to be due to the constraints of the
2. Schwa mobile is a “reduced” vowel in Semitic languages, similar to the first vowel in
English amid (əmɪd).
S o m e I m p l i c at i o n s of D ’A n dr a de ’ s A s s u m p t i o n s [ 25 ]
62
stressed syllables are typically longer than unstressed ones. This is why the
most natural sequence of positions in the feet is “the strong position comes
last” in syllabotonic verse too, even in Hungarian, where stress comes typi-
cally on the first syllable of the word.
The theory I develop here requires that a wide range of plausible and
implausible forms should arise randomly that, in the course of repeated
social transmission, assume a good fit to the natural capacities and con-
straints of the human brain. In a study of Arabic metrics, Golston and Riad
(2010) found that end-accented meters make up about 80–90 percent of
Arabic verse (they call them “iambic meters”).4 If they mean their terms in
the same way as I do, this would suggest that in classical Arab meter a simi-
lar process took place to what we have found in medieval Hebrew poetry.
One of the most dramatic pieces of evidence for the cognitive-constraints
approach, where cognitive constraints override even the stress rules of a
language, is to be found in the continued existence, even dominance, of
iambic lines in Hungarian poetry. “Influence” could at best ensure that the
iambic would continue to exist in Hungarian but not that it would be felt to
be the most natural meter (I discuss this question in chapter 9). My argu-
ment is mainly based on two sets of experiments. One set is H. Woodrow’s
tick-tack experiments back in the 1920s, which found that in a series 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” (1951: 1233): “Intensity
has a group-beginning effect: duration, a group-ending effect: pitch, nei-
ther a group-ending nor a group-beginning effect” (1911: 77). In L. B.
Meyer’s paraphrase, “Durational differences tend to result in ‘end-accented
rhythms,’ and intensity differences tend to result in ‘beginning-accented
rhythms.’ . . . Pitch has neither group-beginning nor group-ending effect”
(1956: 106–107). A recent experiment (with more advanced technical facil-
ities) indicated that variations in pitch lead to a significant shift toward
iambic groupings (Rice 1992).
The other set is D. B. Fry’s experiments on stress perception, which dem-
onstrated that linguistic stress is signaled by a mixture of acoustic cues
including pitch, duration, and amplitude, in decreasing order of effective-
ness. If pitch differences are irrelevant to grouping direction and duration
differences are more effective in stress perception than amplitude differ-
ences, end-accented meters should be more natural in poetry in a variety of
languages. If variations in pitch also lead to a significant shift toward iambic
groupings, they should reinforce this effect. Significantly, even in Hungarian
poetry, where stress is invariably on the first syllable of the word, the iam-
bic meter is far more natural and widespread than the trochaic. The great
Hungarian poet Miklós Radnóti (1943: 167) observes in the epilogue to his
volume of poetry translations that the iambic is the most natural meter for
modern Hungarian poets.
In an interview for the Literature, Cognition and the Brain website concern-
ing the origins and implications of cognitive poetics, Beth Bradburn asked
me: “Cognitive Poetics emphasizes the cognitive processes of individual read-
ers and poets while also assuming the universality of those processes. Does
this tend to set a limit on the extent to which literary texts may be read as
cultural artifacts?” (Bradburn 1998). My answer included the following com-
ments. Cultural artifacts do not spring like Pallas Athena, fully armed, from
Zeus’s head. There is a process of development, until the convention reaches
its optimal fit to the natural capacities of the human brain. But, curiously
enough, these conventions do not become uniform, even when shaped and
constrained by the same cognitive mechanisms. D’Andrade accounted for this
process by repeated social transmission. Thus, when similar cultural programs
are found in most societies around the world, there is reason to search for psy-
chological factors that can account for these similarities. On the other hand,
one might add, if these “cultural programs” contain features that do not con-
flict with the natural capacities and constraints of the human brain, or there
are some good culture-specific (e.g., religious)5 reasons to preserve them, huge
differences or even conflicting patterns may evolve between the various cul-
tures on more concrete levels.
5. An illuminating example of such religious reasons was put forward by John
Skoyles (1997). In ancient times there seem to have been good physical and perhaps
neurological and cognitive reasons for writing from right to left. When vowels were
inserted between the consonants in Greek writing, the cognitive reading process is
hypothesized to have changed as well. Skoyles quotes a wide range of cognitive experi-
ments that support this. Current models of reading suggest that adults use several
kinds of processes involving at least two different routes by which words can be identi-
fied and pronounced. For instance, there are separate routes for reading whole lexical
items (in, e.g., Semitic languages, where only the consonants are written) and reading
words assembled from their consonants and vowels. In Hebrew, for example, the con-
sonant sequence spr can be identified as a unit as sappar (barber), sepher (book), saphar
(counted), or sphar (border). By contrast, the words sappar, sepher, saphar, and sphar
written in Latin characters can be assembled from their parts. Thus, the insertion of
vowels into writing may have entailed the gradual change of writing direction first
into bidirectional (that is, one line from right to left, the next line from left to right)
S o m e I m p l i c at i o n s of D ’A n dr a de ’ s A s s u m p t i o n s [ 27 ]
82
and then into left-to-right writing, finally reaching, in terms of the present book, the
best fit of the changed cognitive route to the natural capacities and constraints of the
human brain. In Hebrew and Arabic these changes did not take place, because of the
prohibition against altering the sacred texts of the Jews and Muslims. These religious
works are considered by believers to be the word of God and thus unalterable. For a
more recent neurological presentation of the two routes of reading, see Dehaene 2009.
6. In medieval Hebrew poetry in Spain, the meter, imported and adjusted from
contemporary Arab poetry, was based on full vowels (termed “chords”) alternating
in all sorts of combinations with units consisting of a schwa mobile plus a full vowel
(termed “pegs”).
should be considered short, whereas all the consonants with vowels proper
should be considered long (Schirmann 1995: 120). In this way, a radically
new system was invented. This system would be inconceivable in most (and
perhaps all) European languages but is germane to classical Hebrew. Thus,
the adoption of foreign models is quite conspicuously constrained by the
properties of the adopting languages.
Syllabic meter would be well suited to any language I know of, and in
quite a few of them there have been relatively short periods in which the
syllabic system was prevalent. Downright dominance of the syllabic ver-
sification system for extended periods, however, is surprisingly rare in
European languages and is found mainly in the (syllable-timed) Romance
languages. We may account for the relative frequency of meters other than
syllabic with the aesthetic demand for unity and complexity. These two
demands constrain each other and seem to exert some pressure on versifi-
cation systems to display greater complexity and unity than mere syllabic
meter. In French, where attempts to accommodate other versification sys-
tems failed, complexity was imposed by means compatible with the syl-
labic system: in the French alexandrine there is an obligatory caesura after
the sixth syllable and a systematic alternation of feminine and masculine
rhymes. This rhyming convention was also restricted by the constraints of
the French language. In English, German, Hebrew, Hungarian, and some
other languages, in some words the main stress is on the last syllable, and
in some words, on the penultimate or antepenultimate syllable. In French,
by contrast, there are no such contrasts of stress. The only way in which
French rhymes can be systematically varied is to contrast words that end
with an e muet (feminine rhyme) and those that do not (masculine rhyme).
This increases the complexity of the rhyme pattern; but at the same time
it groups together a masculine- and a feminine-ending line, thus dividing
a quatrain into two symmetrical halves. Rhymed words ending with an e
muet are not the same as rhymed words stressed on their second to last syl-
lable in some other European languages but are the nearest options avail-
able in French. They are, therefore, usually perceived as equivalent.
When, in Dante’s time, the pegs-and-chords meter was imported from
Spain to Italy, the Hebrew poets were exposed to Italian syllabic meter,
and their pegs-and-chords meter began to resemble the Italian syllabic
meter. Characteristically, in the Hebrew sonnets by Emanuel of Rome some
present-day scholars designate the meter as hendecasyllabic, whereas oth-
ers define it as belonging to the pegs-and-chords system. What appears to
be quite surprising is that this newborn syllabic meter gradually assumed
the characteristics of the syllabotonic iambic. It was the great Hebrew poet
and scholar the late Dan Pagis (1993) who described this development in
S o m e I m p l i c at i o n s of D ’A n dr a de ’ s A s s u m p t i o n s [ 29 ]
03
SCHEMATA
When the information to be processed exceeds the limit of its channel capac-
ity, the organism can have recourse to a variety of cognitive strategies and
devices. One possible way of handling excess information is to recode it in
a more efficient manner, so as to require less processing space.7 Suppose
you are asked to memorize the following series of numbers and then recall
them one week later: 5 8 1 2 1 5 1 9 2 2 2 6. In George Katona’s experiment
(reported by Roger Brown [1958]), none of the subjects could recall them.
Now, suppose you are to learn the same numbers along with their principle
7. In my work on the rhythmical performance of poetry (Tsur 2012a) I show that
in case of metrical deviation, the perceptual solution consists of recoding the sound
material into better gestalts, so as to keep it within the capacity of immediate memory
(see chapter 9).
5 8 12 15 19 22 26
3 4 3 4 3 4
Figure 2.1 A series of apparently random numbers is difficult to memorize. The same num-
bers are better memorized when their principle of organization is given prominence, allevi-
ating cognitive overload.
S o m e I m p l i c at i o n s of D ’A n dr a de ’ s A s s u m p t i o n s [ 31 ]
23
leaving. The brain’s need for parsimony requires us to develop schemata for
remembering large quantities of information; once developed, they allow us
to infer suppressed information, which, in turn, allows further parsimony
in communication. In this sense, schemata are constrained by the natu-
ral capacities of the brain. In different cultures, different schemata can be
developed by the same constraints. The same constraints of the brain may
further shape the different schemata in different cultures.
We have an enormously rich lexicon stored in our long-term memory, in
a way that we can use it with amazing flexibility to communicate about all
possible topics in the world. Linguists and psychologists disagree as to how
these meanings are coded in our memory, but all agree that we have access
to them with amazing ease (for disruptions of this process, see chapter 10).
Suppose we adopt one of the semantic-primitives models, or one of the
semantic-networks models, or a combination thereof, to account for our
versatility in using words. Would it be sufficient to account for our ability
to communicate with words?
Today it is absolutely clear that a model of semantic representation would
be insufficient and that we also need a model for the representation of world
knowledge. In his early work on artificial intelligence, Roger Schank (1973)
included all the information required in the definition of words in a system
called “conceptual dependency.” He gradually discovered that for effective
language processing, he also needed such wider notions as scripts, plans, and
goals. One of the prevalent models is what Bartlett (1932) and Norman and
Rumelhart (1975) called “schemata” and Schank and Abelson (1977) called
“scripts.” I will give two brief examples to indicate how scripts and schemata
help us establish implied relationships between events and make infer-
ences as to what has not been explicitly stated in an utterance. Rumelhart
(1979: 85–86) illustrated the limitations of the semantic approach of com-
ponential analysis with the following sentence frame: “X raised his hand and
stopped the car.” If we substitute the phrase “the policeman” for X, the most
likely account for this event according to Rumelhart is
one involving a traffic cop who is signaling to a driver to stop his car. Note that
this brings under consideration a number of concepts that are not mentioned
in the sentence itself. For example, this interpretation requires that the car has
a driver and that the policeman stopped the car by signaling with his hand to
the driver, who then most likely puts his foot on the brake of the car causing it
finally to halt. (1979: 85–86)
One night two young men from Egulac went down to the river to hunt seals,
and while they were there it became foggy and calm. Then they heard war cries
and they thought; “Maybe this is a war-party.” They escaped to the shore, and
hid behind a log. Now canoes came up, and they heard the noise of paddles and
saw one canoe coming up to them. There were five men in the canoe and they
said; “What do you think? We wish to take you along. We are going up the river
to make war on the people.”
One of the young men said; “I have no arrows.”
“Arrows are in the canoe,” they said.
“I will not go along. I might be killed. My relatives do not know where I have
gone. But you,” he said, turning to the other, “may go with them.”
So one of the young men went, but the other returned home. And the war-
riors went on up the river to a town on the other side of Kalama. The people
came down to the water and began to fight, and many were killed. But presently,
one of the young men heard one of the warriors say; “Quick let us go home. That
Indian has been hit.”
Now he thought; “Oh, they are ghosts.” He did not feel sick, but he had been
shot. So the canoes went back to Egulac, and the young man went back to his
S o m e I m p l i c at i o n s of D ’A n dr a de ’ s A s s u m p t i o n s [ 33 ]
43
house and made a fire. And he told everybody and said; “Behold, I accompanied
the ghosts, and we went to fight. Many of our fellows were killed and many of
those that attacked us were killed. They said I was hit, but I did not feel sick.” He
told it all, and then he became quiet. When the sun rose, he fell down. Something
black came out of his mouth. His face became contorted. The people jumped up
and cried. He was dead. (Bartlett 1920)
Which schema enables the reader to infer “Oh, they are ghosts” from the
observation “He did not feel sick, but he had been shot”? Similarly at the
end of the story, we have no schemata to account for “They said I was hit,
but I did not feel sick” and “Something black came out of his mouth.” But
for the original audience of the story this is precisely to be expected when
you are killed in a war against ghosts.
supplied connecting links, whereas in “The War of the Ghosts” events follow
one another, but their connection is not explicitly stated (cf. 1932: 86):
Or, consider again, at the end of the story, the phrase “Something black
came out of his mouth.” This complies with the original audience’s schema
of being killed in a war of ghosts but is utterly meaningless to a
Western audience. The following sequence of renderings of this phrase is
telling:
“he gave a cry, and as he opened his mouth a black thing rushed from it”
“a black thing rushed out of his mouth”
“a great black thing flew out of his mouth”
“his soul fled black from his mouth”
“his soul passed out from his mouth”
“his spirit fled”
“his spirit left the world”
S o m e I m p l i c at i o n s of D ’A n dr a de ’ s A s s u m p t i o n s [ 35 ]
63
One of the underlying motifs of this book is the natural capacities and con-
straints of the human brain. But in the previous examples I have made no
explicit reference to the brain. It would have been an unnecessary duplica-
tion of terms. I referred to the “cognitive system,” “limited channel capac-
ity,” “immediate memory,” and “schemata,” all of which have been shaped
and constrained by the human brain. In this section I discuss an issue
where an explanation cannot be made without explicit reference to the
brain. This issue concerns the handling of emotive discourse with concep-
tual language:
S o m e I m p l i c at i o n s of D ’A n dr a de ’ s A s s u m p t i o n s [ 37 ]
83
Brain and Language, Jakobson (1980) pointed out that language and speech
are typically related to the left hemisphere of the brain. However, there
are verbal elements such as interjections and exclamations that permeate
speech yet act like immediate signals. These “stand outside the general syn-
tactic patterning of language, and they are neither words nor sentences. . . .
[S]
emantically they are reduced to stereotyped affective expressions”
(Jakobson 1980: 23) and are typically related to the right hemisphere.9
Jakobson quotes an 1874 article by John Hughlings Jackson: “The com-
munist orator did not really make a blunder when he began his oration
‘Thank God I’m an atheist,’ for the expression Thank God is used by careless,
vulgar people simply as an interjection” (Jakobson 1980: 23). Hamlet obvi-
ously uses “God” here as an interjection (not as “the Everlasting,” which
he uses as a full content word). Proper nouns have no meaning either but
are immediate signals used to designate a particular being or thing; that is,
semantically they are reduced to tags (when you say “Herbert” you don’t
have in mind “army + bright” [the name’s Germanic roots] but use it to tag
a person). Vocatives can also act as “immediate signals” and “stand out-
side the general syntactic patterning of language” and are typically related
to the right hemisphere. What is more, repetitions and analogies typically
involve the right hemisphere. The left hemisphere of the brain, the linguis-
tic hemisphere, is generally logical. Emotions and emotional processes are
typically related to the right hemisphere. Given the logical and conceptual
nature of language, it seems to be ill-suited to convey emotional experi-
ences in poetry. One way that poets attempt to overcome this limitation
of their medium is to increase the activation of the right hemisphere by
some verbal device.10 “Laertes” in the king’s speech is not merely a name
but a repeated vocative, a verbal device that activates the right hemisphere,
which in turn bestows affective power on it. Interjections, vocatives, and
repetitions (which are typically related to the right hemisphere) are some of
the most elementary devices. By now they have lost much of their effective-
ness because they are so hackneyed—precisely because they have become a
convention. There are other more sophisticated and poetically more effec-
tive devices as well related to space perception.11
We may better understand the nature of the fossilization process if
we realize that we are not facing an all-or-nothing distinction but, rather,
9. Lateralization (= being largely under the control of one side of the brain) has
become an important topic in brain research. A highly readable introduction to lateral-
ization can be found in Ornstein 1975. A recent summary of research on lateralization
with reference to poetic language can be found in Kane 2007.
10. Another device involves divergent structures (see, e.g., Tsur 1972, 2008a).
11. See, e.g., Tsur 2008a: 285–403, 595–622; Tsur and Benari 2002.
INTERIM SUMMARY
I began this chapter with the suggestion that “in the process of repeated
social transmission, cultural programs come to take forms which have a
good fit to the natural capacities and constraints of the human brain.” In
S o m e I m p l i c at i o n s of D ’A n dr a de ’ s A s s u m p t i o n s [ 39 ]
04
the preceding sections I briefly explored each constituent part of this sug-
gestion. Language is logical, conceptual, and typically related to the left
hemisphere of the brain. Poetry is sometimes assumed to convey illogical,
nonconceptual, highly emotional information. Thus, language is poorly fit-
ted to convey emotional qualities, and poets must find verbal devices that
can increase the activity of the right, emotional hemisphere of the brain.
Interjections, exclamations, and vocatives are typically related to the right
hemisphere, and so are repetitions of all kinds. The repetition of Laertes’s
name by King Claudius in Hamlet may be required by his character, his
“mellifluous excess of speech,” an individual choice to achieve an affection-
ate effect; whereas in Elizabethan drama interjections followed by repeated
vocatives became a conventional means to express strong passion. Hamlet’s
“O God! God!” would be, according to John Hughlings Jackson, an interjec-
tion rather than a vocative. Thus, the conventional device has a good fit
to the emotional capacities of the brain but becomes less effective when
hackneyed: a fossilized cognitive process.
The capacities and constraints of the brain may govern a process of
“natural selection.” Medieval Hebrew poets in Spain imported a wide
range of poetic meters from contemporary Andalusian Arab poetry. They
had no explicit preference for any of them. But in the process of repeated
social transmission one meter gained overwhelming predominance, owing
to its good fit to the Gestalt rules of perception, on the one hand, and to
the limited channel capacity of the cognitive system, on the other (the
longest-comes-last principle).
Bartlett’s ingenious experiment illustrated the process of repeated social
transmission through which the Eskimo story “The War of the Ghosts” was
detached from the original schemata underlying it and then gradually came
to take a form that has a good fit to the schemata entertained in Western
culture. When it achieved that good fit, the story stopped changing in sub-
sequent transmissions. Though “a good fit to schemata” is not quite the
same as “a good fit to the natural capacities of the human brain,” Bartlett’s
experiment reproduced the mental dynamics involved. As he himself said,
“The reproductions themselves illustrate the operation of principles which
undoubtedly help to determine the direction and character of convention-
alisation as it occurs in everyday experience” (1920: 31).
12. Initially I was working with D’Andrade’s 1980 lecture. Later I discovered that he
had also published it (D’Andrade 1981). In the abstract he uses the phrase “a good fit to
the natural capacities and constraints of the human information processing system,”
suggesting that he uses it interchangeably with “a good fit to the natural capacities and
S o m e I m p l i c at i o n s of D ’A n dr a de ’ s A s s u m p t i o n s [ 41 ]
24
The brain of any educated adult contains a circuit specialized for reading. But how
is this possible, given that reading is an extremely recent and highly variable cul-
tural activity? The alphabet is only about 4,000 years old, and until recently, only a
very small fraction of humanity could read. Thus, there was no time for Darwinian
evolution to shape our genome and adapt our brain networks to the particularities
of reading. How is it, then, that we all possess a specialized letterbox area? (2013)
constraints of the human brain.” Fortunately, I discovered this only after I was forced
to find Dehaene’s work to support my thesis in its original formulation.
Neurons in the ventral visual pathway often respond to simple shapes, such as
those formed by the intersections of contours of objects. Even in the macaque
monkey, the inferotemporal visual cortex already contains neurons sensitive to
letter-like combinations of lines such as T, L, X, and *. The ventral visual system
seems to favor those shapes because they signal salient properties of objects that
tend to be robustly invariant to a change in viewpoint. (2013)
S o m e I m p l i c at i o n s of D ’A n dr a de ’ s A s s u m p t i o n s [ 43 ]
4
This text means “No limit may be set to art, neither is there any craftsman
that is fully master of his craft” (The Instruction of Ptahhotep) (Figure 2.2).
Later scribes, by contrast, sought to achieve a good fit to the shapes preex-
isting in the brain’s letterbox, at the expense of similarity to the physical
world. Let me illustrate the process with one highly illuminating example.
Semitic alphabets adapted Egyptian hieroglyphs to write consonantal val-
ues based on the first sound of the Semitic name for the object depicted
by the hieroglyph. ʔalp or ʔālep in Phoenician and in ancient Hebrew meant
“ox.” In the Phoenician script, the first letter of the word ʔalp was derived
from the Egyptian glyph for “ox,” yielding the letter aleph.
Dehaene (2009: 185) begins this string of shapes one step earlier, with
an ox head from the famous Lascaux cave paintings (approximately eigh-
teen thousand years ago; Figure 2.3). At the beginning of the process, the
Figure 2.2 How the alphabet was born from hieroglyphs. (From Goldwasser 2010;
Millmore, http://discoveringegypt.com/egyptian-hieroglyphic-writing/.)
smoothly rounded Egyptian glyph, though a simplified contour, did not con-
form to the abstract shapes in the brain’s letterbox but, rather, to the shape of
the ox head out in the world. The same holds for the first alphabetic writing,
Proto-Sinaitic. But Phoenician scribes “progressively selected their letters and
written characters to closely match the set of shapes that were already pres-
ent in the brains of all primates”—to use Dehaene’s words. Subsequently, in
Greek the aleph was rotated upward, to become A. Significantly for our pur-
pose, Goldwasser observes: “The iconic meaning of the hieroglyphs was so
important that even today, when the Hebrew letters have lost all iconic con-
nection to the old pictorial models (we can’t recognize what the letters are sup-
posed to picture), most letters are still named after the old pictures!” (2010).
In terms of the present argument, this means that here, too, a typical fos-
silization process took place. At the beginning of the process the characters
had a good fit to physical reality, facilitating immediate understanding of the
S o m e I m p l i c at i o n s of D ’A n dr a de ’ s A s s u m p t i o n s [ 45 ]
64
iconic meaning of the characters, but in the course of repeated social trans-
mission they assumed a good fit to the constraints of the brain. To be sure, the
set of iconic characters too had, in a sense, a good fit to the constraints of the
human brain. As Goldwasser suggests, “People learned the letters from one
another orally. For this kind of use, the pictorial nature of the signs was very
important. It was easy to learn the alphabet simply by memorizing the pic-
tures” (2010). But for some reason, in order to gain such universal currency,
writing systems had to assume good fit to “the set of shapes that were already
present in the brains of all primates” (Dehaene 2013)—even at the price of
losing “one of its greater assets: its mnemonic power” (Goldwasser 2010).
The Phoenician script differs, then, in two crucial respects from the hiero-
glyphs: It is alphabetic, not pictographic; and its symbols have a good fit to
the primitive shapes in the primate brain, not to the physical objects out in
the world. The Proto-Sinaitic script performed only one of these two revolu-
tions: Its symbols designate single consonants, not objects in physical real-
ity; but they still retained the shapes of the physical objects. However, once
the icons were used to designate the first sound of the Semitic name for the
object depicted by the hieroglyph and not the object itself, that cleared the
way to progressively match the primitive shapes to which some neurons in
the primate brain were sensitive, detaching themselves from the hieroglyphic
pictograms. It took about half a millennium to accomplish this process.
I have recently coauthored a paper on a cognitive approach to Dao de Jing
(Wagshal-Te’eni and Tsur forthcoming). So I venture to use my experience
gained there to illustrate characters we find in Chinese: 道德經 第十四章
贊玄 (meaning “Dao de Jing, chapter 14: Praising the Mysterious” [http://
www.yellowbridge.com/onlinelit/daodejing14.php]). These figures consist
of complex combinations of primitive intersecting lines and angles.
As Dehaene notes,
In the end, the neuronal recycling hypothesis leads me to believe that we are
able to learn to read because within our preexisting circuits, there is one that
links the left ventral visual pathway to the left-hemispheric language areas. This
circuit is already capable of recognizing many letter-like shapes, and it possesses
enough plasticity, or adaptability, to reorient toward whichever shapes are used
in our alphabet. (2013)
systems of the world. But the capability of recycling such “older cerebral
structures that originally were selected by evolution to address very differ-
ent problems” indicates the plasticity of the brain.
Now we have to face the question of how sensitivity to the same combina-
tions of lines in primate brains can shape and constrain such a wide variety
of writing systems around the world (with the exception of Egyptian hiero-
glyphs). I suggest that one may account for these differences in the world’s
writing systems by invoking Polányi’s (1967) “Principle of Marginal Control”:
namely, that each higher process makes use of the margins left indetermi-
nate by the lower-level processes. In the present instance, existing brain
structures may determine the graphic primitives of alphabets but leave inde-
terminate their specific combinations into characters, as well as the linguistic
units to which the characters may be attached. In the foregoing Chinese text,
for instance, they combine into very complex characters. And in various
writing systems the characters may get attached to phonemes in the Latin
alphabet, to consonants only in the Hebrew and Arab alphabets, to syllables
in Japanese kana notation, or to entire words or morphemes as in Chinese.
In Bartlett’s experiment repeated social transmission is discussed in
great detail, but there is no evidence regarding the constraints of the brain,
only the cultural cognitive schemata involved. In Dehaene’s case, on the
contrary, repeated social transmission through generations of scribes is
only hypothesized; but he provides specific information about the natu-
ral capacities and constraints of the primate brain, on the one hand, and
cultural artifacts (writing systems), on the other. To fill in this gap, in a
rather coarse-grained account of the standard historical facts, I have tried
to trace the gradual development of writing systems, from hieroglyphs that
had a good fit to the physical objects out there in the world, through the
Proto-Sinaitic alphabet, through Semitic alphabets, to the Greek and Latin
alphabets that have a good fit to the abstract shapes to which certain neu-
rons in primate brains have special sensitivity. In all this I relied on more or
less traditional, noncognitive scholarship.
Thus far we have considered the brain’s constraints that governed the
process. In the major part of his article, Dehaene explores the plasticity of
the human brain that enabled the exploitation of existing brain structures
for new cultural ends. He puts the predictions of the neuronal recycling
hypothesis to the test. One of the predictions has to do with the notion of a
cortical competition process. The model predicts that as cortical territories
dedicated to evolutionarily older functions are invaded by novel cultural
objects, their previous organization will slightly shift away from the origi-
nal function (though the original function is never entirely suppressed).
As a result, reading acquisition should displace and dislodge whichever
S o m e I m p l i c at i o n s of D ’A n dr a de ’ s A s s u m p t i o n s [ 47 ]
84
When scanning 9-year-olds who were good readers versus poor readers, we found
two interesting differences in the ventral visual pathway. . . . The poor readers
not only showed weaker responses to written words in the left-hemispheric
visual word form area, but also showed weaker responses to faces in the
right-hemispheric fusiform face area. Thus, the acquisition of reading seemed
to induce an important reorganization of the ventral visual pathway, which dis-
places the cortical responses to face away from the left hemisphere and more
toward the right. (2013)
various brain areas. This is, of course, evidence for the brain’s plasticity.
He further indicates the constraints that required such alterations: the
“cortical competition effect, whereby word responses increased while face
responses decreased” in the visual word form area.
Thus we do not face an either/or choice: the constraints conception and
the plasticity conception of the human brain complement each other in
generating such cultural artifacts as writing systems and poetic conven-
tions. Dehaene’s article has filled two gaps disputed by my reviewer and me.
MORE TROUBLE
I have discussed the principle that in the process of repeated social trans-
mission, poetic conventions come to take forms that have a good fit to the
natural capacities and constraints of the human brain. This does not mean,
however, that there are no exceptions or even poetic conventions that pre-
scribe precisely the opposite of what has a good fit to the natural capacities
and constraints of the human brain. I noted the perceptual principle that
in a series of units of equal importance, the most natural sequence is that
the longest comes last, on the phrase level, the verse-line level, and the foot
level, as well as in music. As we have seen, this is so even on the elementary
“tick-tack” level. A useful explanatory model must be able to account for
apparent counterexamples as well, such as, for instance, when convention
demands a series in which the longest item comes first. At the beginning
of this chapter I invoked an unintentional process of natural selection by
which the metrical pattern that has the best fit to the natural constraints of
cognitive economy is the one that has a better chance to prevail and multi-
ply. Accordingly, in medieval Hebrew poetry, a foot in which the longer part
comes last became the dominant one. In English, Hungarian, and Hebrew
syllabotonic verse the iambic foot became the dominant one, in which the
longer position comes last as well, and this is probably the case in classical
Arab prosody too. In the Serbian folk epic and in French syllabic poetry the
longer segment of decasyllabic lines comes last. This is the case in musical
phrasing too.
In this context, however, there is a disconcerting issue. The natural
selection discussed here works well in some poetic traditions, but other
traditions force rigid limitations upon poets and prevent them from pre-
ferring 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,
S o m e I m p l i c at i o n s of D ’A n dr a de ’ s A s s u m p t i o n s [ 49 ]
05
the iambic, the trochee, the anapest, and the dactyl. The trochaic and the
dactylic feet are relatively unnatural because the strong position (longer
syllable) precedes the weak one(s). The dactylic is less natural than the tro-
chaic because two weak positions lean back upon one strong position. The
appraisal that the trochee and the dactylic hexameter are the least natural
ones among the predominant meters is corroborated by Aristotle’s com-
ment in Rhetoric: “Of the various rhythms, the heroic . . . lacks the tones
of the spoken language,” and “the trochee is too much akin to wild danc-
ing” (1932). Why, then, should precisely that rhythm that has a poor fit to
the constraints of the cognitive system, that is, is the least natural of the
rhythms, prevail and multiply?
For years I had to content myself with the answer that rigid conven-
tion robbed poets of the freedom to follow their own intuitive preferences.
According to the present conception, however, this solution poses another
problem. We would expect rigid conventions to perpetuate the more natu-
ral rather than the less natural meters. I found the solution to this riddle
only recently, when I reread, for a different purpose, Aristotle’s paragraph:
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 common
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. (1932)
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,
heroic action does not demand the most natural rhythms but, rather, those
that take the hearer out of his or her ordinary self—that is, the least natu-
ral meters. This is how the most “dignified” rhythm is generated, so as to
support the intensive 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.” In D’Andrade’s terms, the
most “dignified” meter turns out to be the one that has a less good fit to the
natural constraints of the human brain—precisely because it goes against
them.13 Hence the fact that some, or many, or most conventions acquire a
13. Elsewhere I discuss the question of what “dignified” may mean with reference to
metrical structures (Tsur 2012a: 408–409). The view I put forward there may account
for the “dignified” quality of the spondaic substitution in hexameter verse; the present
argument accounts for the “dignified” quality of the dactyl.
good fit to the natural capacities and constraints of the human brain does
not exclude the existence of conventions whose fit to those capacities and
constraints is poorer. 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.”
Finally, another example shows how poor fit has its own expressive
value. I argued above that in the iambic pentameter line a caesura after
the fourth position is the “unmarked,” more natural caesura: the nearer an
intruding event to the end of the verse line (that is, the shorter the seg-
ment that comes last), the more it increases the need for the missing seg-
ment, and the greater the satisfaction when it “clicks in” (Tsur 1972). Thus,
the caesura after the fourth position may have a natural, relatively relaxed
character, whereas a caesura after the sixth position may elicit a mild sense
of unease or frustration, followed by a mild sense of satisfaction when the
missing segment is supplied.
S o m e I m p l i c at i o n s of D ’A n dr a de ’ s A s s u m p t i o n s [ 51 ]
25
exceeds the frequency of the meter that looms largest in the combined cor-
pus of secular and liturgical poetry (Tsur and Bentov 1996). This can be
explained in the following way: as we have seen above, there is an aesthetic
pressure for unity and complexity. The perceptual unity of a verse line is
preserved only if it is kept within the scope of short-term memory, which
functions in the acoustic mode like an echo box. The span of short-term
memory cannot be extended; only the verbal material can be recoded in a
more efficient manner. Poets want to increase the complexity of verse lines
without sacrificing their integrity. Liturgical poetry is more frequently
sung than secular poetry. Music may thus serve as an additional coding
device by alleviating the load on short-term memory. This allows the poet
to increase the complexity of the verse line without violating its integrity.
The same explanation applies to another phenomenon reported by David
Gil (see Tsur and Bentov 1996). Gil explored such noncanonical poetic
genres as cheers at soccer games, political demonstrations, and the chants
of peddlers and hucksters at the marketplace. His corpus of texts (from
Western Europe, the United States, the Middle East, and the Far East) offers
an impressive intercultural sample. Gil found that most of the cheers were
rhythmical, and only a minority of them were sung. In the vast majority of
them the longest unit came last. But he found it puzzling that all the texts
to which a melody was attached had their prosodic weight at the beginning
of the versification units. Here too, I contend, music is enrolled as an addi-
tional coding device to alleviate the burden on short-term memory when
the text increases the load. Briefly, music makes it possible to increase the
complexity of the prosodic unit without violating its integrity.
Perhaps the most intriguing issue is related to biblical cantillations,14
as reported by Ronit Shoshany (2009). She uses the terms iambic and tro-
chaic in a roughly figurative sense: namely, to define divisions in which the
shorter section precedes the longer one or follows it, respectively. She uti-
lizes my proposal that music may relieve the burden on short-term mem-
ory and make it easier for the longer section to precede the shorter one.
Here is her abstract:
This essay discusses the question of the original purpose of accents in bib-
lical text. Scholars differ as to whether the accents were introduced either as
14. “The Hebrew Bible text is annotated with a system of diacritic marks called ‘accents’
(Hebrew טעמיםṭeʕamim; singular טעםṭaʕam ‘taste, sense, reason’). These accents,
assigned to every word in the Bible, parse each verse in minute detail. This complex
system of representation, developed in and around Tiberias over several generations
up until the 10th century . . . , serves several purposes: among other things, it marks
the position of stress, and guides the musical cantillation of the text” (Dresher 2013).
CONCLUSION
My earlier work has been sometimes praised, sometimes censured for pur-
portedly elaborating a theory of poetic universals. I never intended to do
so and am not qualified to make claims regarding universals of any kind.
The most I can claim is that certain conventions and kinds of response
can be found in a number of cultures; but this is a far cry from univer-
sals. I have a more modest ambition: to contrast two approaches to literary
conventions, the migration (influence-hunting) and the cognitive-fossils
(constraints-seeking) approaches. Both approaches assume repeated social
transmission but different states of affairs as points of departure (full-blown
conventions or responses acquired for adaptation purposes) and usually
rely on different conceptions of mental operations—behavioristic and cog-
nitive, respectively.
Nor do I seek to imply that the constraints of the brain impose an abso-
lute determinism on possible literary forms. In fact, my argument is that
S o m e I m p l i c at i o n s of D ’A n dr a de ’ s A s s u m p t i o n s [ 53 ]
45
all imaginable oddities can or could have been invented at one time or
another. This process of “natural selection” through repeated social trans-
mission renders them into conventions by acquiring a good fit to the natural
capacities and constraints of the human brain. Moreover, as we have seen,
the same constraints of the brain may afford several alternative solutions.
This may be one reason for the extreme cultural variation observed. Being
shaped by the constraints of the brain does not necessarily imply “com-
monsense” or “logical.” I have argued elsewhere (Tsur 2008a: 414–422)
that even such “wild inventions” as the metaphysical conceit could result
from the constraints of the brain, of which Dr. Johnson says in his essay on
Abraham Cowley that the reader wonders “by what perverseness or indus-
try they were ever found” (1951 [1779]) (see, e.g., c hapter 6). I argue that
the same implausible conceit could have been invented more than once
through the same apparatus.
The migration approach is frequently associated with factualism. Its
exponents often object to the cognitive-constraints approach on the
grounds that it goes beyond the hard facts by applying to them cognitive
hypotheses; and like all hypotheses, these cannot be true but, rather, at
best, plausible. On closer inspection, however, migration-hunters cannot
make arguments either without having recourse to hypotheses; they only
disguise them as facts. What they can know for a fact is that a certain con-
vention did occur in places A, B, and C. They can also know the geographic
distance between these places. But that the convention migrated from place
A to B and from B to C in many cases is a very plausible hypothesis rather
than a fact. Apart from this, the migration approach can only point out the
occurrence of a convention at a certain place and time but cannot account
for how it acquired a meaning, an emotional import, or a perceptual quality
such as “dignified” or how these could have been communicated to its first
audience. It can, at best, point out that at that time it was a convention to
attribute such and such a meaning or emotional import to a metaphor or to
regard, for example, the dactylic hexameter as “dignified.”
Moreover, even where deliberate migration is explicitly intended, unex-
pected constraints encountered en route may bring about unexpected results.
When, for instance, English Renaissance poets thought that they were
importing the Greek and Latin versification systems into their stress-timed
language, they unwittingly generated the syllabotonic system. When French
poets attempted to import the Greek and Latin versification systems into
their syllable-timed language, they simply failed. When Hungarian poets
attempted to import the same, their results were even more inextricable.
Hungarian iambic is syllabotonic, whereas Hungarian trochees and dactyls
are basically “quantitative,” as in Greek and Latin poetry. Likewise, when
S o m e I m p l i c at i o n s of D ’A n dr a de ’ s A s s u m p t i o n s [ 55 ]
65
7 5
CHAPTER 3
In order to please patrons, poets had to compose panegyrics (love poems) that
satisfied their patrons by showing mastery of the language. The poet would be
conferred prestige after giving praise for his patron’s accomplishments. . . .
The recital of a panegyric was an important formal occasion and provided
an opportunity for the sovereign to demonstrate his generosity publicly by
handsomely rewarding the poet, who, if he genuinely, but secretly admired his
patron, could be inspired to produce truly excellent work.
. . . Al-Mutanabbi was able to feel safe and secure in the company of
Sayf al-Dawlah, who became his intimate friend and comrade in arms. As a
result, al-Mutanabbi was better able to appraise the heroism of the prince
wholeheartedly. . . .
. . . Having gained intimate knowledge of Sayf al-Dawlah, we can presume
al-Mutanabbi then grew to love him as his patron. (2007)
Abd Razak uses indicative statements about the social and literary for-
malities related to panegyrics: “poets had to compose panegyrics,” and “the
recital of a panegyric was an important formal occasion.” As to the warm
1. I learned at school that William Harvey discovered the circulation of blood in the
human body. Wikipedia, however, tells us that the circulation of blood was discovered
as early as 1242 by the Arabian physician Ibn al-Nafis. So Arab poets and Hebrew poets
in Muslim Spain can be considered to have had circulating blood as of the thirteenth
century on. This, however, would still leave al-Mutanabbi and the Hebrew poets of the
golden age without circulation.
personal relationship between al-Mutanabbi and his patron, she uses the
conditional form “if he genuinely, but secretly admired his patron” and
carefully emphasizes affordances rather than facts, as in “was able to feel
safe and secure in the company of Sayf al-Dawlah” and “was better able
to appraise the heroism of the prince wholeheartedly.” She also uses the
hypothetical construction “we can presume al-Mutanabbi then grew to love
him as his patron.” Linguistically, at least, certain things are presented as
known, and other things emerge as conjectures. Briefly, the social constel-
lation as well as the requirements of the genre prescribed the praise and the
emotions to be expressed in these poems, irrespective of what the poets
felt. Is there any evidence that, in some instances at least, the poets did
indeed experience the emotions that the conventions of the occasion and
the genre prescribed for their poems? Furthermore, what kind of evidence
would be deemed satisfactory? Or is it merely a post-Romantic assumption
that poets attempt to express their genuine feelings?
P o e t i c C o n v e n t i o n s a s F o s s i l i z e d C o g n i t i v e De v i c e s [ 59 ]
06
Based on such lines, some critics (e.g., Shim’on Bernstein) came to grim
conclusions as to the personal relationships in the Ibn Ezra family. I call
this the “biographical fallacy.” Hebrew scholar and poet Dan Pagis (1970),
however, compared the use of the first-person singular in this genre (in
both Arabic and Hebrew poetry of the period) with its use in the genre of
“Poems of Complaints.” The latter uses a personal “I,” whereas the former
has an exemplary “I.” This is an example of the condition humaine—that is,
of how each newer generation forgets the older generation—and has noth-
ing to do with what happened in the Ibn Ezra family.
In elegies, as in the elegy on the death of al-Mutanabbi’s grandmother,
authentic autobiographical elements do occasionally occur. But one can-
not make inferences based on this genre regarding the authenticity of
experiences in the panegyric genre, for instance. Furthermore, the elegy
itself abounds in conventional hyperboles, even if we have good reasons
to suppose that the poet experienced a genuine shock at the death of the
deceased. Thus, for instance, the great eleventh-century Hebrew poet
Shlomo Ibn Gabirol wrote a long elegy on the murder (or execution?) of his
friend and benefactor Yekutiel Ibn Hassan, which begins as follows:
קּות ֵיאל ֲא ֶׁשר נִ גְ ָמרּו אֹות ּכִ י ְׁש ָח ִקים לַ ֲחֹלף יֻ ּצָ רּו
ִ ְִּב ֵימי י
[In the days of Yekutiel that are over
(It’s a sign that the skies are doomed to pass away)]
Probably, the more exaggerated the praise and the expression of love,
the less credible it appears. Poets as well as the targets of praise were well
aware of a certain discrepancy between what the poet thought and what
the poet said in this genre. In his discussion on panegyrics in his trea-
tise on poetics (written in Arabic), Moses Ibn Ezra makes the following
remark: “Once a poet went into excesses in praising one of the high state
officials. . . . But the extolled one resented this and answered: I am less
than what you say but more than what you think of me” (1975). This leads
to the question of how we can determine whether a poet did or did not
feel the sentiments being expressed. Let me state at the outset that I don’t
know the answer.
In sixteenth- and seventeenth-century English poetry we encoun-
ter similar problems. Consider the following well-known couplet from
Shakespeare’s Sonnet XVIII:
At one point in my literary education I was told that these lines indi-
cate (or prove?) that Shakespeare was already aware of the immortality
of his poetry. Such studies as J. W. Lever’s The Elizabethan Love Sonnet
(1966) have amply demonstrated that almost every image or idea found
in Shakespeare’s sonnets could be found in innumerable other sonnets
of the time, but Shakespeare simply wrote much better poetry than they
did.2 Countless poets promised immortality to their beloved through
their verses, but most of them have long been forgotten, and rightly
so. Medieval Arabic and Hebrew boasting poems differ from this only
in the degree of exaggeration, not in the degree of the authenticity of
experience.
Elizabethan poets and editors seem to have been aware that the pas-
sions conveyed in their lyric poems were exemplary rather than authen-
tic personal experience; poets use such linguistic devices as metaphors or
the first-person singular for rhetorical purposes. In his admirable study
of Elizabethan poetry, Hallett Smith (1968) collected some illuminat-
ing comments to that effect from the Mirrour for Magistrates, both by an
anonymous annotator of Thomas Watson’s poetry and by Watson him-
self: “Watson observes to the friendly reader in his preface that he hopes the
reader will excuse any faults escaped, ‘in respect of my trauaile in penning
these loue passions, or for pitie of my paines in suffering them (although
2. The same feature was shown for Villon by Italo Siciliano in his monumental
François Villon et les thèmes poétiques du Moyen Age (1934). About sixty-seven years
later this became a commonplace in Villon studies: “[The naïve reader] is captivated by
Georges Brassens singing La ballade des dames du temps jadis, only to find that Villon
wrote two other poems on precisely the same theme, and that every second poet of the
later Middle Ages had made use of the same well-tried cliché” (Taylor 2001: 6). Siciliano
demonstrates that La ballade des dames du temps jadis is simply more euphonious than
the other poems of this genre. To this one might add that the ubi sunt motive was wide-
spread in eleventh-century Hebrew poetry as well.
P o e t i c C o n v e n t i o n s a s F o s s i l i z e d C o g n i t i v e De v i c e s [ 61 ]
26
We may, then, ask why the publication of an essay applying Maslovian theory
to al-Mutanabbi’s poetry is an important event. But we should also be care-
ful about what we have explained using the application of some psychologi-
cal theory. The psychological theory may explain something that has great
poetic significance, but not necessarily the poet’s subjective experience.
It should be recalled that this poetry is not merely highly conventional
but also ornamental. This book discusses the ways in which poetic conven-
tions and ornaments arise. My answer to this question is based on a psy-
choanalytic concept adopted from Ehrenzweig and a “Darwinian” concept
adopted from D’Andrade’s (1980) cognitive anthropology. Briefly, as dis-
cussed in chapter 1, the reasoning is that poetic conventions are psycho-
logical processes fossilized into verbal devices and that in the process of
repeated social transmission, cultural programs come to take forms that
have a good fit to the natural capacities of the human brain. So, what light
does this theoretical model throw on the application of Maslow’s psycho-
logical theory to al-Mutanabbi’s poetry? It will not necessarily explain the
individual poet’s actual experiences and personal emotional development.
But it may perhaps illuminate the psychological processes whose expres-
sions have fossilized in such highly ornamental poetic conventions as
boasting poetry, love poetry, the panegyric, and so forth.
Let us take a look at Abd Razak’s argument from this point of
view: “Maslow’s theory is primarily concerned with that part of human
motivation and behavior which is based on the higher needs.” Take, for
instance, “the need for self-esteem,” as applied to al-Mutanabbi’s poetry:
Assuming that at this point the first three levels of the hierarchy [the search for
safety, the need to belong, and the need for affection] of al-Mutanabbi’s needs
were adequately satisfied, we would expect him to be concerned with the need
for esteem. Maslow distinguishes two types of esteem needs. The first is esteem
from others. This involves the desire for reputation, status, recognition, fame
and a feeling of being useful and necessary. Individuals need to feel respected
and valued by others for their accomplishments and contribution. Self-esteem,
on the other hand, involves a personal desire for feelings of competence, mas-
tery, confidence and capability.
Self-esteem is therefore closely linked to the desire for superiority and
respect from others. (Abd Razak 2007)
P o e t i c C o n v e n t i o n s a s F o s s i l i z e d C o g n i t i v e De v i c e s [ 63 ]
46
Canzoniere CXXXIV
Pace non trovo e non ho da far guerra
e temo, e spero; e ardo e sono un ghiaccio;
e volo sopra ‘l cielo, e giaccio in terra;
e nulla stringo, e tutto il mondo abbraccio.
Tal m’ha in pregion, che non m’apre nè sera,
nè per suo mi riten nè scioglie il laccio;
e non m’ancide Amore, e non mi sferra,
nè mi vuol vivo, nè mi trae d’impaccio.
Veggio senz’occhi, e non ho lingua, e grido;
e bramo di perire, e chieggio aita;
e ho in odio me stesso, e amo altrui.
Pascomi di dolor, piangendo rido;
egualmente mi spiace morte e vita:
in questo stato son, donna, per voi.
This poetic convention has a long history and goes back at least to
Catullus (circa 84–54 bc):
The same convention appears in Ronsard’s original Sonnet No. XII of his
Amours:
P o e t i c C o n v e n t i o n s a s F o s s i l i z e d C o g n i t i v e De v i c e s [ 65 ]
6
P o e t i c C o n v e n t i o n s a s F o s s i l i z e d C o g n i t i v e De v i c e s [ 67 ]
86
ink matches the name of the color (see Stroop 1935). Further experiments
suggest that (unmentioned) category names too, extracted from parallel
items, may interfere with tasks with which they are incongruous. Thus,
“love” or “contrarious passions,” abstracted from the catalog of contradic-
tions, may be active in the background of Petrarch’s (or Wyatt’s) sonnet,
while the reader attends to the catalog of contradictions.
When one encounters some absurdity, one may laugh it off (the comic is,
indeed, a defense against threat) or may admit the authority of the threat
and try to resolve it. In the former case the effect will be frivolous and
delightful, and in the latter, disquieting. Alternatively, one may be able to
resolve it, in which case it will yield great satisfaction. Such satisfaction
would be particularly significant in light of Maslow’s model as quoted by
Abd Razak, with reference to “self-esteem”: it “involves a personal desire
for feelings of competence, mastery, confidence and capability” (2007).
When these absurdities occur in close succession, one is unlikely to be able
to work out the details of the resolutions. This leaves one with several pos-
sible strategies. If the succession of absurdities occurs in a situation that
affects one’s well-being, one will experience disorientation, discomfort,
and distress. If they occur in situations that are unrelated to concrete reali-
ties, specified objects, or actual instances, there are no clues to resolving
the absurdity; hence the most convenient way to cope with it is to laugh it
off. The third possibility is to direct one’s attention away from the contra-
dictions to the common elements of the items on the list.
There is a tendency in the human mind to impute coherence to what is
incoherent. If it can organize disconnected elements into a coherent story
or state of affairs (i.e., into some schema), it will do so. If not, it will search
for common elements (i.e., generate a category). This is what typically hap-
pens in catalogs: attention is directed away from the discontinuous items
(in this case the baffling logical contradictions) to a category that includes
them, thus reducing the force of bafflement.
Suppose the common denominator of the catalog is “The Contrarious
Passions in a Lover.” The reader can divert the baffling contradictions into
the unitary abstraction of love. Such an abstraction is active in the back of
one’s mind, without preempting everything else, just like emotions, thus
generating an emotional quality attributed to the poem. This tendency to
abstract a common denominator is enhanced by the rhythmic quality of the
poem, the recurrence of the verse line as a versification unit, and the rhyme
pattern, which reinforces a sense of parallelism between the items.
Scholars of the source-hunting tradition have claimed that Petrarch’s
sonnet grew out of Provençal riddles, the solution to which was “love.” One
P o e t i c C o n v e n t i o n s a s F o s s i l i z e d C o g n i t i v e De v i c e s [ 69 ]
07
logical contradiction, and then toned down its baffling effect through the
catalog technique, to finally reach a poetic structure that evokes a less
disquieting emotional quality of love, one that is now perceived in the
poem rather than experienced.
As we have seen, the need to impute coherence may force the reader to
ignore individual contradictions to secure a common superordinate that
can resolve the conflicting statements and account for the co-occurrence of
various “contrarieties”—in this case, “love.” There are instances, however,
in which no such superordinates are available. In this case the human mind
searching for coherence may settle for a formal common denominator, the
category “contradictions.” In such cases, since the incongruities cannot be
resolved by meanings, the only way to cope with the succession of absurdities
is to “laugh it off” and treat them as frivolous trifling. At this juncture, having
recourse to personal or impersonal grammatical constructions may make a
substantial difference.
Now, what happens when the catalog of paradoxes cannot be traced back
to conspicuous ambivalent feelings that have been sharpened and fossil-
ized? When the meanings suggested by the logical contradictions can-
not be ascribed to a superordinate category or state of affairs, the brain,
eager to impose coherence upon the disconnected contraries, will not give
up but may abstract the formal category “contradictions,” which is then
repeated throughout the series. Such a poem would quite plausibly be per-
ceived as frivolous playing around. But sometimes something happens to
the frivolous trifling, and the poem may emerge as somehow more serious.
Consider, for instance, the following two poems:
P o e t i c C o n v e n t i o n s a s F o s s i l i z e d C o g n i t i v e De v i c e s [ 71 ]
27
Figure 3.1 “Nu comme ung ver, vestu en president.” David Kleinman’s illustration of
Reuven Tsur’s Hebrew translation of François Villon’s poems Danse Macabre.
Farewell to Love
Since there’s no help, come let us kiss and part;
Nay I have done, you get no more of me;
And I am glad, yea, glad, with all my heart,
That thus so cleanly I myself can free;
Shake hands for ever, cancel all our vows,
And when we meet at any time again,
Be it not seen in either of our brows
That we one jot of former love retain.
Now at the last gasp of loves latest breath,
When his pulse failing, passion speechless lies,
When faith is kneeling by his bed of death,
And innocence is closing up his eyes,
Now if thou would’st, when all have given him over,
From death to life thou might’st him yet recover.
P o e t i c C o n v e n t i o n s a s F o s s i l i z e d C o g n i t i v e De v i c e s [ 73 ]
47
paradoxes. Consider Drayton’s “When first I ended, then I first began.” This
line contains a straightforward logical contradiction between two incompat-
ible statements that are unrelated to any concrete realities, specified object,
or actual instance. Consequently, there are no clues for, or constraints on,
its interpretation. There is not sufficient information to decide whether they
are relevant to the speaker’s well-being. Villon’s line “I die of thirst beside
the fountain,” by contrast, contains an indirect or empirical opposition. In
fact, the inferences from being “beside the fountain” and from “dying of
thirst” conflict. The opposing terms are actually compatible, and they occur
within a concrete situation. It could be literally true as, for instance, in
Or it may suggest another tragic situation, as when one reaches the foun-
tain but dies before being able to drink, and so forth. This situation is clearly
relevant to the speaker’s well-being and suggests a tragic flavor. The effect
is not dependent on Villon’s personal disposition. The line was given to him
by Charles, Duke of Orléans, himself a fine poet, at a poetry contest, where
eleven contestants began their entries with this line. Having set the tone, the
next three paradoxes can also be construed to conform to a tragic experience.
Charles d’Orléans’s own entry went as follows:
While Villon’s shivering with cold and the heat of the fire are both literal,
Charles d’Orléans has “Shaking from cold or the fire of lovers,” choosing
the Petrarchan solution for his contrarieties. This may bestow some figu-
rative meaning on the first line, for example, I am near the object of my
yearning, but I cannot reach her.
Goethe uses the dying-of-thirst-beside-the-fountain metaphor in a
very un-Petrarchan context. In his poem “Das Tagebuch” (The Diary) he
recounts how he found himself in bed with a beautiful virgin, but unex-
pectedly, presumably for moral scruples, he had no erection:
P o e t i c C o n v e n t i o n s a s F o s s i l i z e d C o g n i t i v e De v i c e s [ 75 ]
67
attain a state of tranquility. But even so, this paradoxical statement can be
construed as, for instance, being confronted with a particularly inquisitive
spirit. Thus the image of a complex personality begins to emerge. Then one
may attempt to relate such statements as “Wise in conceit, in act a very sot”
to this emerging personality, which, again, could be literally true.
Villon’s ballade begins with the line “I die of thirst beside the foun-
tain,” where the contradiction is not between two abstract statements but
between a scene term and the speaker’s action. This can be construed as sug-
gesting a tragic situation, followed by three lines in which the paradoxes can
be construed as suggesting some intense physical or psychological anguish.
This sets the scene for the emergence of an intense and coherent human
experience, suggesting, in turn, a complex human personality. Within this
framework, such lines as “I keep winning and remain the loser” suggest a
sophisticated but plausible character. To this one may add the grotesque
image of “Naked as a worm, dressed like a president” or the self-sarcasm of
the misfit: “At dawn I say ‘I bid you good night.’ ” The better the integration
of these features into a unified character, the more credible the character.
The more diversified the features suggested by the paradoxes—assuming
that unity is taken care of—the more complex the emerging character.
Is there any evidence that readers go such a long way to integrate such
a diversity of features? And if so, is there anything in the texts themselves
that would advance or obstruct such an integration? Meir Sternberg (1976)
quotes a series of experiments by Luchins and others on primacy and recency
effects that may clarify two aspects of this issue. The experimenters com-
posed two diametrically opposed passages describing a hypothetical per-
son, Jim. One passage described him as friendly and extroverted; another,
as unfriendly and introverted. Then they combined the two passages in two
different ways: the friendly paragraph first or the unfriendly paragraph
first. The subjects were asked to give their impression of the person and
were divided into two groups, one that read the friendly-paragraph-first
version and another that read the unfriendly-paragraph-first version. The
results showed a strong primacy effect. Subjects tended to interpret the
first part of the description as the real character of the person described.3
Surprisingly enough, in the interviews, subjects in both conditions reported
that they had not noticed the discrepancy in the description. Sternberg
accounts for this curious fact as follows:
Due to the successive order of presentation, the first block was read with an open
mind, while the interpretation of the second—in itself as weighty—was deci-
sively conditioned and colored by the anterior, homogeneous primacy effect. In
other words, the leading block established a perceptual set serving as a frame of
reference to which subsequent information was subordinated as far as possible.
In each case, accordingly, the leading block was taken to represent the “real”
Jim, the “essential nature” of Jim, while the second was taken to describe excep-
tional behavior, which is to be explained away in terms of temporary variations
in mood or circumstances—in short, as a mere qualification of the previously
established conception of character. (1976: 298)
P o e t i c C o n v e n t i o n s a s F o s s i l i z e d C o g n i t i v e De v i c e s [ 77 ]
87
A word must be said about Villon’s rhymes. As the Romantic poets argued,
rhyme is unity in diversity. The more similar the sounds, and the more dif-
ferent the meaning structure of the rhyming words, the more effective the
rhyme. The ballade form has a sophisticated rhyme pattern. In the present
instance, the a and d rhymes of the stanza have two members each; the b and
c rhymes have three. To make things more difficult, all the later stanzas rhyme
with the rhymes set by the first stanza. As a result, poets take the liberty of
rhyming similar grammatical structures (what Jakobson [1960] calls “gram-
matical rhymes”) or even have recourse to the same words with different pre-
fixes. A similar problem arises in the quassida of medieval Hebrew and Arabic
poetry, where the monorhyme needs to rhyme with an indefinite number of
similar-ended words. As to rhyme quality, medieval Hebrew poets explicitly dis-
tinguished between “passable,” “becoming,” and “excellent” rhymes, according
to the number of shared phonemes in the rhyme words. Consider the following
rhymes in Villon’s ballade: espoir–desespoir, aucun–chascun, incertaine–certaine.
In view of such monotony, the complexity of the a and b rhymes in the first
stanza will be all the more salient, for instance. Fontaine is a noun, whereas
loingtaine is an adjective, displaying a contrast on the part-of-speech level. At
the same time, the two words have four consecutive speech sounds in common.
I have already mentioned that the sequence dent à dent necessarily repeats the
same speech sounds and that the nasal feature of the vowels is foreshadowed
in tremble and fontaine. Moreover, the alternating a-b-a-b rhymes have a special
phonetic relationship, which Kenneth Burke called “colliteration” (1957: 369;
see chapter 10). The t of fountaine and the d of dent are cognates: the d is a
voiced t. Whereas fountaine ends with a full consonant, a (nasal) n, the nasal
vowel ẽ is generated by an attenuated n. Thus, in the a-b-a-b rhyme, two sets
of cognate but different speech sounds alternate, yielding a particularly eupho-
nious effect. The three b rhyme words, dent–ardent–président, rhyme a nasal
vowel preceded by a d. Dent is a unitary word, a noun; ardent and président
end with the formative -ent, which derives a noun or an adjective from a verb.
In this case, ardent is an adjective, and président is a noun. What is more, as
far as I can judge as a foreigner, and from another century, président is also
perceived as a unitary word. The adjective ardent intervenes between the two
nouns in the rhyme pattern. Furthermore, the rhymed syllables of both the a
and b rhymes (-taine and -dent) preserve all their speech sounds throughout
the three stanzas. This is not necessarily so in the poetry of the period.
Below I quote the opening stanza of Alain Chartier’s ballade of proverbs
and the opening stanza of Villon’s parody of that poem from Jane Taylor
(2001). Taylor points out significant similarities between them, including
the identical rhyme patterns. By identical rhyme patterns, I suppose she
also means identical specific rhyme endings. At this point, however, I want
to point out the difference in the quality of the rhymes. In both stanzas, the
a rhymes (villain–plain, fain–faing) are unitary, suffixless words (but Villon
happens to rhyme a pair of homonyms with very different meanings). The
d rhymes (including the refrain) are not very exciting: both poets rhyme
the grammatical suffix -eux. This is not very respectable or euphonious. But
while Chartier rhymes a single phoneme (d’amoureux → joieux), Villon adds
a common preceding consonant: r (peureux → qu’amoureux). This r persists
throughout Villon’s poem. The b rhymes in these two poems consist of mas-
culine /i/ rhymes; the c rhymes, of a pair of feminine endings (/i/ plus e muet).
But while in Chartier’s stanza a different consonant precedes the vowels in
each member (enrichi → amy → jalousie → seignori), in Villon’s stanza in all
members the vowels are preceded by a nasal consonant (in the b rhyme by
/m/, in the c rhyme by an /n/ and a palatalized /n/: d’ennemy → endormy →
félonnie → regnie). In the ensuing stanzas as well Villon assiduously inserts
pairs of supporting consonants before the final vowel. In the third stanza
and the “envoi” there are six c rhymes, in all of which there is a supporting
/d/. In the last three paragraphs subtle euphony plays a substantial part in
the appeal of some of Villon’s ballades, and this degree of euphony should by
no means be taken for granted in the poetry of the period.
This rhyme quality is probably not accidental in Villon. He seems to
have been fascinated by rhyme words that are exceptionally similar from
the phonetic point of view and exceptionally dissimilar from the semantic
point of view, as the following rondeau may testify:
Jenin l’Avenu,
Va-t-en aux estuves;
Et toy la venu,
Jenin l’Avenu,
Si te lave nu
Et tu baigne es cuves.
Jenin l’Avenu,
Va-t-en aux estuves.
[Jenin Avenue,
Go there to the heat chambers;
And you came there,
Jenin Avenue,
If you wash yourself naked
And you bathe in tubs,
Jenin Avenue,
Go there to the heat chambers.] (my translation)
P o e t i c C o n v e n t i o n s a s F o s s i l i z e d C o g n i t i v e De v i c e s [ 79 ]
08
(c) Trif ling/Parody
P o e t i c C o n v e n t i o n s a s F o s s i l i z e d C o g n i t i v e De v i c e s [ 81 ]
28
The identical rhyme pattern, per se, of the two poems is not very remark-
able in my mind, because this is one of the most widespread rhyme patterns
in the ballade form, but the identical specific rhyme endings, by contrast,
are. Likewise, the endlessly repeated formula “No a but b” is more than tell-
ing. Moreover, consider Chartier’s line “Ne hault vouloir que d’amoureux.”
Its second half recurs in Villon’s refrain “Ne bien conseillé qu’amoureux.”
Its first part occurs in Villon’s line “Ne haut vouloir que couardie,” amplify-
ing the ironic twist.
“Villon’s ‘proverbs,’ ” Taylor writes, “though just as lapidary, just as
respectable-looking as Alain’s, are pseudo-proverbs whose sense is, to a
moment’s thought, patently absurd: they replace the irritating predictabil-
ity of Chartier’s poem with proverbs which are not trite—but not ‘true.’
What he demonstrates, in other words, is the sheer fatuity of Alain’s exer-
cise.” “But a string of proverbs like Alain’s,” she continues, “by definition has
no function; a poem consisting solely of proverbs subordinates universal
wisdom to the adventitious constraints of rhyme and metre” (2001: 146).4
Taylor ignores the fact here that Villon himself wrote two ballades of prov-
erbs just as “irritatingly predictable” as Chartier’s, as well as the fact that
in his “Ballade du concours de Blois” he used the technique of catalogs of
contradictions to a very different purpose and effect. Nevertheless, her
conception of this parody is well taken. I would add that Villon does not
explicitly criticize Chartier’s platitudes and complacent values of “respect-
able” people, unlike what he does in his parody on the pastoral values of
the “Franc Gontier” poems. But if you put the two poems side by side, this
intertextual relationship casts an ironic light on Chartier’s ballade and
renders it ridiculous (very much in the manner in which Lewis Carroll’s
“Father William” renders Robert Southey’s “Father William” ridiculous).5
4. In a Marcello Mastroianni film (Stanno tutti bene) there is an anecdote in the fam-
ily about an old winemaker who, on his deathbed, reveals a professional secret to his
sons: wine can be made of grapes too, and then it has a special flavor. In the same vein
I want to reveal that sometimes people write poetry not to instruct but for fun. People
write and read ballades that contain catalogs of proverbs not in order to make readers/
listeners wiser but to cause them to have fun. Presumably, the reader knows all these
proverbs by heart. The ballade form has a complicated rhyme scheme but can despite all
this be very euphonious. Readers, for their part, may derive pleasure from the fact that
everything falls into place. The more rigorous the formal constraints, the greater the plea-
sure when they are overcome. That is how I propose to (mis)read Taylor’s comment: “a
poem consisting solely of proverbs subordinates universal wisdom to the adventitious
constraints of rhyme and metre.” Alain Chartier’s ballade is inferior to Villon’s not in that
he “subordinates universal wisdom to the adventitious constraints of rhyme and metre”
but in that “the adventitious constraints of rhyme” are less rigorous in it than in Villon’s
ballades and the solution of the problem is, therefore, less pleasurable (cf. c hapter 7).
5. In what appears to be a Villon entry in an online encyclopedia, Professor Gert
Pinkernell writes: “Besides, V. apparently had a love affair at that time, because the
refrain of the ballade repeats ‘No well-advised but the enamored’ ” (n.d.; my transla-
tion). This is a fine example of the biographical fallacy refuted above by Dan Pagis that
involves mistaking the exemplary nature of the genre for personal experience. This
refrain is based on the reversal of a commonplace of the time, is conspicuously in line
with the other countertruths of the list, and does not indicate an actual love affair any
more than his refrain “But where are the snows of yesteryear?” would indicate that
Villon went on a skiing trip the preceding year.
P o e t i c C o n v e n t i o n s a s F o s s i l i z e d C o g n i t i v e De v i c e s [ 83 ]
48
In this chapter I have focused on certain issues from the broad perspec-
tive of medieval and Renaissance poetry in general. My argument has
three stages. First, in a highly conventional kind of poetry it is dangerous
to make inferences from what the poet writes to what the poet feels (or
thinks). In this respect, I refer to the “biographical fallacy,” (mis)applied
to al-Mutanabbi’s, Moses Ibn Ezra’s, and François Villon’s poetry. I warn
against applying psychological hypotheses to a poet’s empirical personal-
ity through his or her use of poetic conventions. In this respect, in many
genres, when poets say “I,” they do not even necessarily mean themselves
but, rather, a certain exemplary “I,” Everyman. In this case, the first-person
singular is meant to induce the audience to activate certain processing
strategies, to generate certain kinds of effects. Impersonal constructions
serve to induce different processing strategies, with different effects.
This does not imply that psychological theories are irrelevant to lit-
erary research. We, however, need to carefully scrutinize what we have
explained using these psychological theories. In a highly conventional kind
of poetry one may learn very little about the empirical personality of the
poet. I strongly suspect that in this period, a poet’s recourse to this or that
genre reflects a fashion rather than stages in the development of his or
her personality. But we may learn a lot about the origin of conventions.
With reference to such genres as the poetry of boasting and the panegyric,
for instance, Maslow’s theory can describe the psychological processes that
have fossilized into these genre conventions.
Second, following Ehrenzweig and D’Andrade, I have tried to explore the
nature of poetic conventions and their evolution. As Laozi said, every jour-
ney, even the longest one, must begin with a single step. Likewise, every
poetic tradition, even the longest and widest one, must begin with one
poet composing one poem, or one verse line, for one particular audience or,
perhaps, several poets, each composing one poem for the same audience or
several different audiences. How, from such humble beginnings, did large
poetic traditions emerge? In a world in which there was no radio or televi-
sion, no Internet, no airplanes, only horses, camels, and sailboats? I am not
proposing an actual historic process but, rather, a hypothesis that must be
verified against a wide variety of actual processes.
On this issue I have adopted D’Andrade’s position: “In the process of
repeated social transmission, cultural programs come to take forms which
have a good fit to the natural capacities and constraints of the human
brain.” One promising beginning for such a study would be, with the neces-
sary changes, Sir Frederick Bartlett’s classical experiment with the Eskimo
story “The War of the Ghosts.” Bartlett’s purpose in this study was slightly
different. He tried to determine how a story for which we have no mental
schemata assumes, through repeated transmission, a good fit to the men-
tal schemata we entertain. When the story achieved a good fit to a mental
schema prevalent in Western culture, the story stopped changing in sub-
sequent transmissions (see chapter 2). I take up the process at a relatively
late stage, when defense mechanisms are used against expressive devices,
which are turned, according to Ehrenzweig, into style. In this respect,
I claim, artistic conventions and ornaments are fossilized psychological
processes.
Third, I illustrate this model with reference to a widespread convention
in medieval and Renaissance poetry: catalogs of incongruous or incom-
patible statements. I start at a point well known to psychoanalytic the-
ory: ambivalence. Ambivalence is a potential source of undesirable stress.
Dreams and fairy tales handle ambivalence by splitting the object of ambiv-
alent attitudes into two opposing figures, for example, a good mother and
a wicked stepmother. It would appear that a straightforward wicked step-
mother is less threatening than an ambivalent attitude. The poetic conven-
tion of contradictions originates not in a split of the ambivalent attitude
but in applying to it the cognitive mechanism of sharpening,6 thus generat-
ing explicit contrarieties. Again, baffling contradictions seem to be less dis-
tressing than ambivalent attitudes, but they are still threatening. One way
to defuse the disquieting element in contradictions is to generate a catalog
of contradictions. The constraints of human memory induce the mind to
abstract a common category from parallel items. At the same time, this
helps shift attention from the contradictions to their common element or
superordinate category. The most widespread variety of this convention is
that in which the common superordinate category is “love,” which is active
in the back of the mind, without preempting everything else. At the same
time, the repetition of parallel items generates a rhythm that amplifies and
diffuses the superordinate category “love,” and via interaction with other
poetic devices, such as the use of the first-person singular, the mnemonic
device is turned into an emotional quality of the poem. It is claimed that
the ambivalence of love and the need to remove the disquieting element has
a good fit to the capacities and constraints of the human brain. I suggest
that even catalogs of contradictions that historically do not originate in
the ambivalent emotions of love may eventually be accommodated in this
tradition and thus assume a good fit to the natural capacities of the human
6. See chapter 1.
P o e t i c C o n v e n t i o n s a s F o s s i l i z e d C o g n i t i v e De v i c e s [ 85 ]
68
brain. There are, however, instances in which the “love” solution simply
does not work. Then, to get rid of the disturbing element of the unresolved
contradictions, the human mind may try other strategies, including the
construal of some complex personality roughly compatible with the list
of paradoxes. And if this fails as well, we can still laugh it off and treat it
as frivolous trifling, making fun. We have seen one instance in which this
solution is subordinated to a higher degree of making fun, namely, parody.
CHAPTER 4
Paradoxically enough, this passage gives us a clue not only to how “the rep-
ertoire of rhetorical figures” provides insight but also to how “pigeonholing
gives certainty, but no insight” (de Mourgues 1953). The crucial clause seems
to be “not so much the operations that are important as the reassurance.”
However, “instructions” and “reassurance” are not and should not be taken to
be two independent factors, as this clause implies. “Reassurance” should not be
considered as a goal in itself but, rather, as a psychological precondition under
which the instructions can be carried out. Pigeonholing occurs when readers
or critics are content with “the reassurance that what seems odd is perfectly
acceptable since it is figurative expression,” or a convention of a genre, or a
symptom of a psychological syndrome, but have no expectation of getting (or
executing) any further instructions. For these readers, carrying out instruc-
tions could at best lead to the final “reassurance that what seems odd is per-
fectly acceptable” and at worst might unnecessarily complicate things or even
lead to new ambiguities. So why bother? In George Klein’s (1970: 134–136)
terminology, they tend to level out differences between unique works. In such
cases, explication of the poem may be of decisive importance.
For instance, we are taught that the key characteristics of the ballad style
are repetition (often merely decorative) and ellipsis (often quite easily com-
pleted). When we ask, “What does this repetition or that ellipsis contribute
to the ballad?” we often are given the diagnosis, “It is a characteristic of bal-
lad style,” or, “It contributes to the balladic atmosphere.” Whenever there
is threefold repetition, one can expect to be told that “it contributes to the
popular-ballad style.” In other words, repetition and ellipsis are treated as
merely symptoms of a style.
I would by no means deny that repetition does contribute to the balladic
character of a ballad. I would even admit that these characteristics give a
certain charm and naïveté to the ballad. Nevertheless I would submit that
this naïveté is sometimes the consequence of repressing some utterly non-
naïve and expressive feature of the poem.
Let us have a look at “Edward”:
1
“Why dois your brand sae drap wi bluid,
Edward, Edward,
Why dois your brand sae drap wi bluid,
And why so sad gang yee O?”
“O I hae killed my hauke sae guid,
Mither, mither,
O I hae killed my hauke sae guid,
And I had nae mair bot hee O.”
2
“Your Haukis bluid was nevir sae reid,
Edward, Edward,
Your Haukis bluid was nevir sae reid,
My dear son I tell thee O.”
“O I hae killed my reid-roan steid,
Mither, mither,
O I hae killed my reid-roan steid,
That erst was sae fair and frie O.”
3
“Your steid was auld, and ye hae gat mair,
Edward, Edward,
Your steid was auld, and ye hae gat mair,
Sum other dule ye drie O.”
“O I hae killed my fadir deir,
Mither, mither,
O I hae killed my fadir deir,
Alas and wae is mee O!”
4
“And whatten penance wul ye drie for that,
Edward, Edward,
And whatten penance wul ye drie for that,
My deir son, now tell mee O.”
“Ile set my feit in yonder boat,
Mither, mither,
Ile set my feit in yonder boat,
And Ile fare ovir the sea O.”
5
“And what wul ye doe wi your towirs and your ha,
Edward, Edward,
And what wul ye doe wi your towirs and your ha,
That were sae fair to see O?”
“Ile let thame stand tul they doun fa,
Mither, mither,
Ile let thame stand tul they doun fa,
For here nevir mair maun I bee O.”
6
“And what wul ye leive to your bairns and your wife,
Edward, Edward,
F r o z e n F or m u l a e a n d E x p r e s s i v e F or c e [ 89 ]
09
* The shocking revelation at the end of stanza 3 and the brutal statement
at the end of stanza 7 contribute to the quality of sharpness.
* All the last lines, except for in stanzas 3 and 7, are syndetic; all the syndetic
phrases have a connecting character (that, for, and), never contrasting.
* The abrupt, asyndetic lines at the end of the third and the seventh
stanza are perceived as indicating a break. In fact, the second to the last
line of the stanza becomes “sharper” when it is end-stopped, in virtue of
the asyndetic new start in the next line.
* The last line of the ballad reveals a hitherto concealed, highly essential
piece of information. The repetitions in this stanza, although antici-
pated, are highly functional.
* “The curse of hell frae me sall ye beir” strikes the reader by its unex-
pected character. The apostrophe “Mither, mither” has a double func-
tion here: (a) In this unexpected context, it regains some of its original
emotive quality; (b) it defers the solution to this striking riddle. The rep-
etition, in spite of its expectedness, further stretches the reader’s antici-
pation, until the last line finally conveys the unexpected, long-awaited
solution.
* In the last line of the third stanza, “Alas and wae is mee O!” the formu-
laic “O” occurs after a sequence of two parallel interjections, reviving its
meaning and restoring its high emphasis.
This ballad has action in the Aristotelian sense: It has a peripeteia (rever-
sal) with an anagnorisis (revelation), or to be precise, two revelations (one
at the end of the third stanza and one at the end of the seventh); that is,
there is a shift from one extreme situation to its opposite, in this case from
ignorance to knowledge.
The ballad itself is seriously limited in its expressive resources. It consists
of a series of questions and answers, expressed in a rigidly formulaic way
that leaves virtually no room for dramatic manipulation. Nevertheless the
peripeteia is brought about with pointed theatricality. There is no continu-
ous action but, rather, a series of fragmentary, abrupt, formulaic dialogues;
it is up to the reader to complete the situation and generate hypotheses to
make a coherent whole out of the fragments.
Stanza 1, besides advancing the action through dialogue, sets the stage
for the action and provides the necessary information, simultaneously
serving as exposition. We learn that the dialogue occurs between mother
and son. Metonymies bring information to the foreground that lingers in
the background (in the immediate past or within the protagonist). From
“your brand sae drap wi bluid,” the reader infers that Edward has just
committed a murder. Sad in medieval usage frequently meant “heavy,
firm, steadfast.” Consequently, “so sad gang yee” is primarily a physical
metonymy, indicating Edward’s state of mind. Both together awaken the
reader’s curiosity and seem to be sufficient motivation for the mother’s
inquisitiveness.
F r o z e n F or m u l a e a n d E x p r e s s i v e F or c e [ 91 ]
29
not know the end of the ballad. They observe the mother’s steady ques-
tioning. During the first stanza, they could conclude that these are the
questions of a loving mother, troubled by her son’s unusual appearance.
During the second and third stanza, readers are led to modify this appre-
ciation somewhat: The mother makes statements showing that she doesn’t
believe Edward, and Edward produces one pretext after the other, but his
mother is cleverer than that; she presses him with his back to the wall.
Revelation I (as opposed to the preceding pretexts) throws new light on
this hide-and-seek. Obviously, Edward may be trying to conceal the terrible
truth from his mother. Revelation II changes the readers’ point of view
and creates a new puzzle: Why does Edward conceal the murder from his
mother, when he knows that she is bound to welcome it?
There is an obvious Oedipus-like situation in this ballad. The son kills his
father, in order to please his mother. Over the course of the dialogue, readers
gradually discover its meaning, as it emerges little by little, through their
more or less adequate assumptions and expectations, while the ballad
keeps refuting or confirming them. This emerging meaning is more com-
plex and rich and qualifies the basic Oedipal situation. Thus it cannot be
summed up but, rather, must be followed step by step. The mother is wor-
riedly inquisitive upon seeing her son so troubled. Although the readers’
perception of the events is changed by the two revelations, they are not
expected to abandon their interpretation but, rather, qualify it constantly.
At the beginning we hear that the son looks sad—or, rather, walks sad. Sad
may have a wide range of degrees, even when qualified by so. This vague-
ness renders the word ambiguous, and as we shall see, this ambiguity will
prove crucial.
Readers may be puzzled by the mother’s expertness in comparing blood
of different hues: “Your Haukis bluid was nevir sae reid.” One may even
wonder whether there is any real difference between the possible degrees
of redness. Therefore, readers are induced to find some metaphorical mean-
ing. Assuming that the mother does not know more at this stage than the
readers, one might conjecture that there is some metonymic connection
between this blood and Edward’s sorrow, and red being an intense color,
the mother’s question might imply something like “Your sorrow is too
intense to let me believe that it is only your hawk you have killed, so let
me question you further.” But since this is not explicitly stated, there is
some mysterious flavor about the mother’s illogical statement that may be
readily ascribed to the “balladic atmosphere.” Since the readers do not see
Edward, they may attribute several meanings to sad of varying intensities
for having slain his hawk or having murdered his father. When he reveals
the murder, redness acquires connotations of guilt (without lessening the
F r o z e n F or m u l a e a n d E x p r e s s i v e F or c e [ 93 ]
49
he has done, was not afraid that his mother would be too hard upon him.
One might even be tempted to interpret her “And why so sad gang yee O?”
as an invitation, “Let us now make merry.” Nevertheless, Edward tries to
conceal what he has done, as if he were rather unwilling to let his mother
feel satisfaction, let alone express her affection for him. This shameful
confession of his crime becomes a shameful confession of his crime (that
is, both mother and son know all the elements of the situation very well,
but Edward now shies away from her participation in his father’s murder).
This throws some new light on the mother’s stubborn inquisitiveness: not
being satisfied with the mere effects of the murder, she wants to have her
share of her son’s affection (emphasized, again, by “your ain mither deir” in
the last stanza). One might then reinterpret their disguised discussion of
the hawk’s blood and its hue as though each of the two had said: “I know
exactly what you mean, but I want you to speak out first.”
This reading of “Edward” captures the hypothetical nature of our inter-
pretation. Throughout the first three stanzas, readers can project a con-
sistent situation on the abrupt dialogue. They imagine an initial set of
conditions, compatible with the questions and answers presented in the
text, as well as with known human behavior and mental processing. This
initial set of conditions can be considered highly plausible in view of the
available evidence but not strictly true:
F r o z e n F or m u l a e a n d E x p r e s s i v e F or c e [ 95 ]
69
A final word about the mother’s reasons for having her husband mur-
dered by her son. Saying that she wanted to enact her part in the Oedipal
situation would be a platitude. It seems to me that an illuminating, though
disquieting, answer can be found in Aristotle’s Poetics, chapter 9. When
dealing with the difference between history and poetry, he remarks that
the latter is more philosophical because history deals with particulars such
as “what Alcibiades did or suffered,” whereas poetry deals with universals
such as “how a person of a certain type will on occasion speak or act, accord-
ing to the law of probability or necessity” (1951: 37–38). Following this line
of thought, the mother’s immediate motives seem of no relevance to the
ballad. The murder provides the opportunity for people of certain types to
act, while its lack of explicit motivation contributes to the mysterious bal-
ladic atmosphere—an atmosphere that is easily eliminated when we supply
a reason for her action such as in the case of the Oedipus complex, which
thus enables a rapid closure on the readers’ part.
Elsewhere (Tsur 1975, 2006c: 11–77) I have criticized the psychological
syndromes of rigidity and intolerance of ambiguity derived from research
on perception and personality. One of their most notorious symptoms is
the inability to be “in uncertainty, mysteries, doubts,” without any “irrita-
ble reaching after fact and reason”—in Keats’s famous phrase. It is readily
apparent that the Aristotelian answer to the question “Why did the mother
want to have her husband murdered by her son?” is bound to frustrate any
“irritable reaching after fact and reason.” Another possible symptom of this
syndrome is the concern with concrete details and their functioning—in
the present case, namely, a preference for dealing with such particulars as
“what Alcibiades did or suffered” rather than with such universals as “how
a person of a certain type will on occasion speak or act, according to the law
of probability or necessity.” The law of probability in Aristotle’s proposition
poses another difficulty to rigid individuals, since one of the symptoms of
rigidity is an unwillingness to assume an attitude toward the merely pos-
sible (Goldstein and Scheerer 1941; see also Frenkel-Brunswik 1968: 136).
I began this chapter by making a distinction between pigeonholing and a
set of instructions that readers can apply. This metacritical distinction is by
no means recent in critical theory. Apparently pigeonholing in literary criti-
cism may thus satisfy some personal need rather than the requirements of
any discipline or critical school. There is probably no critical art so readily
associated with labeling as classical rhetoric. Nevertheless, the following
observation can be found in one of the classical authorities on rhetoric:
For it makes no difference by which name either is called, so long as its stylis-
tic value is apparent, since the meaning of things is not altered by a change of
name. For just as men remain the same, even though they adopt a new name,
so these artifices will produce exactly the same effect, whether they are styled
tropes or figures, since their values lie not in their names, but in their effect.
(Quintilian n.d.: IX, 7–8)
These pairs of terms contrast two different critical activities, which, in turn,
are frequently behavioral corollaries of different psychological attitudes.
Pigeonholing, instead of bringing out the complexity of a given experience,
tends to reduce experience to a single item that falls into acknowledged
and well-classified categories (the Oedipus complex, the ballad genre).
Classifying moods, feelings, or human situations is hence a means of tak-
ing the disquieting element out of them.
F r o z e n F or m u l a e a n d E x p r e s s i v e F or c e [ 97 ]
89
CHAPTER 5
1. Coauthored by Idit Einat-Nov. The theoretical framework and the discussion of
Merkabah mysticism are reproduced here from Tsur 2003. The analysis of poems from
the Spanish period is taken from Einat-Nov 1995.
01
Elior), on the one hand, and on cognitive poetics and Ehrenzweig’s studies
in the psychology of art, on the other. We will utilize two further illumi-
nating insights of Ehrenzweig’s (1965), which he applies to music and the
visual arts.
The first insight is the one underlying this book: we apply to poetic
language the concept Ehrenzweig developed with reference to the visual
arts and music regarding the generation of ornaments. He elaborates at
great length, as we have seen, on the defense mechanisms human soci-
ety invokes to protect itself from the excessive expressive force of artistic
devices and turn them into style, that is, harmless ornaments. As we saw
in chapter 1, Ehrenzweig distinguishes three stages in the development of
artistic devices. In what he defined as Stage 1, artistic devices are perceived
subliminally and affect what he termed the “depth mind.” As these devices
become more emphatic, their emotional appeal increases, so long as they
do not become consciously perceptible. If they do become somewhat per-
ceptible, they are considered in bad taste or cheaply emotional. Ehrenzweig
called this the second stage. In the third stage, these devices turn into orna-
ments, with drastically reduced emotional appeal. In c hapter 4 we added a
fourth one, when artistic structure revives a frozen formula. We are dealing
here with some more extreme cases of formulaic poetry than in c hapter 4.
As to Ehrenzweig’s second insight, following Freud, he speaks of “oce-
anic dedifferentiation” in relation to mystical experience. William James’s
Varieties of Religious Experience indicates that this concept is consistent with
the subjective experiences reported by various individuals. Ehrenzweig
translated these psychoanalytic notions into a set of cognitive-Gestaltist
terms that can be directly applied to the subtle details of works of art. He
speaks of “a creative ego rhythm that swings between focused Gestalt and
an oceanic undifferentiation”:
The London psychoanalysts D. W. Winnicott and Marion Milner stressed the
importance for a creative ego to be able to suspend the boundaries between self
and not-self to become more at home in a world of reality where objects and self
are clearly held apart. . . . Seen in this way, the oceanic experience of fusion, of
a “return to the womb,” represents the minimum content of all art; Freud saw
in it only the basic religious experience. But it seems now that it belongs to all
creativity. (1970: 135)
A r t i s t i c De v i c e s a n d M y s t i c a l Q ua l i t i e s [ 101 ]
021
ANCIENT MERKABAH HYMNS
4. Rabbi Solomon Ibn Gabirol (1021–1058) was the greatest Hebrew poet and a phi-
losopher in Arabic in the Spanish golden age.
5. Rabbi Itshak Ibn Ghiyat (1038–1089) was a minor Hebrew poet in the Spanish
golden age.
period of almost a thousand years, from the first century bc to the tenth
ad (Scholem 1961: 40). Most of the tracts are called “Hekhalot books,”
that is, descriptions of the hekhaloth, the heavenly halls or palaces through
which the visionary passes and in the seventh and last of which there rises
the throne of divine glory (Scholem 1961: 45). Scholem writes, “There is
little hope that we shall ever learn the true identity of the men who were
the first to make an attempt, still recognizable and describable, to invest
Judaism with the glory of mystical splendor” (1961: 41):
The Greater Hekhaloth presents us with a large number of Hebrew hymns, which
it treats in an unusual manner: the very same hymns are characterized by the
text as representing two different types of songs. On the one hand, the hymns
are addressed to the throne and to Him who sits upon it, and are described as
celestial songs of praise sung by “the Holy Living Creatures” (Hayyoth ha-Kodesh)
who, in Ezekiel 1:5 ff., are the bearers of the throne. On the other hand, these
same hymns are the ones the mystic is instructed to recite before and during his
ecstatic ascent to heaven (which, in a very curious and so far unexplained change
of phraseology, is always referred to in this text as a descent unto the Merkabah).
The hymns describe, in a plethora of solemn phrases, the spirit of majesty and
solemnity that permeates the heavenly realm, “the Palaces of Silence” in which
God’s Shekhinah dwells. (1960: 20)
Before turning to the poetic text itself, we propose two short semantic
exercises concerning two phrases in the last sentence of the foregoing quo-
tation: “the Palaces of Silence” and “God’s Shekhinah.” A palace is a large,
impressive, and majestic building of imposing magnificence; it is usually
the official residence of a sovereign, bishop, or other exalted personage.
The orientation mechanism collects information from the environment
that specifies one’s place in and relationship to the surrounding space. The
huge hall of a palace casts self-specifying information on the perceiving
self that makes the individual feel extremely small (which appears to be the
reason why ancient sovereigns chose such buildings for a residence—and
elevated thrones for seats—to inspire their subjects’ awe). Great silence
deprives the person of information about the surroundings and thus may
reinforce such a feeling. “Palaces of Silence” may suggest real palaces domi-
nated by great silence, casting self-specifying information of smallness.
One of the assumptions underlying the present work is that it is the orien-
tation mechanism that handles such self-specifying information, involving
a ready integration of many inputs at once and producing a diffuse output.
This has an intense nonrational effect that, when heightened, generates a
quality similar to altered states of consciousness, such as meditation and
A r t i s t i c De v i c e s a n d M y s t i c a l Q ua l i t i e s [ 103 ]
0
41
Excerpt 1.
וָ יוֹם מ ֶּׁש ַבח וְ ִשׁ ָירה ֶשׁל יוֹם ִ
ִמּגִ ילָ ה וְ ִרּנָ ה ֶשׁל עִ ִּתים וְ עִ ִּתים
דוֹשׁים ִ וּמ ִהּגָ יוֹן ַהּיוֹצֵ א ִמ ִּפי ְק ִ
מ ִּפי ְמ ָשׁ ְר ִתים ִ וּמּנִ ּגוּן ַה ִּמ ְתּגַ ֵּב רִ
ָה ֵרי ֵאשׁ וְ גִ ְבעוֹת לֶ ָה ָבה
נִ צְ ָּברוֹת וְ נִ גְ נָ זוֹת וְ נִ ָּתכוֹת ְּבכָ ל־יוֹם
'וכו כדבר שנאמר קדוש קדוש קדוש
the Hekhaloth literature does not deal with ecstatic outbursts and their inciden-
tal individual interpretation. . . . The unified conception and the routine descrip-
tions of the contents of the visions in the Hekhaloth literature indicate that they
do not leave speculative or descriptive space for the interpretation of the vision
of the descender unto the merkabah. Pluralism of conception, view, observation
or interpretation is not allowed; the meaning, order and conditions of the vision
are firmly determined. (1987: 93)
In other words, the descriptions of the heavenly visions were already fro-
zen in ancient mysticism into “meaningless formulae.” They do not express
individual experience but become “an objective fact and pattern of refer-
ence for metaphysical reality” (Elior 1987: 13). Thus, for instance, the
unique, sublime biblical event, “suddenly a chariot of fire appeared with
horses of fire . . . ; and Elijah went up by a whirlwind into heaven” (2 Kings
2:11), is turned into a routine practice in the Merkabah (Chariot) literature.
This could epitomize the phenomenon pointed out by Ehrenzweig.
Later in this chapter we shall see that the frozen formulae of the
Merkabah hymns become even more rigid in medieval Hebrew poetry in
Spain. In light of Ehrenzweig’s hypothesis, the rigid formulae of this poetry
may still have had too much expressive power for the later poets.
6. The grammatical form of the Hebrew verbs for “proceed” and “welleth forth” is
more like the present participle than a finite verb. Unless otherwise noted, all transla-
tions are ours.
A r t i s t i c De v i c e s a n d M y s t i c a l Q ua l i t i e s [ 105 ]
061
When one of us (Tsur) first read Scholem’s book forty years ago, he had
a strong intuition that after the four incantatory lines there was a switch
to some overwhelming quality that powerfully engulfed him. At that time
he could not account for it. Today we know that it suggests some struc-
tural resemblance to oceanic dedifferentiation. The last two lines, contain-
ing the subject and the predicate phrases, also shift the poetic strategy of
the hymn. Fire and flame in the second-to-last line continue the pattern of
obtrusive formulaic reiteration. But by the same token they introduce a
gestalt-free, unstable, threatening entity, radiating excessive energy. From
the figurative point of view, the genitive phrase “Mountains of fire and
hills of flame” may be construed similarly to the genitive phrase “Palaces
of Silence.” When fire is the referring member of the phrase, Mountains
suggests “enormous mass”; the feature [+ stability] is deleted. However,
Mountains and hills reinforce in each other their literal meaning; when
they are the referring members, they indicate a vast landscape to which
the perceiving self may relate, activating the orientation mechanism. The
three parallel verb phrases “Are piled up and hidden and poured out” not
only recapitulate the incantatory effect, in that the contrast between the
absence of finite verbs in the first five lines and the sudden pileup of three
consecutive finite verbs in the last line introduces much movement; by
the same token, the meaning of the verbs indicates the intense fluctua-
tion of stable scenery. “Poured out” is associated with liquids (in fact, the
Hebrew verb suggests both pouring and alteration from a solid to a liquid
state, usually by heat), and the sequence as a whole seems to subvert the
surrounding landscape with its self-specifying information. The landscape
itself is unstable: it comes into existence, disappears, and reappears.
Finally, the whole presentation has an interesting existential aspect. The
syntactic agents (“holy ones,” “Servants”) are conspicuously inactive. It is
the praise, the song, the jubilation, the exultation, the utterances, and the
melody that are active, displaying voluntary or spontaneous impulses to
act by employing “linguistic devices that shift the agent to a non-volitional
role” (Balaban 1999: 259). Balaban found empirically that people who report
overwhelming religious experiences “tend to manipulate their own self into
a non-volitional syntactic role” (1999: 259). The praise, the song, the jubi-
lation, the exultation, the utterances, and the melody occur, as it were, of
themselves. Such a consistent syntactic strategy indicates a suspension of
voluntary control. As we have observed, “Mountains of fire” and “hills of
flame” serve as referential terms for the orientation of the self, but at the
same time they have no objective existence at all. They issue from, or their
pileup is caused by, the praise, the song, the jubilation, the exultation, the
utterances, and the melody, which, in turn, issue from the mouth of the agents.
A r t i s t i c De v i c e s a n d M y s t i c a l Q ua l i t i e s [ 107 ]
0
81
Excerpt 2.
אוּרים וּלְ ַמ ָּטה ִּב ְבנֵ י ְּב ִח ִירים ִ לְ ַמעְ לָ ה ְּב ֵאלֵ י ֵאל נִ ְק ָּדשׁ
וּלְ ַמ ָּטה ְּב ַד ַעת ְּדגָ לִ ים לְ ַמעְ לָ ה ְּבגַ עַ שּׁגַ לְ ּגַ ּלִ ים
ת ַה ְּבזָ ִקים וּלְ ַמ ָּטה ְּבוַ ַעדוְ ִת ִיקים לְ ַמעְ לָ ה ְּב ִהּלוֹ
שׁוּקיםִ וּלְ ַמ ָּטה ְּבח ֶֹסן ֲח לְ ַמעְ לָ ה ְּבזִ ְמ ַרתזִּקִים
וּלְ ַמ ָּטה ְּביִ ְק ֵרייְ ָח ִסים לְ ַמעְ לָ ה ְּבט ַֹה ר ָט ִסים
בוּבים ִ ְ וּלְ ַמ ָּטה ְּבלֶ ַקחל רוּבים ִ ְלְ ַמעְ לָ ה ְּבכֶ ֶתרּכ
וּלְ ַמ ָּטה ְּבנֵ צַ חנְ ִסיכִ ים לְ ַמעְ לָ ה ְּב ַמעֲ נֵ ה ְמלָ כִ ים
טוּפים
ִ וּלְ ַמ ָּטה ְּב ֶע ֶתר ֲע שׂ ָר ִפים ְ לְ ַמעְ לָ ה ְּב ִשׂיחַ
רוּפים ִ ְ וּלְ ַמ ָּטה ְּבצֶ ַרח צ רוּפים ִ לְ ַמעְ לָ ה ְּב ֶפצַ ח ְּפ
אשׁים ִ וּלְ ַמ ָּטה ְּב ִרּנַ ת ָר דוֹשׁים ִ לְ ַמעְ לָ ה ִּב ְק ַהל ְק
קוּפים ִ קוּפים וּלְ ַמ ָּטה ְּבתֹכֶ ן ְּת ִ שׁ ְ לְ ַמעְ לָ ה ְּב ִשׁ ַירת
A r t i s t i c De v i c e s a n d M y s t i c a l Q ua l i t i e s [ 109 ]
0
1
celestial worship to human prayers (cf. Elior 1994: 33). On the pho-
nological level, three formal devices are prominent in Excerpt 2: the
acrostic from aleph to tav, each speech sound in the acrostic is repeated
(alliteration), and in each verse line the first hemistich rhymes with
the second.
These piyutim and the Hekhalot hymn share an insistence on songs of
glory conveyed by a pattern of repeated identical verbal structures. There
is, however, a conspicuous difference in this respect regarding the length
of the catalogs, which in piyutim are, as a rule, far more extensive. The
longish catalogs are one of the factors that contribute to the artificial
and frivolous impression of these Spanish piyutim. Another difference
between the Hekhalot hymn and the Spanish piyutim concerns the degree
of similarity between the verbal patterns. In the hymn it is not a verba-
tim repetition but, rather, the repetition of a syntactic-semantic scheme;
moreover, this abstract scheme also changes from lines 1–2 to lines 3–4.
Consequently, “unity in variety” is perceived, resulting in relative com-
plexity. In the majority of the Spanish piyutim discussed here, by con-
trast, verbatim repetition is prominent in the verbal schemes, evoking an
impression of “tediousness” or “lack of progress.” Owing to the acrostic,
the fixed length of verse lines, and the exact repetitions in these piyutim,
rigid order and thought-out organization become their most prominent
features. The single words that change in accordance with the needs of
the acrostic, occurring in strictly reiterated syntactic patterns, reinforce
the common elements in each other. Thus they obliterate each other’s
unique semantic features and contribute to the impression of artificial-
ness and technical frivolity. An additional difference between repetition in
the hymn and repetition in the piyutim concerns syntactic suspension: in
the former, the meaning unit is completed only after the catalogs in lines
1–4—in fact, only in line 6. In the latter, by contrast, each one of the
analogous verse lines is, at the same time, a complete syntactic-semantic
unit. In the catalogs consisting of such syntactically complete units, the
impression of monotony and the sense of “saturation” are amplified: there
is no development from one verse line to another, and the completed
units arouse no expectations for new information, unlike the incomplete
units of the hymn, which require their completion in a verbal unit beyond
the opening catalog. As we have said, the verse lines “Mountains of fire
and hills of flame / Are piled up and hidden and poured out each day” do
indeed contain a new syntactic structure, but its relationship to the pre-
ceding units is not sufficiently clear.
A r t i s t i c De v i c e s a n d M y s t i c a l Q ua l i t i e s [ 111 ]
2
1
Excerpt 3.
שׁירוֹת יַ ְשׁ ִּביצוּ ִ ְּב ֶשׁ ֶקד ְּב ֶשׁ ַבח,קוּפים ְ ּב ֶשׁנֶ ן
ִ ְּת
ְּבלַ ַהב ְּבלֶ ֶמד ְּבלֶ ַקחלְ ָבבוֹת יָ לִ יצוּ
מּלוֹת יַ ְמ ִריצוּ ִ ְּב ַמעַ ן ְּב ֶמלֶ ץ ְּב ֶמ ֶר ץ
ְּב ֶהגֶ ה ְּבהוֹד ְּב ָה ָדר ָהגוּת יְ ַה ְר ִּביצוּ
הוּרים יָ צִ יצוּ ִ ְְּבצַ ַהל ְּבצֶ וַ ח ְּבצֶ ַר חצ
ְּבעֶ ֶתר ְּבעֶ ֶרץ ְּבעֶ לֶ ץעֲ נָ וִ ים יַ ֲעלִ יצוּ
ְּבי ֶֹשׁר ְּבי ֶֹתר ְּבי ֶֹק ריְ ֵשׁנִ יםיָ ִקיצוּ
ְּב ַר ַחשׁ ְּב ַרעַ שׁ ְּב ֶרגֶ שׁ ַרגְ לָ ם יָ ִריצוּ
ֹלהי יִ ְשׂ ָר ֵאל יַ ֲע ִריצוּ
ֵ וְ ִה ְק ִּדישׁוּ ֶאת ְקדוֹשׁ יַ עֲ קֹב וְ ֶאת ֱא
tǝquphim bǝšenen bǝšɛqɛd bǝšɛvaḥ širot jašbiṣu
bǝlahav bǝlɛmɛd bǝlɛqaḥ lǝvavot jaliṣu
bǝmaʕan bǝmɛlɛṣ bǝmɛrɛṣ milot jamriṣu
bǝhɛgɛ bǝhod bǝhadar hagut jǝharbiṣu
bǝṣahal bǝṣɛvaḥ bǝṣɛraḥ ṣǝhurim jaṣiṣu
bǝʕɛtɛr bǝʕɛrɛṣ bǝʕɛlɛṣ ʕanavim jaʕaliṣu
bǝjošɛr bǝjotɛr bǝjoqɛr jǝšenim jaqiṣu
bǝraḥaš bǝraʕaš bǝrɛgɛš raglajim jariṣu
vǝhiqdišu ʔɛt qǝdoš Jaʕqov vǝʔɛt ʔɛlohe Jisrael jaʕariṣu
(Ibn Gabirol 1971: 140)
A poem like the one in Excerpt 3 cannot be translated even literally; only its
semantic structure can be explicated. It is a sequel to Excerpt 2. It begins
with the last word of the earlier poem, קּופים
ִ ּת,ְ that is, “the firm ones” (= the
children of Israel). This is the subject phrase for all the ensuing predicates. In
the rest of the poem, each verse line consists of three consecutive adverbials
of mode or instrument that suggest “talking” or “devotion” of one kind or
another. The last two words of each line contain a direct object followed by a
verb. Some of the words do not exist in the dictionary, and those that do are
used here in a slightly or radically different sense. The nonexistent words can
be shown to be derived from an existing word of the same root. The adverbi-
als have one semantic ingredient in common: “talking” or “adoration.” The
semantic features that distinguish them from one another are effaced.
Let us take a closer look at the first three lines of this poem:
The first three words of each line begin with the prefix bǝ-, whose func-
tion is to turn a noun into an adverbial of mode or instrument. Bǝšenen
is derived from šinnun (repetition, learning by rote); Yarden in his com-
mentary construes it as “speech.” Bǝšɛqɛd is derived from šǝqida, meaning
“industriousness.” Bǝšɛvaḥ means “in praise.” Širot jašbiṣu means “inlay
songs [with words].” Bǝlahav means “with flame, ardently.” Bǝlɛmɛd is
derived from lǝmida (learning). Lɛqaḥ means “moral, lesson.” Lǝvavot jaliṣu
means “cheer up hearts.” Bǝmaʕan is derived, according to Yarden, from
maʕane lašon (response, reply). Bǝmɛlɛṣ is derived from mǝliṣa, “flowery
speech.” Bǝmɛrɛṣ means “vigorously.” Jamriṣu, derived from the same root,
means “urge, encourage, stimulate.” Thus, milot jamriṣu means “stimulate
words.” And so forth.
The graphemic-phonemic dimension of the poem is remarkable.
Ignoring the repeated prefix, in every line each word begins with the same
consonant (alliteration). The letters designating the alliterating speech
sounds yield, in the consecutive lines, the acrostic ( שלמה צעירthat is, “young
Šǝlomo”). We use the cumbersome phrase “the letters designating the allit-
erating speech sounds” because the letter ה, for instance, designates the
consonant [h]in the alliterating sequence but serves as a special character
designating a vowel in the acrostic “Šǝlomo.” Such a shift emphasizes that
the same graphemes change their very nature from text to acrostic.
Many of the consecutive adverbials are near-homonyms (that is, they
differ in only one phoneme), such as bǝmɛlɛṣ bǝmɛrɛṣ, bǝṣɛvaḥ bǝṣɛraḥ, bǝʕɛrɛṣ
bǝʕɛlɛṣ, bǝjošɛr bǝjotɛr bǝjoqɛr, and bǝraḥaš bǝraʕaš. Note that bǝṣɛvaḥ bǝṣɛraḥ
became stock epithets and occur in Excerpt 3 as well. All nine verse lines
rhyme on the same monorhyme ending with -iṣu, with the stress on the
penultimate vowel.
Excerpt 3 is similar to Excerpt 1 in that it too contains obsessive repeti-
tions of words suggesting song and adoration. It presents a rich catalog of
words, most of which contain the semantic component “talking” or “voic-
ing.” These words display an enormous variety of meanings. However, here
too the words change according to the requirements of the acrostic and
occur in reiterated syntactic patterns. They reinforce their shared semantic
components and lose their distinguishing features. Thus they contribute to
the effect of monotonous repetition, with no novel information, and ulti-
mately give the impression of artificiality and technical trifling. This prin-
ciple is so effective here that readers “understand” even the words whose
meaning they do not know and construe them as “some kind of saying” or
“some kind of adoration.”
As we have pointed out above, the letters/phonemes in this text are
treated in a special way (which is quite characteristic of this corpus). Apart
A r t i s t i c De v i c e s a n d M y s t i c a l Q ua l i t i e s [ 113 ]
4
1
from their function in ordinary language, they are organized into three
additional patterns, all arbitrary from the point of view of ordinary com-
munication. First, in every line each word begins with the same letter (after
omitting the prepositional and conjugational prefixes). Second, these let-
ters add up to the poet’s “signature,” his name and nickname, שלמה צעיר
(young Šǝlomo). Third, the last words of the lines make up an apparently
“innocent” monorhyme. However, this monorhyme is part of a fourth pat-
tern: When the last line appears, the monorhyme takes on new significance
after the event; it heralds, as it were, the biblical verse “They will hallow the
Holy One of Jacob, / And fear the God of Israel” (Isaiah 29:23). The occur-
rence of this biblical verse is expected and surprising at the same time and
ends the poem with a powerful closure. Likewise, Excerpt 4 (below) termi-
nates with the verse “ֹלהינוּ ַה ַּמגְ ִּב ִיהי לָ ֶשׁ ֶבת
ֵ ”מי ּכַ יְ יָ ֱא
ִ [who is like the Lord our God
who dwells on high], quoted from Psalms 113:5. Such a closure is a conven-
tion in this corpus, and as such, it is expected. But readers do not know
which verse they should expect or at what point of the sequence. Only after
the event, when the verse occurs, do they feel that the sequence has ended,
that they have “reached home.”
The signature of the poet’s name is widespread in liturgical poetry. It
is usually regarded as a device to indicate the poet’s identity and nothing
more. Alphabetic acrostics are found in the Bible. In this poem, however,
the fact that the letters occur in three different patterns beyond their
ordinary linguistic use foregrounds most emphatically the arbitrary, non-
linguistic combination of letters. A comparison to Psalm 34 highlights
the unique treatment of letters in this poem. In ordinary language, as
well as in poetic language, the letters are “transparent” signs: we do not
linger on them, since their function is to point to the phonological sig-
nifieds (speech sounds), which, in turn, serve as signifiers of semantic
units. Though in Psalm 34 the syntactic coherence is less smooth than
in other psalms, the semantic-thematic richness of the verses is not
impaired. Conversely, the blurring of the semantic load of the words in
this poem and the increase in the number of patterns in which the letters
of the rhyme and the acrostic participate result in the disruption of the
chain of signifiers and signifieds, thus directing the reader’s attention to
the combination of letters.8
In light of these differences, the “innocent” signature of the author’s
name takes on special significance in the context of the processes discussed
8. It should be noted that when the early mystics talk of “letters,” it is not always clear
whether they mean the written signs or the speech sounds; but this does not affect our
essential argument.
the letters of the Hebrew alphabet, out of which words are composed, are the
fundamental building blocks of creation. The letters have ontic capacity and
can be—indeed, have been—employed by God to create the world and every-
thing within it. In this reading, the creation accounts wherein “God speaks” are
taken with extreme, if original, literalness. The process of creation resides in the
manipulation of the alphabetical ciphers. (Katz 1992: 16)11
9. Book of Yetsira is the title of the earliest extant book on Jewish esotericism.
10. Theurgy is the art or technique of compelling or persuading a god or beneficent or
supernatural power to do or refrain from doing something.
11. “Writing has a very magical quality—not because of anything divine about
its origins, but because it greatly increased our brain’s capacities. It is close to
miraculous that Homo sapiens, a mere primate, was able to dramatically increase
its memory by making a few marks on paper,” says the neuropsychologist Stanislas
Dehaene (2009: 172). But notice this. When mystics believe in such magic, it is one
thing; when critics insist on it, it is another. A lot of nonsense has been written,
for instance, Kabbalistic speculations on symbolic meanings in Coleridge’s “Alph the
sacred river” (e.g., Beer 1959: 209–212; Tsur, 2006c: 25–26).
A r t i s t i c De v i c e s a n d M y s t i c a l Q ua l i t i e s [ 115 ]
6
1
graphemic component. Katz quotes a short passage from Sefer Yetzira that
describes such “meaningless” manipulation of letters:
How did He combine them, weigh them and set them at opposites?
Aleph with all of them, and all of them with Aleph,
Bet with all of them, and all of them with Bet.
It rotates in turn, and thus they are in two hundred and thirty-one gates.
And everything that is formed and everything that is spoken goes
out from one term. (in Katz 1992: 17)
Active types of meditation begin not with the intention to clear the mind of
thoughts, but instead, to focus it intensely upon some thought or object of
attention. A Buddhist might chant a mantra, or focus upon a glowing candle or
a small bowl of water, for example, while a Christian might pray with the mind
trained upon God, or a saint, or the symbol of a cross. (2001: 120)
Similarly, a Jew may focus on the heavenly halls or the combination of the
letters of God’s names.
An anonymous disciple of Abulafia’s gave an extended, masterly descrip-
tion of such an experience; a short passage is reproduced below:
When I came to the night in which this power was conferred on me, and
midnight—when this power especially expands and gains strength whereas the
body weakens—had passed, I set out to take up the Great Name of God, consisting
of seventy-two names, permuting and combining it. But when I had done this for
a little while, behold, the letters took on in my eyes the shape of great mountains,
strong trembling seized me and I could summon no strength, my hair stood on end,
and it was as if I were not in this world. At once I fell down, for I no longer felt the least
strength in any of my limbs. And behold, something resembling speech emerged from
my heart and came to my lips and forced them to move. (in Scholem 1961: 150–151)12
12. We quote Abulafia’s disciple not as someone who influenced Ibn Gabirol (who
lived 150 years before him) but as someone who gave an exceptionally lively descrip-
tion of the experience whose fossilized remnants we encounter in our corpus.
A r t i s t i c De v i c e s a n d M y s t i c a l Q ua l i t i e s [ 117 ]
8
1
Excerpt 4.
א ִּדיר ַּב ָּמרוֹםנִ ּצֶ ֶב
ת ַ קר ְתָך
ְָ ִּת
יוֹשׁ ֶבתֶ נִ ּצֶ ֶבת ֶמ ְמ ַשׁלְ ּתְָךוְ לָ נֶ צַ ח
שׁלְ ֶה ֶבת ַ יוֹשׁ ֶבת ִס ְת ָרתְָך ַּב ֲח ֻתּלַ ת
ֶ
וּמ ֹּפהלַ ֶה ֶבת ִ ַשׁלְ ֶה ֶבת ִמּפֹה
מ ְרּכֶ ֶבתִ ת לַ ֶה ֶבת ְתלַ ֵהט לְ עֻ ַּמ
ִמ ְרּכֶ ֶבת עֻ ּזְָך ִהיא ַהּנִ ּצֶ ֶבת
יוֹשׁב ַּב ֶּשׁ ֶבת
ֵ ַהּנִ ּצֶ ֶבת מוּל ִט ְפ ָסר
ַּב ֶּשׁ ֶבת ִמּיָ ִמין ְשׁ ָח ִקי םרוֹכֶ ֶבת
רוֹכֶ ֶבת ֲא ֻפ ַּדת ְשׁ ִביעִ י ְּברֹאשׁוֹ ְּב ַמּצֶ ֶבת
יוֹשׁ ֶבתֶ ְּב ַמּצֶ ֶבת ִשׁ ְבעָ ה ְּב ַפ ְרּגוֹ ד
13. Ibn Gabirol’s, but not Ibn Ghiyat’s, poem has the extra formal constraint: an elab-
orate acrostic. If you combine the first letters of each line, you obtain “אני שלמה ברבי יהודה
אבן גבירול,” that is, “I Šǝlomo son of Rabbi Yǝhuda Ibn Gabirol.”
יוֹב ֶבת
ֶ יוֹשׁ ֶבת ִמּלְ ָפנִ ים נִ ְשׁ ְק ָפה ֶ
יוֹב ֶבת מוּל ִמ ְרּכַ ְב ְּתָך ַהּנִ ְשׁלֶ ֶבת ֶ
רוּח וְ עָ נָ ןוְ צָ ֶר ֶבת
ַ ַהּנִ ְשׁלֶ ֶבת ְּב
ּדוֹא ֶבת
ֶ וְ צָ ֶר ֶבת ּתוְֹך ֶק ַרח ֵמ ַאיִ ן
רוֹפ ֶפת עֲ לִ ּיָ ה ַהּנֶ ְחצֶ ֶבתֶ ְּדוֹא ֶבת ו ֶ
אוֹה ֶבת
ֶ ַהּנֶ ְחצֶ ֶבת לְ קוֹל ֲה ֻמּלָ ה אוֹתְָך
אוֹה ֶבת עֻ ּזְָךּגַ ם ְּבקוֹל ּגַ ם ְּב ַמ ֲח ֶשׁ ֶבת ֶ
נוֹב ֶבת
ֶ ְּב ַמ ֲח ֶשׁ ֶבת ְוּבקוֹל ָרּנֵ י זְ ִמירוֹת
ּגוֹב ֶבת
ֶ נוֹב ֶבת ִרגְ ַשׁת ָה ַא ְר ַּבע ָס ִביב ֶ
לשׁים וָ ֵשׁשׁ נֶ גְ ּדָך ְּב ִמ ְקצֶ ֶבת ִ ּגוֹב ֶבת ְשׁ ֶ
יוֹה ֶבת
ֶ וּשׂמֹאל לְָך ֹעז ְ ְּב ִמ ְקצֶ ֶבת יָ ִמין
יוֹה ֶבת ֶק ֶדם וְ ָאחוֹר ְּב ַאלְ ֵפי ְר ָבבוֹתרוֹכֶ ֶבת ֶ
סוֹב ֶבת ֶ ְקוֹמָךהוֹלֶ כֶ תו ְ רוֹכֶ ֶבת עַ ל ְמ
סוֹב ֶבת לִ ְראוֹת ְּבזִ יו ַה ִמ ְרּכֶ ֶבת ֶ ְו
ֹלהינוּ ַה ַּמגְ ִּב ִיהי לָ ֶשׁ ֶבת ֵ ִמי ּכַ יְ יָ ֱא
tikratkha ʔadir bamarom niṣɛvɛt
niṣɛvɛt mɛmšaltǝkha wǝlanɛṣaḥ jošɛvɛt
jošɛvɛt sitratkha baḥatulat šalhɛvɛt
šalhɛvɛt mipo umipo lahɛvɛt
lahɛvɛt tǝlahet lǝʕumat mirkɛvɛt
mirkɛvɛt ʕuzkha hiʔ haniṣɛvɛt
haniṣɛvɛt mul tifsar jošev bašɛvɛt
bašɛvɛt mijamin šǝḥaqim rokhɛvɛt
rokhɛvɛt ʔafudat šǝviʕi bǝroʔšo bǝmaṣɛvɛt
bǝmaṣɛvɛt šivʕa bǝfargod joṣɛvɛt
joṣɛvɛt milfanim niṣqǝfa jovɛvɛt
jovɛvɛt mul mirkavtǝkha hanišlɛvɛt
hanišlɛvɛt bǝruaḥ bǝʕanan wǝṣarɛvɛt
wǝṣarɛvɛt tokh qɛraḥ meʔajin doʔɛvɛt
doʔɛvɛt wǝrofɛfɛt ʕalija hanɛḥṣɛvɛt
hanɛḥṣɛvɛt lǝqol hamula ʔotkha ʔohɛvɛt
ʔ
ohɛvɛt ʕuzkha gam bǝqol gam bǝmaḥṣɛvɛt
bǝmaḥṣɛvɛt uvǝqol rane zǝmira novɛvɛt
novɛvɛt rigṣat haʔarbaʕ saviv govɛvɛt
govɛvɛt šǝlošim wašeš nɛgdǝkha bǝmiqṣɛvɛt
bǝmiqṣɛvɛt jamin usmol lǝkha ʕoz johɛvɛt
johɛvɛt qɛdɛm wǝʔaḥor bǝʔalfe rǝvavot rokhɛvɛt
rokhɛvɛt ʕal mǝqomkha holɛkhɛt vǝsovɛvɛt
vǝsovɛvɛt lirʔot bǝziw hamirkɛvɛt
mi kaʔadonaj ʔelohenu hamagbihi laṣɛvɛt. (1971: 244–246)
A r t i s t i c De v i c e s a n d M y s t i c a l Q ua l i t i e s [ 119 ]
201
Anadiplosis foregrounds both the initial and final boundaries of the verse
lines. Thus it may generate an effect of blocking or disjoining the lines.
On the other hand, repetition of the same word at the end and begin-
ning of two consecutive lines can have the effect of conjoining them, blur-
ring the distinction between the separate lines, and blending them into
one semantic sequence. Anadiplosis thus has two potentially opposite
effects (as the great modernist Hebrew poet Yehuda Amichai wrote of
the sea: “perhaps connects, perhaps separates”). We argue that one of
the texts realizes the former and the other, the latter option. We label the
former option “the blocking effect” and the latter “the flux effect.” The
flux effect connects the linguistic units in the consecutive verse lines and
can also be labeled “the mystic effect” because it blurs the boundaries and
differentiation of the various verse lines, generating weak, ambiguous,
indistinct verbal structures. In Ibn Ghiyat’s poem (see below) anadiplosis
Let us consider the first verse line: “”ּת ְק ָרתְ ָך ַא ִּדיר ַּב ָּמרוֹם נִ ּצֶ ֶבת
ִ [Your ceiling,
Mighty, in the sky [is] positioned]. In Hebrew, both the adverbial “in the
sky” and the adjective positioned may serve as predicates without the
copula is. Consequently, in a first reading the adverbial may be perceived,
momentarily, as the predicate completing the utterance: “Your ceiling,
Mighty, [is] in the sky.” In such a reading, the concatenating word posi-
tioned will be perceived as unnecessary for the utterance of the line and
freed to link with the utterance in the next line, which, in turn, begins
with the same word. In this reading, readers can treat the two tokens of
A r t i s t i c De v i c e s a n d M y s t i c a l Q ua l i t i e s [ 121 ]
21
positioned as reduplication and thus join the two lines into one seman-
tic sequence. However, while this is a plausible reading, there is a more
obvious organization of the syntactic-semantic information in this verse
line: when readers reach the line-final positioned, they reorganize the ver-
bal elements to suggest “Your ceiling, Mighty, [is] positioned in the sky.”
In such an organization of the verbal information in the line, “[is] posi-
tioned” becomes the nucleus of the predicate and is no longer perceived as
superfluous for the utterance in the line. At this point it becomes clear that
the two performances—the one that assigns a central role to positioned in
the predicate and the one that relegates it to the next line—generate two
sentences that are near-synonymous: “Your ceiling, Mighty, in the sky [is]
positioned” = “Your ceiling, Mighty, [is] in the sky.” What William Empson
(1955) called “double syntax” conveys here no additional meaning. So, what
is their effect? They blur each other, enhancing the fluidity of the passage,
and thus contribute to the mystical effect of the sequence.
A more striking example of the effect of double syntax generated by the
inversion of the adverbial of place and the concatenating word is found
in line 4: “ּומּפֹה לַ ֶה ֶבת
ִ ”שׁלְ ֶה ֶבת ִמּפֹה
ַ [flame on this side and on this side blaze].
Here as well there are two possible readings: one that sees the concatenat-
ing word blaze as redundant in its line, thus freeing it to link with the
next line, and another that regards it as required by the preceding utter-
ance. Idiomatic Hebrew would require saying “flame on this side and on
this side” for “flame on each side” or “flame all around.” The present-day
Hebrew speaker has difficulty reading this sequence of words otherwise,
even after becoming overfamiliar with this verse line. But readers then
realize that the last word enables them to reorganize the line into two par-
allel sentences, that is, into a symmetrical closed shape. Again, the double
syntax conveys no added meaning; rather, the two possible sentences blur
each other. Thus, the sequence becomes well articulated and fluid at the
same time. The conjunction and changes its function from one reading to
another. In the former it serves to conjoin two parallel adverbials of place
in an idiomatic expression, whereas in the latter it conjoins two sentences,
thus enhancing the sense of uncertainty, instability, and insecurity.
Let us briefly present some additional examples of double syntax gen-
erated by such verb-adverbial inversion at the concatenation junction.
In line 19 we find: “ּגוֹב ֶבת ֶ ”נוֹב ֶבת ִרגְ ַשׁת ָה ַא ְר ַּבע ָס ִביב
ֶ [talks the excitement of
the four around crowding]. Here again “talks the excitement of the four
around” can be read as a self-contained, complete, though slightly obscure
sentence (and one that is utterly awkward in English): it has a straightfor-
ward verbal predicate (talks), a subject (the excitement of the four), and an
adverbial of place (around), meaning “the excitement of the four is talking
all around” (in Hebrew, transposition of the verb before the noun is per-
fectly legitimate). In this way, the concatenating word, crowding, is made
redundant and linked to the utterance in the next line. In this reading,
the syntactic-semantic status of the line-final concatenating word becomes
obscure; by coupling with the word at the beginning of the next line, it is
perceived in the sweep of reading as an irrational, gratuitous repetition,
reinforcing the emotional import displayed by the already obscured pat-
tern of anadiplosis. On second reading, however, the same constituents
of the sentence allow for an alternative organization that renders the
line-final concatenating word highly required by the preceding phrases in
the line: “talks the excitement of the four—all around crowding together.”
Likewise, line 22 may be read as “”יוֹה ֶבת ֶק ֶדם וְ ָאחוֹר ְּב ַאלְ ֵפי ְר ָבבוֹת—רוֹכֶ ֶבת
ֶ [give
before and behind by thousands of thousands—it chariots]14 or “יוֹה ֶבת ֶק ֶדם ֶ
[ ”וְ ָאחוֹר— ְּב ַאלְ ֵפי ְר ָבבוֹת רוֹכֶ ֶבתgive before and behind—by thousands of thou-
sands it chariots].
יוֹה ֶב
ת ֶ וּשׂמֹאל לְָךעֹז ְ ְּב ִמ ְקצֶ ֶבת יָ ִמין
יוֹה ֶבת ֶק ֶדם וְ ָאחוֹר ְּב ַאלְ ֵפי ְר ָבבוֹת רוֹכֶ ֶבת
ֶ
[rhythmically right and left honor to you give
give before and behind by thousands of thousands it chariots]
There are verbs that cannot stand alone and inherently require comple-
tion. “He closed” cannot be considered a complete utterance unless we can
reconstruct the missing object from the context. “He closed the door,” by
contrast, is a complete utterance. The objects are indispensable for, inher-
ently required by, such verbs. If we say, “He closed the door yesterday,”
the adverbial of time is not inherently required but an optional addition,
not mandatory. Ibn Gabirol proceeds in these lines as follows: At first
(line 21), he presents the verb gives with its two required object phrases
14. This is one of the typical obscurities of this poem. It’s a verb we took from Shelley’s
“Ode to the West Wind” (“Oh thou, / Who chariotest”).
A r t i s t i c De v i c e s a n d M y s t i c a l Q ua l i t i e s [ 123 ]
2
41
(“[give] honor to you”); the second time (line 22), he does not provide
the internally required complements but adduces, as an afterthought, an
optional one (the adverbial “before and behind”) that becomes natural—if
at all—only after the internally required objects have already been sup-
plied. In this way, in order to perceive a complete utterance, the reader may
tend to accept the reduplication of gives as suggesting an afterthought, so
as to perceive the two verse lines as complementing each other to consti-
tute one meaning unit: “rhythmically right and left—honor to you give /
(give) before and behind.” The line boundaries are obscured and ruptured,
and the verbal units (verse lines) extend into each other such that the flux
effect is generated.
A similar phenomenon occurs in lines 12–13:
The third way Ibn Gabirol renders the boundary words ambiguous
is to attach the relative particle to the concatenating word. Consider
lines 15–16:
How should this verbal sequence be realized? Again there are two possibili
ties. One possibility is to foreground the verse lines as closed, complete units,
reading as follows: “אוֹה ֶבת
ֶ ַהּנֶ ְחצֶ ֶבת לְ קוֹל ֲה ֻמּלָ ה—אוֹתְָך/ רוֹפ ֶפת עֲ לִ ּיָ ה ַהּנֶ ְחצֶ ֶבת
ֶ ְ”ּדוֹא ֶבת ו
ֶ
[painful and unstable [is] the heaven that is carved / that is carved to the
voice of tumult—that loves you]. Line 15 means, roughly, that heaven
was carved from an unstable state and the process was painful. (“Carved”
inherently requires some hard material as its object, suggesting here a
paradox.) Line 16 means that the carving of heaven was accompanied by
A r t i s t i c De v i c e s a n d M y s t i c a l Q ua l i t i e s [ 125 ]
261
the loving voices of a tumult (of angels?). With some goodwill we can say
that line 15 is complete and self-sufficient, closed, and that line 16 is a
separate unit providing some additional information that the painful carv-
ing of heaven was accompanied by some loving voices. This stable, closed
structure is reinforced by a symmetric chiastic pattern: line 15 ends with
the relative clause preceded by all the rest, whereas line 16 begins with a
relative clause, followed by all the rest. One cannot ignore, however, that
the relative clause in line 16 refers back to the subject (“heaven”) in line 15.
Hence neither of the two verse lines “begins or ends, but . . . extend into
each other” (Ehrenzweig 1965: 35). The two tokens of “that is carved” are
perceived as reduplication for emotive purposes, or the sake of obscuring
the juncture between the two lines, or both. The two lines will be read as
“painful and unstable are the heavens that are carved (that are carved) to
the voice of tumult / that loves you.”
ANADIPLOSIS IN IBN GHIYAT
Excerpt 5.
מוֹשׁ ָביו
ָ ַא ָּתה ֵאל ִמ ְס ַּת ֵּתר ְּב ֶח ְביוֹן
. . .
עוֹב ָריו וְ ָשׁ ָביו
ְ ְנוֹדעוּ ל ְ בוֹתימוֹ ל ֹא ֵ נְ ִת ָיביו עִ ְק
טוֹביו
ָ ְנוֹה ִרים ל ֲ ְּפוֹח ִדים ֵאלָ יו ּכַ ּיוֹם וֲ וְ ָשׁ ָביו
רוֹביו
ָ טוֹביו צָ ֲהלוּ ֵמהוֹדוֹ ְּבשׂוּמוֹ ְרחוֹקוֹ ִּב ְק ָ ְל
דוֹשׁיו נְ ִד ָיביו ָ דוֹשׁיו ְוּק
ָ רוֹביו ְק ָ ְק
נְ ִד ָיביו ְרצוֹנוֹ לְ ָה ִפיק ְּב ָק ָריו וַ ֲע ָר ָביו
אוֹריו ְּב ַמ ֲע ָר ָביו ָ וַ עֲ ָר ָביו ָשׁ ֲחרוּ ּכִ ְשׁקֹע
ʔ
ata ʔel mistater bǝḥɛvjon mošavav
. . .
nǝtivav ʕiqvotemo loʔ nodʕu lǝʕovrav vǝšavav
vǝšavav poḥadim ʔelav kajom vǝnoharim lǝtovav
lǝtovav ṣahalu mehodo bǝsumo rǝḥoqo biqǝrovav
qǝrovav qǝdošav uqǝdošav nǝdivav
nǝdivav rǝṣono lǝhafiq bǝqarav vaʕaravav
vaʕaravav kǝšaḥar kišqoʕa ʔorav bǝmaʕaravav (in David 1987)
[you are a God who hides in the secret place of his residence
. . .
the footsteps of his routes are unknown to his passers and his returners
[= passersby]
and his returners dread to him today and throng to his goods
to his goods they rejoice in his majesty turning remote ones into those
close to him
those close to him are his saints his saints are his nobles
his nobles attempt to please him in his mornings and his evenings
[= always]
and his evenings are black while his sun is setting in his West]
Ibn Ghiyat follows verbal strategies that are very different from Ibn
Gabirol’s. Owing to the syntactic function of the majority of the line-final
concatenating words in Ibn Ghiyat’s poem, they are highly required in
order to complete the line as a syntactic-semantic unit. Most of the con-
catenated verse lines in this poem close with words that, because of their
syntactic-semantic function, must belong to the line in which they occur,
either because they are the prepositional phrase of a genitive construct, as
in “the secret place of his residence”; or because they are nouns preceded
by the conjunction and, creating a pair of parallel nouns in a contracted
clause, as in “in his mornings and his evenings” (their symmetry is rein-
forced by the facts that they are antonyms and that the expression is an
idiomatic stock phrase in medieval devotional poetry, meaning “always”);
or because they are nouns that are inherently required by the verb, as in
“turning remote ones into those close to him” (the verb turn into inher-
ently requires a complement; the same holds true of throng to, requiring an
adverbial of direction). In some cases the second token of the concatenat-
ing word is used with a radically different meaning, indicating a completely
new start. The collocation “his mornings and his evenings” is an interesting
case. The conjunction and conjoins an idiomatic pair of parallel (antonymic)
nouns. In the next token of “and his evenings” it conjoins two independent
sentences, dismantling the idiom, where the repeated word refers to an
entirely different aspect of “evening” and starts a statement that is con-
spicuously irrelevant to the preceding utterance; moreover, quite untypi-
cally in this corpus, it gives a description of nature: black evening as the
background to the setting sun. The same holds true of “the footsteps of his
routes are unknown to his passers and his returners,” where and conjoins
an idiomatic pair of parallel (near-synonymic) nouns meaning his (casual)
“passers-by.” The next token of “returners” suggests, literally, return to a
A r t i s t i c De v i c e s a n d M y s t i c a l Q ua l i t i e s [ 127 ]
2
8
1
place (from exile, as in the book of Jeremiah?) or, possibly, religious repen-
tance; both construals are conspicuously irrelevant to the “passers-by” of
the preceding utterance.
These three syntactic functions and the semantic change in the last
member of an idiomatic phrase determine the status of the line-final con-
catenating word as the last link demanded to complete the utterance and
close the verse line. This emphasizes the “blocking effect” of anadiplosis,
whose task is to foreground a clear-cut boundary between verse lines.
We are confronted with two different aesthetic conceptions. The “block-
ing” and “flux” effects are associated with different semantic conceptions
of concatenated words. Two meanings of the word ambiguous highlight
these different conceptions. Merriam-Webster’s Collegiate Dictionary defines
ambiguous as “doubtful or uncertain especially from obscurity or indistinct-
ness” and “capable of being understood in two or more possible senses.”
The former characterizes Ibn Gabirol’s usage, and the latter, Ibn Ghiyat’s.
Ibn Gabirol typically obscures the difference between the two tokens of the
concatenated words (as we have pointed out with reference to cases of “dou-
ble syntax”), whereas Ibn Ghiyat typically sharpens it (as in “in his morn-
ings and his evenings [= always] / and his evenings are black”). Versification
units with clear-cut boundaries have “good gestalt” and suggest strong
intellectual control, inspiring certainty, whereas units with blurred bound-
aries have “weak gestalt,” inspire uncertainty, are regarded as affectively
charged, and indicate emotional excitability. This view of Gestalt theory is
supported by the Rorschach inkblot test. Furthermore, the abrupt change
of meanings in the repeated words introduces considerable movement into
the frozen formula in Ibn Ghiyat’s poem as well.
SUMMARY
A r t i s t i c De v i c e s a n d M y s t i c a l Q ua l i t i e s [ 129 ]
301
1
3
CHAPTER 6
logics. People usually do this with amazing ease. But, en route, something
happens to them, as the following joke confirms. A mental asylum inmate
is pulling a shoe box on a long string and says: “Come on, come on, come
on!” A doctor who is passing by says, “What a lovely dog you have there,
what is its name?” “Are you crazy, doctor, don’t you see this is merely a
shoe box?” “Oh yes, now I realize, it is a shoe box. Why are you pulling
it on such a long string?” “Because I am afraid it will bite me.” This joke
involves several reversals of mental sets and a readiness to enter, one after
the other, into several situations, each of which has an internal logic that
differs from the preceding one. In each such reversal, the reader receives a
shock but is able to readjust immediately to the new situation. The shock
and the ensuing readjustment are perceived as wit. Although in real life
attention is directed toward the shifting situation appraisals that affect
one’s well-being, in the joke the audience’s attention is directed away from
them to their phenomenological quality. This is what I mean throughout
this book by the statement that “in the response to literature, cognitive
devices evolved for adaptive purposes are turned to aesthetic ends.”
Metaphysical conceit deals with grave issues in a witty mode, such as
the pangs of love or the existential problems of facing Divinity. When
experiencing the processes involved, readers may feel reassured that their
adaptation mechanisms function properly and derive pleasure from this
reassurance. It should be noted that in Excerpts 3 and 4 below the conceits
develop, simultaneously, in two different situations, one in the physical
and the other in the spiritual domain, following radically different logics.
Nevertheless, the two situations are closely related, owing to the fact that
they consistently develop two respective aspects of one image. Thus the
mechanism underlying thought, the image, becomes the center of aware-
ness. A similar story, with minor changes, fits Excerpts 7–12 (below).
This approach sees language as a hierarchical system of signifiers
and signifieds: from the point of view of reading, for instance, a system
of graphemes (letters) on paper signifies a system of phonemes (speech
sounds). Combinations of phonemes designate semantic representation
units; semantic representation units, in turn, are used to refer to entities in
extralinguistic reality. Individuals, as sign-using animals, are programmed
to get as quickly as possible from the signifiers (which have little survival
value) to the signifieds (which may have high survival value). When we read
the instructions for using some piece of electrical equipment, we remember
the instructions rather than the type font of the letters. After some time we
may not even remember what language the instructions were in. According
to Roman Jakobson (1960), poetic language typically directs attention back
to the various signifiers and forces us to linger at them. Meter and rhyme
As true Wit generally consists in this Resemblance and Congruity of Ideas, false Wit
chiefly consists in the Resemblance and Congruity of single Letters, as in Anagrams,
Chronograms, Lipograms, and Acrosticks: Sometimes of Syllable, as in Ecchos and
Doggerel Rhymes: Sometimes of Words, as in Punns and Quibbles; and sometimes
of whole Sentences or Poems, cast into Figures of Eggs, Axes or Altars. . . . As true
Wit consists in the Resemblance of Ideas, and false Wit in the Resemblance of
Words, according to the foregoing Instances; there is another kind of Wit which con-
sists Partly in the resemblance of Ideas, and partly in the Resemblance of Words;
which for Distinction Sake I shall call mixt Wit. This Kind of Wit abounds in Cowley,
more than in any Author that ever wrote. Mixt Wit is therefore a Composition of
Punn and true Wit, and is more or less perfect as the Resemblance lies in the Ideas
or in the Words. (1951 [1711–1712]: The Spectator, no. 62)
The Acrostick was probably invented about the same Time with the Anagram,
tho’ it is impossible to decide whether the Inventor of the one or the other were
the greater Blockhead. (1951 [1711–1712]: The Spectator, no. 60)
F i g u r at i v e L a n g ua g e a n d S o c i o c u lt u r a l B ac k g r o u n d [ 133 ]
341
Excerpt 1.
F i g u r at i v e L a n g ua g e a n d S o c i o c u lt u r a l B ac k g r o u n d [ 135 ]
361
Excerpt 2.
בֹורה ָ עֲ ֵׂשה לָ ֶהם ּכְ ִס ְיס ָרא וַ עֲ ֵׂשה לִ י ּכְ ַמ ְע ֶׂשיָך לְ ָב ָרק ְּוד
ֹאבדּו ִמּגְ ָע ָרה
ְ ֹלהי וַ ֲה ִפיצֵ ם ּגְ ַער ָּב ֶהם וְ י ַ ְּברֹק ָּב ָרק ֱא
ʕ
ase lahεm kǝsisraʔ vaʕase li kǝmaʕsεkha lǝvaraq udǝvora
bǝroq baraq εlohay vahafiṣem gǝʕar bahem vǝyʔovdu migǝʕara
ʔ
in it at all, only the patronymic of the person called “Baraq.” However, the
ear trained on parallelisms and medieval inlay language will find the paral-
lelism of “the thunder fell” and the explicit “the son of Avinoam came” less
than satisfactory. Indeed, we are confronted with a school example of what
in medieval Hebrew poetics is called a “hidden pun.” Without an aware-
ness that the unmentioned word Baraq has two unrelated meanings, the
verse line cannot be understood. So, readers are compelled to suspend their
“communicative competence” and exit the chain of signifiers and signifieds.
They must pass sidewise from the explicit patronymic “the son of Avinoam”
to the unmentioned part of the proper name, “Baraq,” and become aware,
first, that this name is the signified of the phonological sequence [baraq];
second, that the phonological sequence has another semantic signified as
well, unrelated to the previous signified, meaning “lightning”; and, third,
that this semantic signified has a good fit with a nature description. In
Excerpt 2 readers may comprehend the meaning of the two verse lines even
in translation into a language in which the similarity of sounds does not
exist. In the Hebrew original, the phonological sequence will reverberate
in the back of readers’ minds as musicality: At most they will regard it as
paronomasia (a play on words, pun); at worst, they will miss the sound
patterns altogether but will get the message. In Excerpt 1, by contrast,
one may not attend away from the phonological sequence [baraq] and its
two sign functions: if one attends away from them, the entire verse line
becomes meaningless. Furthermore, one cannot give rules as to at which
point one must exit the ordinary semiotic chain of signifiers and signifieds
to construct a meaningful utterance. Consequently, the expression is more
similar to a riddle than to a metaphor. The description is received as wit or,
rather, as Mannerist witticism.
So far I have made two crucial distinctions. One concerns poetic and
nonpoetic language. Both are based on a hierarchy of signs. In the latter,
however, readers or listeners attempt to reach the final referent as fast
as possible by attending away from the signifiers. In the former, they are
forced, by various poetic devices, to attend back to the signifiers. The sec-
ond distinction concerns types of poetic style. In Excerpt 1 (the Mannerist
style type), the signifiers and poetic devices are more obtrusive than in
Excerpt 2. The difference can be described in structural terms or in terms
of the duration of delay. Assuming that in both style types the reader or
listener must linger longer on the chain of signifiers than in nonpoetic
language, in Excerpt 2 the reader may, eventually, attend away from the
verbal expressions to an abstraction of human significance. The concrete
images are perceived as illustrating some idea. John Crowe Ransom (1951)
dubbed such poetry “Platonic poetry.” In Excerpt 1, the structure of the text
F i g u r at i v e L a n g ua g e a n d S o c i o c u lt u r a l B ac k g r o u n d [ 137 ]
381
prevents the reader from attending away from the concrete verbal images
to an abstraction. This is Mannerism of the Précieux type. “Romantic” and
“classic” styles are usually regarded as opposites. But in this respect, in con-
trast to Mannerism, they are similar in that both are typically “Platonic.”
In the two preceding excerpts, the text forces the reader to attend back
to the phonological signifier to some extent. In the following two excerpts,
both of the “Metaphysical” kind, the reader is forced to attend back to the
semantic component:
Excerpt 3.
י ּד ְמ ִעי וְ ִּת ְׁש ָט ֵחם ְּב ֶׁש ֶמׁש זָ ֳה ָרּה
ִ עָ ְפ ָרה ְּתכַ ֵּבס ְּבגָ ֶד ָיה ְּב ֵמ
ʕ
Ofra tǝkhabbes ʔεt bǝgadεha bǝmey dimʕi vǝtištaḥem bǝšεmεš zohora
Ofra [the Gazelle] washes her clothes in the waters of my tears
and spreads them against the sun of her splendor (Yehuda Halevy;
my translation)
Excerpt 4.
F i g u r at i v e L a n g ua g e a n d S o c i o c u lt u r a l B ac k g r o u n d [ 139 ]
4
01
Excerpt 5.
זָ נ ַֹח אתיָך ֲא ִבי
ִ ֹד עד ּכִ י ְק ָר
ַ ָאכֵ ן ְמיֻ ָּדעִ י זְ נַ ְח ַּתנִ י ְמא
נֹוח
ַ ׁשֹוטט ָּב ֱאנֹוׁש ל ֹא ָמצְ ָאה ּכִ י ִאם ְּבָך ָמ ֵ ְיֹום ֻׁשּלְ ָחה נַ ְפ ִׁשי ל
נ ַֹח ַּב ֶּמה ְּתכַ ֶּבה ַא ֲה ַבת לֵ ָבב?—וְ ֵאל נִ ְׁש ַּבע ְּברּוחֹו ֵמ ֲעבֹר ֵמי
ʔ
akhen mǝyudaʕi zǝnaḥtani mǝʔod ʕad ki qǝraʔtikha ʔavi Zanoaḥ
yom šulǝḥa nafši lǝšotet baʔɛnoš loʔ maṣǝʔa ki ʔim bǝkha manoaḥ
bamɛ tǝkhabbɛ ʔahavat levav?—vǝʔel nišbaʕ bǝruḥo meʕavor mey Noaḥ
F i g u r at i v e L a n g ua g e a n d S o c i o c u lt u r a l B ac k g r o u n d [ 141 ]
421
to see if the waters had subsided from the face of the ground; but the dove
found no place [of rest] to set her foot” (Genesis 8:8–9). This allusion gen-
erates a metaphorical interplay between the Flood story and the verse line.
Noah’s ark, the Flood, and the waters are deleted, thus foregrounding and
amplifying the feeling of stormy unrest and instability in the “wanderings”
of the soul among humans.
But what has been deleted—the many waters, Noah, and the
Flood—is restored in two additional powerful metaphysical conceits in
verse line 3: “With what will you quench the love of the heart?—as God /
Swore by His Soul that the waters of Noah should no more go over.” The
first hemistich alludes to the verse “Many waters cannot quench love, /
neither can floods drown it” (Song of Songs 8:7). In fact, the hemistich
contains a rhetorical question, to which an obvious answer would be the
biblical verse. The verb quench transfers to “love” the transfer feature
<fire>, thus rendering love intense like a fire that consumes the lover.
The phrase “Many waters cannot quench” suggests two different possible
construals: (1) love is a kind of fire that water cannot affect, or (2) love
is a fire that can be affected by water, but in this case it is so power-
ful that no amount of water can extinguish it. The verse in the Song of
Songs seems to suggest the first possibility. The question “With what?”
in the poem apparently proceeds in the same direction. Then, however,
the sequel creates a surprise: Many waters may perhaps quench love,
but in the present instance you do not have that amount of water; my
love for you is so powerful that a flood is needed to extinguish it. This,
however, is impossible, owing to God’s covenant that “never again shall
there be a flood to destroy the earth” (Genesis 9:11). In this instance, Ibn
Gabirol uses a verse from Isaiah as an inlay that, in turn, alludes to the
verse in Genesis:
This suggests yet another twist: In Ibn Gabirol’s poem the flood is not
merely a means to destroy the Earth; it is beneficial, and God’s covenant
is harmful.
The most outstanding English metaphysical poet of the seventeenth
century, John Donne, cannot be suspected of ever having heard of Ibn
Gabirol or read his poems. But in one of his Holy Sonnets he develops the
same metaphysical conceit and in a very similar manner:
Excerpt 6.
You which beyond that heaven which was most high
Have found new spheres, and of new lands can write,
Pour new seas in mine eyes that so I might
Drown my own world by my weeping earnestly,
Or wash it if it must be drowned no more.
Here as well many tears serve as a metonymy for intense emotion; all the seas
would not be enough to have the required amount of water for the tears. Then
Donne alludes to a passage from 2 Peter in the New Testament referring to
the destruction of the world by water and fire and to God’s covenant to Noah.
F i g u r at i v e L a n g ua g e a n d S o c i o c u lt u r a l B ac k g r o u n d [ 143 ]
41
In eleventh-century Spain things are less clear. But the encounter of the
three great monotheistic religions in Spain generated tensions, with no solid
ground for preferring one religion over another. Yehuda Halevy’s treatise
Hakuzari (The King of Khazars) is an illuminating document from this point
of view. The Khazars were a nation who founded an empire in the seventh
century in the northern Caucasus along the Caspian Sea, where over time
Judaism became the state religion. Halevy’s treatise is a Platonic dialogue
that tells the fictional story preceding the king’s conversion and discusses a
wide range of controversial issues, from religious to cultural. The king invites
a Christian, a Muslim, and a Jewish scholar and questions them about their
religion to find out which one is the true religion. The mere need to write
such a treatise highlights the problems facing a society in which not one but
three religions vied for primacy. For our purposes it makes little difference
that in the treatise of the Jewish poet-philosopher the Jewish position wins
the day. It is the very need to convince a neutral outsider that is significant.
From a literary point of view it is important to note that Yehuda Halevy
had an attitude of cultural ambivalence even toward poetic prosody. As
is well known, he was one of the greatest poets (if not the greatest) who
wrote Hebrew poetry in the pegs-and-chords meter imported from Arabic
poetry.1 But he has his Jewish scholar say harsh things about the inferiority
of the pegs-and-chords meter, as compared with the prosody of the book
of Psalms (cf. Rosen 1994).2 Furthermore, as Excerpt 4 suggests above
(as well as some other poems by Ibn Gabirol), poets and educated readers
of the time had to cope with a perplexing problem in theology: namely,
the need to integrate the biblical concept of a personal Creator with the
Neoplatonic conception of creation as an emanation of light.
This approach to the sociocultural background of Hebrew poetry in
eleventh-century Spain gets massive support from Ross Brann’s 1991 book
The Compunctious Poet: Cultural Ambiguity and Hebrew Poetry in Muslim Spain:
1. In medieval Hebrew poetry in Spain, meter, imported and adjusted from con-
temporary Arab poetry, was based on full vowels (termed “chords”) alternating in all
sorts of combinations with units consisting of a schwa mobile plus a full vowel (termed
“pegs”). The schwa mobile is, in Semitic languages, a reduced vowel, similar to the first
vowel in English against.
2. The critical sources for evaluating Halevy’s ambivalence toward Hebrew verse are
the two complementary texts: book 2 of The Kuzari (64–78) and the Judeo-Arabic
“Treatise on Hebrew Meters” (Brann 1991: 96).
divine judgment and life in the hereafter. The ideal man lived in both worlds and
found the ambiguity most attractive. (1991: 11)
On one point, however, Brann expresses a view opposite to the one underly-
ing the present work: “As liturgical poets, the courtier-rabbis of the Golden
Age therefore strike us as ‘poets of high seriousness’ seeking truth, serving
the community of Israel, and giving voice to its spiritual yearnings. But as
secular poets they present themselves to us as ‘poets of wit and whimsy’ in
the service of courtly ideals” (1991: 10).
As Excerpt 4 above indicates, some of the best samples of medieval devo-
tional poems are examples of metaphysical wit. Elsewhere (Tsur 2008b)
I have compared a devotional poem by Shlomo Ibn Gabirol with a poem by
Levy Ibn Altaban offered as a tribute to Ibn Gabirol. They have the same
monorhyme, the same meter, and similar phrases for opening and clos-
ing. Yet the two poems differ in two crucial respects: Ibn Altaban’s poem is
far inferior to its model, and it is typical of “Platonic” poetry, whereas Ibn
Gabirol’s is an exquisite specimen of metaphysical poetry. Perhaps, how-
ever, all the disagreement between Brann and me is a matter of terminology
rather than essence. In the terminology of the present chapter, some of the
best specimens of liturgical poetry “seek truth” in the style of Metaphysical
wit, whereas the majority of secular poetry of love, wine, and garden descrip-
tions are in the Précieux line. Brann does not go into such stylistic details,
and perhaps he does not use “high seriousness” in Matthew Arnold’s sense.
We will need to account for the relationship between the kind of poetics
discussed here and the kinds of sociocultural periods in which it tends to
occur. But before doing so, I would like to point out a figurative construc-
tion for which Moses Ibn Ezra had a particular predilection—genitive
phrases of the “concrete of the abstract” form:
Excerpt 7.
ר ּפ ְרסּו ֲעלֵ י ֹֻשלְ ַחן ְמזִ ָּמה ָע ְרכּו
ָ וַ ּיעֲ ִבירּו קֹול לְ לֶ ֶחם ָּדת ֲא ֶֹש
עּודה ָמ ְסכּוָ שֹכֶ ל וְ ֶאל יֵ ינֵ י ְת
ֵ לְ כּו ֶאל נַ ֲהלֵ י :ולִ כָ ל צְ ֵמ ֵאי ִּבין
They passed the word to the bread of justice which
they sliced on a table of judgment they set
And to all those thirsting for wisdom: go to the rivers of reason
and to the wines of moral obligation they poured. (my translation)
F i g u r at i v e L a n g ua g e a n d S o c i o c u lt u r a l B ac k g r o u n d [ 145 ]
46
1
Excerpt 8.
לֹוח ִמים ָּפ ָרסּו
ֲ ֻׁשלְ ַחן ּכְ ִזָבים עָ ְרכּו בֹו ַמ ֲאכָ ל ֶ ק ֶׁשר וְ לֶ ֶחם
They set a table of lies with food of conspiracy
on which they sliced bread of animosity. (my translation)
Excerpt 9.
Excerpt 10.
, ֵּבית ַהעֶ צֶ ב,ּגּומיָך
ֶ ּכְ ַב ָּנ ִאים עַ ל ִּפ
.ִמּיָ ד ֶאל יָ ד נִ ְמסֹר ֶאת ֶא ֶבן ַה ֵּטרּוף
Like masons on your scaffolding, House of Sorrow,
We pass on, hand to hand, the stone of Insanity.
Excerpt 11.
ָשֹמונִ י ּפֹה ּכַ ָּתף לָ ֵׂשאת ַא ְבנֵ י ַהּבֹהּו
.אֹובים
ִ ְּגומי ַה ַּמכֵ ִּב ִפ
11. I have been made a bearer, to shoulder the stones of Chaos
On the scaffoldings of sufferings.
“Waiter” does not illustrate “Fate,” nor does “winery” illustrate “Strangers’
Land.” Likewise, “stone” was not chosen as an apt representative of
“Insanity” or of “Chaos,” nor is “scaffolding” an apt representation of
“sufferings.” “Stone” was chosen because it fits into a narrative including
masons, bearers, and scaffolding, just as “Waiter” and “winery” make up
a coherent narrative. It is only when the abstract terms (which also rein-
force each other) are inserted that two incompatible universes of discourse
emerge, each striving to establish itself in the reader’s perception. No
“Platonic” representation of ideas can occur.
Ḥayyim Naḥman Bialik was, indisputably, the greatest figure in
so-called modern Hebrew poetry. His early poetry grew out of what is
termed “Enlightenment” poetry, but its main bulk was conspicuously
Romantic. Toward the end of his career he adopted some of the prosodic
features of modernism, as well as some features of Mannerist figurative
language. In the next two excerpts I compare two poems in which the
same image occurs in a “Platonic” structure and an “arbitrary genitive,”
respectively:
F i g u r at i v e L a n g ua g e a n d S o c i o c u lt u r a l B ac k g r o u n d [ 147 ]
4
81
Excerpt 12.
י־מעִ י ִמ ַזְּבח לְ ַב ְבכֶ ם ֶֹש ָֹּש ַמם
ְ ֵוַ עֲ ל
.יְ ִֵילִ יל וִ ַיפ ֵהק ֲהתּול ַה ִֹּש ָּממֹון
And upon the desolate3 wreck of the altar of your hearts
The cat of desolation will be wailing and yawning. (my translation)
Excerpt 13.
,ִּב ְהיֹות יֵ ֵמי ַסגְ ִריר ְּוב ֶנ ֶֹשף ַהּלַ יְ לָ ה
,ִּב ְֹשלֹוט ְּב ֵבית ָא ִבי ְּד ִמי ַק ְדרּות ּגְ דֹולָ ה
דֹומם ֵ ְוַ ֲחלָ לֹו ַה ֶמ ֻדּכָ א ּכְ מֹו הֹזֶ ה ו
.ֹשֹומם
ֵ וְ חֹולֵ ם ֲחלֹום עַ גּום ִּבכְ נַ ף ִֹשּקּוץ
, ְּד ִמי ַדּלּות ְמנֻ ּוֶ לֶ ת,זֹה עֲ ַקת ַה ַּד ֲהקּות
—ֹשֹואלֶ ת ֶ ִּבנְ ׂשא ֶֹש ַבע נְ ָפֹשֹות עַ יִ ן
; ִּד ְמעָ ה נֶ עֱ צֶ ְרת,ֹשֹוממֹות ֵ זָ וִ ּיֹות
,ַעל־ּגַ ֵּבי ַהּכִ ָירה ֲחתּול ְמיַ ּלֶ לֶ ת
,אֹור—ּב ִמ ֶׁש ֶא ֶרת
ַ ְׂש,ַּב ַּסל ֵאין פַּ ת־לֶ ֶחס
— ּוגְ ִר ִיסין ַּב ִּסיר,לַ ֲה ַמם ֵאין ּגַ ֶתלֶ ת
ָאז ֵהצִ יץ ַהּצְ ָרצַ ר ִמּנִ ְק ַרת ַה ִּקיר
, ָה ֵר ָיקה,וִ ינַ ֵּסר ִֹש ָירתֹו ַה ֵיֵב ָֹשה
.ּבֹוק ָקה ֵ ִנ ְֹש ָמ ִתי,קֹוס ָסה לִ ִּבי ְ ּכָ עָ ׁש
, ָּבכָ ָתה, ל ֹא־נִ ֲח ָמה,ל ֹא–זָ עֲ ָפה ִֹש ָירתֹו
;ֹשֹומ ָמה ָהיָ ָתה ֵ —גַּ ם־קֹב ל ֹא יָ ָדעָ ה
, ּכַ ֲה ֵבל ַחּיַ י ְּת ֵפלָ ה,שֹומ ָמה ּכַ ַּמוֶ ת ֵ
. ְּבלִ י ַא ֲה ִרית וְ ִתכְ לָ ה ֲא ֵבלָ ה,וַ ֲא ֵבלָ ה
In the days of rainstorm, and in the darkness of night,
With the silence of great blackness, reigning in my father’s house,
Its depressed void like a daydreamer, silent—
Dreaming a grievous dream, under the wings of a desolate abomination
It is the hardship of need, the silence of degenerate poverty,
With seven souls raising questioning eyes—
The corners desolate, the tears stopping
Upon the stove the cat is wailing,
No bread in the basket, no leaven in the kneading trough,
No ember for heating, no groats in the pot,
Then the cricket peeked from the crack in the wall,
3. In Hebrew, “desolate” is expressed by a line-final relative clause that can qualify
either “the wreck,” or “the altar,” or “your hearts” but most likely the whole package.
Such thing-free and gestalt-free entities do not conflict visually with the
solid objects described. Furthermore, the inner space of a room is absence
that looks as though it were something. When the void, the inner space,
“dreams,” its activity is imperceptible to the senses, but its mental energy
bestows an intense presence on this absence. Thus, desolateness and a cat
on an altar/stove may contribute to metaphysical wit in one poem but to
an emotional atmosphere in another. Significantly enough for the present
argument, Bialik had recourse to such metaphysical figures in a group of
F i g u r at i v e L a n g ua g e a n d S o c i o c u lt u r a l B ac k g r o u n d [ 149 ]
501
At the beginning of this chapter I said that the aesthetic response is based
on adaptation devices turned to aesthetic ends. I also claimed that the
Mannerist style—especially of the Metaphysical variety—tends to occur in
eras and societies marked by strife, radical changes, and disorientation in
which cultural disintegration exceeds a certain limit, there is more than one
scale of values, and there is no agreed-upon basis to decide between them.
Such a description, however, would suit Romanticism as well, for example.
In fact, the Romantic era, on the one hand, and the seventeenth and twen-
tieth centuries, on the other, differ mostly in their relative degree of dis-
integration. Romantic poetry handles disintegration with the help of the
orientation mechanism. The greater the disintegration, the more pointed
the effect of the mechanism of integration and orientation—but only up to
a certain point. Beyond that point, the disintegrating environment escapes
from the orienting mechanism’s control, and a different kind of coping pro-
cedure must be instantiated. Orientation is the ability to locate oneself in
one’s environment with reference to time, place, and people. This process is
marked by speed and relative imprecision. One gathers much information
via all the senses about one’s environment as well as about oneself, inte-
grating it as fast as possible. This is the source of the intuitive and impre-
cise nature of orientation. This makes it possible to only distinguish a few
solid objects; its essence is to find relationships that enable the individual
to make fast decisions regarding overall directions.
When disintegration exceeds a certain point, this mode of adaptation
becomes inadequate. In such cases we must find alternative modes of adap-
tation and orientation. Information “about oneself, like all other informa-
tion, can only be picked up by an appropriately tuned schema,” says the
cognitive psychologist Ulrich Neisser (1976: 116). When, in circumstances
of extreme disintegration, something suddenly seems to go wrong, one has
to check the tuning of one’s own schemata: “Consciousness, according to
Bartlett, enables an organism ‘to turn around its own schemata’ ” (Miller and
Johnson-Laird 1976: 150). I have called such awareness “meta-awareness.”
It should be pointed out that critical philosophy is characterized by similar
terms. Not unlike the poetry of disorientation, it tends to be prevalent in
societies dominated by more than one set of values, where there appear
to be no unquestionable truths, and when philosophy cannot take its own
When human thought turns around and examines itself, where does the inves-
tigation start? . . . The short answer . . . is that there are two forms in which the
data to be investigated may be presented. They may be presented in a psycho-
logical form, as ideas, thoughts and modes of thought, or they may be presented
in a linguistic form, as words, sentences and types of discourse. Kant’s critique
starts from data of the first kind, and the second wave of critical philosophy, the
logico-analytic movement of this century, starts from data of the second kind.
(Pears 1971: 27–28)
F i g u r at i v e L a n g ua g e a n d S o c i o c u lt u r a l B ac k g r o u n d [ 151 ]
521
When Ibn Gabirol in Excerpt 5 has recourse to the phrase “”א ִבי זָ נ ַֹח
ֲ (father
of Zanóaḥ [father of neglect]) in two different senses that follow two dif-
ferent kinds of logic, the common communicative logic and the “riddling”
logic, or in some of the conceits discussed above, precisely such smashing
of language may occur that deal the reader a shock, followed by recovery
from the shock and readjustment to the universe of the poem.
This view may provide a clue to solving another riddle. Why should a reli-
gious or mystical poet ever want to have recourse to such ingenious, artifi-
cial modes of expression? As I have just argued, the “smashing of language”
deals a shock to readers, making them suddenly doubt their comfortable
relationship with language. This shock is not unlike the sense of disorienta-
tion and confusion associated with mystical paradoxes, as argued by Steven
T. Katz:
Such linguistic ploys exist in many places throughout the world, usually con-
nected with the conscious construction of paradoxes whose necessary violation
of the laws of logic are intended to shock, even shatter, the standard epistemic
security of “disciples,” thereby allowing them to move to new and higher forms
of insight or knowledge. That is, mystics in certain circumstances know that
they are uttering nonsensical propositions, but in so doing they intend, among
other things, to force the hearers of such propositions to consider who they
are—to locate themselves vis-à-vis normal versus transcendental “reality.”
(1992: 7–8)
F i g u r at i v e L a n g ua g e a n d S o c i o c u lt u r a l B ac k g r o u n d [ 153 ]
541
more than once in different cultural periods and centers, under similar
psychological and cultural pressures. Shlomo Ibn Gabirol and John Donne
could take the metonymic relationship between tears and certain emotions
from their everyday experience, whereas they certainly took the story of
the Flood and God’s covenant never to destroy the Earth again by flood
from the same source. But they may have independently generated their
respective conceits of joining tears with flood via violence by applying to
them the same “metaphysical” or psychological logic. This chapter has
explored the raison d’être for such “unlikely” conceits.
F i g u r at i v e L a n g ua g e a n d S o c i o c u lt u r a l B ac k g r o u n d [ 155 ]
561
571
CHAPTER 7
the assumption that constitutive elements are the same from language to lan-
guage, if adjusted by a touch of equivalence. The iambic pentameter is metrical
in the same way that the alexandrine is metrical (though this is clearly untrue).
Rhyme in English is the same as rhyme in French, even though rhyme in French
recognizes different degrees of rhyme, makes alternating rhyme-gender a prin-
ciple of construction, and rhymes on endings and suffixes, which necessitates
a certain practice of avoidances, all features unknown to English rhyming. To
encourage readers to think of French and English rhymes as equivalent is seri-
ously and irresponsibly to mislead them. (2011: 72–73)
efface fine distinctions like the ones suggested by Scott’s paragraph above,
those distinctions can serve as a basis for translators’ decisions: they must
choose the nearest option available in the target language to the device in
the source poem. I even tend to agree with Scott’s next assertion, though,
as we shall see, in a less categorical version and with different implica-
tions: “Literary translation makes no sense to me if the reader of the tar-
get text (TT) is ignorant of the source language and source text (ST). My
approach to translation always presupposes that the reader of the TT is
familiar with the ST. This presupposition makes the act of translation a
linguistically dialectical act, and an act of textual comparison” (2011: 67).
In my work I have had recourse, in my own way, to both kinds of solu-
tions. I was brought up in the Hungarian language and literature. When,
at the age of sixteen, I mastered Hebrew, I devoted myself to a realization
in Hebrew of the conception of translation I adopted from Babits, Tóth,
and other great Hungarian translators: that is, a rhymed and metered
poem in the target language that makes as good a poem as possible and
chooses those options from the target language that are closest to the ones
in the source language. The precision of the translation depends on how
fine-grained the sign units of the target system are. If the target system
is sufficiently fine-grained and its nearest alternatives are chosen to rep-
resent a source phenomenon, the translation may evoke a perception that
the two are “equivalent.” I virtually gave up these activities while writing
my D.Phil. dissertation, when I drifted in the direction of Scott’s concep-
tion and beyond, for practical rather than theoretical reasons. In my disser-
tation and many of my ensuing publications I frequently analyzed poems
in languages other than the one I was writing in. I quoted the original
poem, providing a literal translation (and, if necessary, a transliteration).
Then I presented a close reading, pointing out the subtleties in the poem’s
sound stratum, units of meaning stratum, and projected world stratum
(the extralinguistic possible world to which the poem refers) and even tried
to integrate all of them (see the discussion below of the Hungarian and
Hebrew translations of Verlaine’s “Chanson d’Automne” and the Hebrew
translations of the line from Hamlet). I felt that in this way I could con-
vey more of the subtleness of the source text. In my artistic translations
I had the satisfaction of conveying as many subtleties of the original poem
as possible; later the suppressed subtleties had their way, paralyzing my
translation activity.
Readers can derive two different kinds of experience from these two
types of output. They can have a more direct, imaginative experience of an
integrated aesthetic object that is equivalent in some sense to the original;
or, alternatively, they can have a more purely rational understanding of a
T h e T r a n s l at e d P o e m a s a n A e s t h e t i c O b j e c t [ 159 ]
6
01
greater number of the source poem’s subtleties, which may or may not lead,
eventually, to a direct imaginative experience of it. Clive Scott’s solution is
somewhere in the middle.
Which solution is right? This question makes no sense to me. Each one is
the right solution relative to the kind of experience a translation is meant
to evoke. Translators and literary scholars are perfectly free to choose one
conception or another; the same person may have recourse to different con-
ceptions at different times. This is not an argument of the anything-goes
type. The final result will be judged not by whether one kind of solution
or another is selected but by how well the chosen solution instantiates its
respective conception. Each kind of solution does something better than
the other kinds do and fails to do things that the other kinds may do very
well. Good reasons can be given to prefer precisely this kind of solution to
the other ones—with reference to each kind. To use Morris Weitz’s (1962)
term, a “crucial recommendation” determines which package of gains and
losses to prefer. From this point of view, Clive Scott’s article does not pro-
pose the best kind of solution but is a crucial recommendation for one kind
and enumerates excellent reasons to support it. The present chapter makes
a crucial recommendation for a different kind of solution and supports it
with the best reasons I can mobilize. As we shall see, Douglas Hofstadter
states, no less categorically, a position that collides head-on with Scott’s.
There is a parable by Izmailov about the cuckoo who tells her neighbors
in the provinces about the wonderful song of the nightingale she heard
in a faraway country. She learned this song and is willing to reproduce it
for the benefit of her neighbors. They are all eager to hear that marvelous
song, so the cuckoo starts singing: “kukuk, kukuk, kukuk.” The moral of
the parable is that this is what happens to bad translators of poetry. My
thesis is that Izmailov does an injustice to the cuckoo (not to some transla-
tors). When you translate from one semiotic system to another, you are
constrained by the options of the target system. The cuckoo had no choice
but to use cuckoo language for the translation. The question is whether she
utilized those options of cuckoo language that were nearest to the nightin-
gale’s song. After all, Izmailov himself committed exactly the same kind of
inadequacy he attributes to the cuckoo. The bird emits neither the speech
sound [k]nor [u]; it uses no speech sounds at all. But every poet in human
language is constrained by the phoneme system of their language; they can
translate the cuckoo’s song only to those speech sounds. Their translation
will be judged adequate if they choose those speech sounds that are most
similar in their effect to the cuckoo’s call.
How do systems of music sounds and verbal signs take on perceptual
qualities endemic to other systems, such as human emotions or ani-
mal calls? At this stage of my argument I only want to point out that the
resources available in the target systems impose severe strictures on the
process. Usually only very few features or configurations thereof are avail-
able in the target systems that can be shared with the source phenomena.
So, the best one can do is to choose the nearest options available in the
target system. Minute differences may suffice to transform the perceived
character of a complex whole. As Krueger observed, the overall perceived
qualities of “total complexes” are determined by minute differences: “It
has been observed over and over that the smallest changes in experi-
ence are felt emotionally long before the change can be exactly described”
(1968: 100–101).
In onomatopoeia, the phonological system of a language cannot repro-
duce the actual sounds of, for instance, the cuckoo’s call: neither the
minor-third interval, nor the sound quality, nor the abrupt onset. The bird
says neither [k]nor [u]. The only thing one can do is to choose the speech
sounds with the nearest formant structure (see Figure 7.1). A symphony
Figure 7.1 Wave plot, the first and second formants of the cardinal vowels i-a-u, and the
European cuckoo’s call. (Formants are concentrations of overtones that determine vowels
and sound color.) Note that the formants of the bird’s call are most similar to, but not iden-
tical with, the vowel [u]. (Produced on SoundScope.)
T h e T r a n s l at e d P o e m a s a n A e s t h e t i c O b j e c t [ 161 ]
621
Figure 7.2 Wave plot and pitch abstract of the European cuckoo’s call and the cardinal vow-
els read by a professional reader.
Figure 7.3 Sound waves and pitch extract of the imitation of the cuckoo’s call in Leopold
Mozart’s “Toy Symphony.” (Produced on Praat.)
631
orchestra, by contrast, can reproduce the minor-third interval but not the
formant structure of the call.
The nearest option available to codify the abrupt onset of the call in
human speech is the abrupt consonant [k]—all the other features of [k] are
irrelevant. This use of voiceless stops to indicate abrupt onset appears to
have some intercultural validity. The Chinese word for “cuckoo” is pu-kû. In
Japanese we have the semantically based hatodokei = “dove” + “clock” but
also poppoo, kokyu, and kakkou. In orchestral versions, the abrupt onset is
indicated more directly (see Figure 7.3). Thus, the voiceless plosive [k] is a
bundle of perceptual features, a subset of which is frequently exploited by
the context to suggest some abrupt metallic noises such as “ticktack” or
“click”; but in the case of “cuckoo” only the perceptual feature [+ abrupt] is
utilized. Thus, the same elements or configurations in a target system can
serve as the “nearest option” for a wide range of source phenomena.
The philosopher John Dewey (1980) and others following him conceive
of an aesthetic object as an elegant solution to a problem. Such an elegant
solution is relevant to a work of art only if both the problem and its solu-
tion can be discerned in it at the same time. Flute players are frequently
praised as follows: “The flute sounds exactly like a soprano singer.” If the
soprano’s voice is so much more beautiful than the flute’s sound, why not
give the part to a singer? But no less frequently we hear the opposite praise
as well: “The voice of this singer sounds like a flute.” The praise does not
indicate preference for the soprano’s voice or the flute’s sound but, rather,
refers to the artistic achievement that one produces a certain sound qual-
ity using a basically different sound quality and that this can be perceived
even if one cannot see the source of the sound (on a record, for instance).
In such a case, both sound qualities are perceptible. Then one may say that
the problem of producing the sound quality of a flute using a human voice
has been elegantly solved.
There is an old Greek parable about a swineherd who entertained his
audience by imitating a piglet’s shriek. Another swineherd, who envied his
colleague’s success, hid a real piglet under his coat and whenever asked to
imitate a piglet’s shriek would pull its tail. The trick failed because the real
piglet was received with scorn. This parable throws an unfavorable light on
the audience’s taste, since it could not differentiate an imitation from the
real thing, giving, by the same token, a good lesson to the envious swine-
herd. The great Hungarian poet János Arany, in his “Vojtina’s Ars Poetica,”
gave a different interpretation of this parable that is in line with certain
nineteenth-century aesthetic views: “Not the truth, but its heavenly image”
is the essence of art. The imitator shrieked like the piglet shrieks in general,
whereas the real piglet may have shrieked as it had never shrieked before.
T h e T r a n s l at e d P o e m a s a n A e s t h e t i c O b j e c t [ 163 ]
641
I will interpret this parable in the light of my foregoing argument. The real
piglet will not please the audience more than the imitator even with its most
habitual shrieking, because it provides no solution to the problem. The imita-
tor’s voice, by contrast, remains basically human but weaves characteristics of
the piglet’s voice quality into it, thus creating the illusion of a piglet’s shriek.
With these two parables in mind, let us return to the question of how
a poetic translation can offer some equivalent in the target text of the
sounds and sound patterns of the source text. Speech sounds are per-
ceived as unitary events. If you replace a French word with its Hungarian
equivalent, you are bound to get different sequences of speech sounds: “To
translate alliteration by alliteration, or assonance by assonance usually
entails two second-bests: . . . one does not alliterate the same sounds”
(Scott 2011: 73). This is one way to view the issue. Structuralist phonology
since Jakobson, Fant, and Halle (1952), by contrast, views speech sounds
as bundles of distinctive features on which a closed system of binary oppo-
sitions is constructed. These distinctive features are systematically related
to certain nonlinguistic perceptual qualities. Thus, for instance, the front
vowel [i]is perceived as higher and brighter than its back vowel counter-
part [u] pronounced at the same pitch. Similarly, front vowels, in general,
are perceived as higher and brighter than their back vowel counterparts,
as, for instance, [e] and [o] as well. This approach is more fine-grained: it
gives the translator the opportunity to choose distinctive features and
nonlinguistic perceptual qualities in the target language that, if not the
same speech sounds as in the source language, are nearer to, or more
remote from, the ones in the source language. Let me give an example.
The French word for “violin” is violon; the Hungarian word is hegedű. The
Hungarian word is conspicuously unlike its French counterpart. The for-
mer approach can only acknowledge this conspicuous unlikeness. The lat-
ter approach can assign a structural description to it. Thus, for instance,
in the French word, a voiced, continuous, aperiodic and a voiced continu-
ous, periodic consonant as well as two back vowels (oral and nasal, respec-
tively) are present.1 In the Hungarian word, two abrupt, voiced stops [g,
d], the front vowel [ɛ] (twice), and a middle vowel [ű] are dominant. The
oral back vowel [o] is perceived as relatively dark, and the nasal back vowel
[õ] is perceived as even darker, whereas the front vowel [ɛ] is relatively
bright, while [ű] has a particular sheen. Thus, we can say that they are not
1. In periodic speech sounds the same wave shape is repeated. Voicing is periodic.
The consonants [l, m, n, r, w, j] and all the vowels are periodic. The other consonants
are aperiodic. In voiced stops and fricatives, such as [b, v], the [p, f] ingredients are
aperiodic, whereas the voicing ingredient is periodic. Periodic speech sounds are near
to musical tones; aperiodic speech sounds are near to noises.
merely as different as any other two words but are contrasted on a great
number of distinctive and perceptual features.
Below I compare Verlaine’s “Chanson d’Automne” with two Hungarian
translations and a Hebrew one. The Hungarian translator Lőrinc Szabó
uses the straightforward translation hegedű, generating a conspicuously
inappropriate sound effect. Árpád Tóth, by contrast, has recourse to a
metonymy of the violin, húr (string), which contains a dark long back vowel
and a voiced continuous periodic consonant. Owing to the vowel harmony
in Hungarian, the possessive suffixes hegedűje and húrja also contrast a
front vowel with a back vowel. Thus, we can say that Tóth chose a nearer
available alternative in Hungarian than Szabó, even though, from a seman-
tic point of view, Szabó used a straightforward translation, whereas Tóth
had recourse to a conspicuous metonymy. In the word húr no nasal vowel
is available; even the best translator cannot remedy such a state of affairs.
But, as we shall see, Tóth amply compensates for this in other words.
Problem solving implies “constraints” or “stringencies.” In poetry, the
grammar and vocabulary of the language as well as the various kinds of
poetic conventions are the most notable ones. These stringencies are fre-
quently incompatible. Syntax may demand one word order; meter, a differ-
ent order of the same words; and rhetorical emphasis, yet another, while
manipulating a certain word into the rhyme may demand still another
word order. The poet must find an elegant way to satisfy the demands of
one solution without infringing on the demands of another.
Thus, for instance, a poetic style demanding more unpredictable figura-
tive language than usual may facilitate the manipulation of a word required
by rhyme for the line ending, without violating word order, rhetorical
emphasis may justify certain deviations from word order required by syn-
tax for the sake of meter or rhyme, and so forth. Poets can be quite inven-
tive and unpredictable in finding such elegant solutions. This holds true
for all poetry. As to translations, the translator of poetry must meet one
more all-important stringency: the translated poem must be as similar as
possible to the original poem in reproducing all these stringencies. Another
imperative is, of course, that the final result must make a good poem in the
target language. This is the sense in which I embrace Scott’s statement that
“literary translation makes no sense to me if the reader of the target text
. . . is ignorant of the source language and source text” (2011: 67), which is,
of course, not the meaning he intended.
One can, of course, compare a translated poem with its original and
point out all the figures of speech and sound patterns and other poetic
devices of the source text that were lost in the target text. Likewise, one
can point out all the poetic devices in the target text that do not occur in
T h e T r a n s l at e d P o e m a s a n A e s t h e t i c O b j e c t [ 165 ]
61
the source text. No translation in the world can survive criticism based on
such distinctions. But if the objective is to discover whether the translator
chose the closest options in the target language to the poetic devices of the
source text and whether he or she found a way to bring together as many of
them as possible without one violating the other, one may find that some
translations are more successful than others. Then, indeed, some target
texts may be reasonably akin to the source text in their sound patterns,
figurative language, syntactic structures, and idea content; some of them
may even be masterpieces in the target language. The translation will fail to
be identical with the source text but may present a reasonable equivalent
in the reader’s mind. The reader may derive aesthetic pleasure not from
the point-by-point resemblance between the source text and the target
text but from experiencing the elegant solution of a problem posed by a
great number of stringencies. In this context, equivalence with the source
text (in the sense discussed here) is just another stringency, and complying
with it may increase the reader’s aesthetic pleasure.
Les sanglots longs
Des violons
De l’Automne,
Blessent mon coeur
D’une langueur
Monotonne. (Paul Verlaine)2
Ősz húrja zsong,
Jajong, búsong
A tájon,
2. Listen to three readings of this stanza, by two native speakers of French and one
probably nonnative. https://www.tau.ac.il/~tsurxx/Soundfiles_counterproposal/
Soundfiles_counterproposal.html
S ont monoton
Bút konokon
És fájón. (trans. Árpád Tóth)3
Zokog, zokog
Az ősz konok
Hegedűje,
Zordúl szivem
S fordúl szivem,
Keserűre. (trans. Lőrinc Szabó)4
ִּבנְ ִהי ַמ ְמרֹור
הֹומה ּכִ ּנֹור
ֶ
, ֵַט ֵבת ָּפרּוע
הּלֵ ב ל
ַ וְ ֶא
ּכְ ֵאב חֹודר ֵ
וְ גַ עֲ גּוע
Binhi mamror
Homɛ kinor
Tevet paruʕa,
Vǝʔɛl halev
Ḥoder kǝʔev
Vǝgaʕaguʕa. (trans. Zǝʔev Jabotinsky)5
One of the prime sources for this poem’s striking musicality is derived
from its nasal vowels and the sound sequence -eur. This sound sequence
and the nasal vowels have two characteristics in common: from an acous-
tic point of view both are continuous and periodic, and from the point of
view of the infant’s phonological development, both are late acquisitions.
Following Jakobson (1968), I have argued elsewhere (Tsur 1992a) that the
infant’s latest phonological acquisitions have the greatest emotional and
aesthetic potential in adult language, for better or for worse. Among the
late acquisitions, such abrupt consonants as the affricates [pf] and [ts]
are “ugly” or express unpleasant emotions; the speech sounds that are
continuous and periodic from the acoustic point of view are musical and
“beautiful” and express pleasant emotions. The “beautiful” sounds abound
T h e T r a n s l at e d P o e m a s a n A e s t h e t i c O b j e c t [ 167 ]
6
81
the closed phonological system of the target language and, in the present
instance, whether he chose as many alternatives as possible involving the
distinctive features [+ continuous + periodic + nasal] that contribute to
the poem’s effect in the source language.
Tóth’s preference for nasal vowels and consonants in this translation
is obvious. Recall that the Hungarian word for “its violin” (hegedűje) was
retained in Szabó’s translation, whereas in Tóth’s it is replaced by a meton-
ymy, húrja (its string), introducing a dark back vowel ([u:]) and a sonorant
([r]), as well as eliminating, by the same token, such plosives as the /g/ and
/d/ of the proper term. The ű is a rather late developmental acquisition but,
according to Jakobson (1968), has a certain sheen that would be foreign to
the atmosphere of this poem (Hungarian ő and ű are brighter than their
French equivalents).
Encouraged by the Hungarian translator’s feast of nasality, I attempted
many years ago to render Verlaine’s poem in Hebrew; lines 4–6 run as
follows:
These words make ample use of dark back vowels and nasal consonants, as
well as of the voiced velar stop [g]in close vicinity. Nonetheless, the poem
refused to assume a musical quality comparable to the French original
or Tóth’s Hungarian translation. This was most clear-cut precisely where
I expected the greatest similarity, in the rhyme words yagon–hagon. At
that time I could only describe the difference in an intuitive fashion: the
gon sequence in the Hebrew rhyme sounded somehow too decisive, too
conclusive, too assertive, too solid, as compared with the correspond-
ing sequences in French and Hungarian. Phonetically and phonologically
speaking, in French we are dealing with nasal vowels proper, whereas in
Hungarian we are confronted with an allophone of the oral vowel, strongly
nasalized by coarticulation with the subsequent nasal consonant, which,
in turn, is attenuated by the ensuing voiced velar stop. In Hebrew, by con-
trast, though both [g] and [n] are present, there is no coarticulation, no
nasalizing effect.
T h e T r a n s l at e d P o e m a s a n A e s t h e t i c O b j e c t [ 169 ]
701
In what follows, I give a further example, this time from the domain of met-
rical equivalence, when different kinds of semiotic systems are involved. In
chapter 9, I use this same example to illustrate the perceptual difference
between the iambic and the ternary meters; here I use it to demonstrate
the issue of equivalences between versification structures in syllabic and
syllabotonic verses.
What is the equivalent of a poem’s meter when we translate it into a
language in which a different metrical system is prevalent? More specifi-
cally, I explore the problem of translating French alexandrines (based on
the syllabic metrical system) into languages in which the syllabotonic is
the dominant metrical system. I look at the problem of meter in translating
Baudelaire’s “Correspondances” into such languages as English, Hebrew,
and Hungarian. In an epilogue to my Hebrew volume of poetry transla-
tions I put forward a cognitive theory of translation similar to the pres-
ent one. Among other things, I compare the metrical organizations of six
Hebrew translations of this poem. Two have no recognizable meter, one
is in a mixture of ternary meters, and three are in the iambic hexameter.
Which one of these solutions best suits the spirit of the French meter? And
what explanation can be given for this choice?
The metrical system dominant in English, Hebrew, Hungarian, and
some other modern literatures is the syllabotonic system: that is, the sys-
tem that determines, ideally, the number of syllables as well as the number
of stresses and their placement in the verse line. In iambic pentameter, for
instance, there are ten syllables in a line, and every even-numbered syl-
lable ought to be stressed. In the preceding sentences I used the phrases
“determines,” “ideally,” and “ought to be stressed” because in fact there is
such a discrepancy between the ideal and the real stresses that the most
fruitful way to speak of poetic rhythm is to define the metrical pattern
and the stress pattern separately and identify where the two converge
and where they diverge. In French poetry (and in certain other Romance
languages), by contrast, the syllabic system is the dominant metrical sys-
tem: that is, the system in which the syllables are counted, whereas the
T h e T r a n s l at e d P o e m a s a n A e s t h e t i c O b j e c t [ 171 ]
721
number of stresses and their placement are ignored (Scott would qual-
ify this statement: they are not ignored, only unpredictable). In French
poetry, usually two more organizing prosodic principles are added: the
caesura at a fixed point (the middle) of the verse line and a predictable
arrangement of “masculine” and “feminine” rhymes, according to certain
principles of symmetry. When translating a poem from French to English,
Hungarian, or Hebrew, the question necessarily arises of what syllabo-
tonic meter will be equivalent to the French poem’s meter from the point
of view of perceived effect.
In English, Hebrew, and Hungarian, French alexandrines are usually
translated into iambic hexameter. My argument will rest on a Hebrew
translation that deviates from the prevalent practice. So, unfortunately,
I will have to keep my discussion at a highly general level.
In all my work in metrics I distinguish metrical pattern, stress pattern,
and pattern of performance. When stress pattern and metrical pattern
converge, they yield strong prosodic gestalts with a psychological atmo-
sphere of certainty and patent purpose. When they diverge, the verse is in
danger of falling into chaos. The coherence of the metrical foot depends,
to a considerable extent, on the downbeat. In binary meters (the iambic
and trochaic) only one upbeat “leans on” each downbeat for support; in
ternary meters two upbeats lean on one downbeat. Consequently, binary
meters are more stable than ternary meters and more resistant to disin-
tegration. In ternary meters, to prevent disintegration, performers are
inclined to subordinate the prose rhythm to the regular metrical beat. The
iambic foot, with its stronger gestalt, seems to tolerate greater deviance
and complexity. In extreme cases of deviation, to prevent chaos the per-
former needs to accommodate the divergent patterns in a strong gestalt
of additional grouping. This underlying strong gestalt, whether in the pat-
tern of meter or performance, makes rhythmicality possible when diver-
gent elements are present. As can be inferred from Woodrow’s tick-tack
experiments (see chapter 9), end-accented meters (such as the iambic and
the anapest) allow for greater flexibility in manipulating the time fac-
tor than beginning-accented meters (such as the trochaic and dactylic).
Hence the relatively greater rigidity of trochaic verse, which is manifest
in its “compelling” nature, observed by so many critics from Aristotle
to Zhirmunsky, Chatman, and the generative metrists. Consequently,
the iambic is more tolerant of deviations than the trochaic and the
ternary meters.
At this point I will refer once again to Scott’s objection to the claim that
“the iambic pentameter is metrical in the same way that the alexandrine is
metrical (though this is clearly untrue)” (2011: 72). My attitude toward this
claim depends very much on what is meant by “is metrical in the same way
that.” If it implies that the iambic pentameter is the most frequent meter
in English, Hungarian, and Hebrew syllabotonic systems, just as the alex-
andrine is the dominant meter in the French syllabic system, I wholeheart-
edly agree. But there are other possibilities as well. Whereas in English
and Hungarian syllabotonic verse the iambic is, indeed, the most frequent
meter, in Hebrew it is not; for reasons relating to average word length in
Hebrew, the anapest and amphibrach are more frequent. Yet, as we shall
see, in Hebrew as well, the iambic is felt to be more appropriate for the
translation of the alexandrine than some ternary meter.
The alexandrine, the most widespread meter in French, is syllabic. It
consists of twelve syllables (thirteen in “feminine”-ended lines), with a
compulsory caesura after the sixth. It makes no systematic use of contrast
between prominent and nonprominent syllables. Alexandrines by Racine,
Baudelaire, and other French poets are usually translated into the iambic in
languages as dissimilar as Hebrew and Hungarian. In my corpus of Hebrew
translations of Baudelaire’s “Correspondances,” however, one translation is
in a ternary meter, a mixture of the amphibrach and the anapest. Though
it is remarkably polished, the rhythm sounds strikingly unsuitable. It has
a vigorous, straightforward quality, quite unlike the vague impression the
poem makes in French.
One’s first response to such an encounter is surprise. Then it comes to
mind that surely it is only a matter of habit that the ear feels the iamb
rather than the amphibrach as the syllabotonic equivalent of the alexan-
drine. From a purely arithmetic point of view, both possibilities are equally
plausible. After all, 4 × 3 = 12 just as 6 × 2 = 12. The caesura after the sixth
syllable will coincide with a foot boundary, whether binary or ternary.
Something perceptual, however, resists this arithmetic equivalence. As
I have said, compared with the iambic meter, ternary meters are inflexible
and generate “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 direction” (Meyer 1956: 160) in
the poem, conflicting with the vague, ambiguous atmosphere, the elusive
apprehension of another reality, characteristic of the French original (cf.
Hofstadter’s translation of Armand Silvestre’s poem below).
At first sight, there appears to be no reason why Baudelaire or Racine
should not be translated into the anapest or the amphibrach. A closer look
at the French poem “Correspondances” even reveals that the “stress pat-
tern” of the first hemistich confirms the ternary anapest. But the second
hemistich confirms the binary iambic. The third one nearly confirms the
iambic (except for the “inverted first foot”), while the fourth confirms
T h e T r a n s l at e d P o e m a s a n A e s t h e t i c O b j e c t [ 173 ]
7
41
anapest, and so forth. Consider the first two lines, indicating the position
in which the “stress” occurs:
3
6
4
6
T h e T r a n s l at e d P o e m a s a n A e s t h e t i c O b j e c t [ 175 ]
761
I illustrate this essential tension and the lack of it by comparing two metered
passages with a text where there is no versification pattern (reproduced,
with the necessary changes, from Tsur 2008a: 145–146). The three excerpts
below are versions, by three different authors, of the same speech. One
of the most memorable instances of Gloucester’s villainy (in Shakespeare’s
Richard III) is the scene in which he gets rid of Lord Hastings:
the reader’s perception, thus enhancing each other. The group boundaries
intrude on the line, but the line strives to retain its integrity.
Shakespeare actually amplifies a technique that he found in Dolman’s
poem in The Mirrour for Magistrates: “Yf, traytor quod he? playest thou
with yfs and ands? / Ile on thy body avowe it with these hands.” Notice
the isolated Yf followed by a short vocative (traytor), followed by a short
parenthesis (“quod he”). The quoted yfs and ands not only diversify but also
intensify the subdivision of the line. They give rise, at the same time, to a
bold “antigrammatical” rhyme, by rhyming a noun with a conjunction (in
the plural!), involving a leap from first-order language to metalanguage and
back. It strengthens the closure of the rhymed couplet, thus heightening
its sense of unity.
It is revealing to compare this couplet of Dolman’s to Sir Thomas More’s
prose account of the same incident: “What quod the protectour thou seru-
est me I wene wt iffs & with andes, I tell the thei have so done, & that
I will make good on thy body traitour” (1963: 48). Verbally, More’s account
does not differ greatly from Dolman’s. It displays the same (or even more)
emphatic segmentation of the syntactic stream. Nevertheless, there seems
to be a perceptual difference between them, in that the segmentation
stands out less clearly in the unmetered version. The present comparison
highlights an illuminating aspect of the issue. The first two excerpts clearly
demonstrate how shifting phrases can enhance the prosodic unit. The third
excerpt foregrounds an unexpected, obverse, aspect: namely, how the pro-
sodic structure imposed on the phrases in Dolman’s poem renders their
shift far more emphatic than in More’s prose account.
T h e T r a n s l at e d P o e m a s a n A e s t h e t i c O b j e c t [ 177 ]
78
1
T h e T r a n s l at e d P o e m a s a n A e s t h e t i c O b j e c t [ 179 ]
8
01
HOFSTADTER ON TRANSLATION
He finds poetry that eschews the constraints of regular verse, like atonal music,
aesthetically objectionable because in it, he cannot see content wedded to form.
Instead, he sees a purely intellectual art in which “form is seen as the dog’s tail,
content as its body” (527), and the artist refuses to let the formal tail wag the
dog of content. (Dayan 2012: 12, quoting Hofstadter 1997)
On the whole he does a beautiful job: the English poems are good poems in
their own right and fairly convey the meaning. The versification is highly
polished. In this stanza, the meaning comes through, and so does the
rather moderate figurative language. There is one interesting exception. In
the English version, the dawn spreads “its canvases white,” whereas in the
French original, it spreads “the whiteness of the canvases.” There is a huge
difference between the two. In “canvases white” a physical action predicate
is applied to a physical object, resulting in physical movement in space. In
“the whiteness of the canvases” the physical action is applied to an abstrac-
tion, a thing-free quality, generating an intense, thing-free presence. To
be sure, we are speaking of the dawn’s canvases, which are immaterial and
have no stable visual shape, but, still, “the whiteness [or “candor”] of the
canvases” is an abstract quality of that immaterial entity. In other words, in
a concrete noun many abstract properties have “grown together”; the geni-
tive construction loosens the relationship between the abstract quality and
the object whose quality it is; the application of a predicate to the abstract
quality as an object suggests that it has some independent existence.
This change is not forced on the translator by the constraints of ver-
sification; he could solve the problem with a simple apostrophe, by writ-
ing “spreading its canvases’ white.” “Canvases white” suggests relative
T h e T r a n s l at e d P o e m a s a n A e s t h e t i c O b j e c t [ 181 ]
821
stability, whereas “the whiteness of the canvases” as the object of the verb
spread suggests massive presence, diffuseness, elusiveness, and intangibil-
ity. Thus “canvases white” generates a psychological atmosphere of relative
certainty, security, and purpose.7
As to versification, there seems to be a problem here relating to meter.
Hofstadter translates the French alexandrine into English anapests. A closer
look at the French stress pattern reveals that, just as in “Correspondances,”
the first hemistich is clearly anapestic. The second, third, fourth, and fifth
hemistichs, however, are clearly iambic. The sixth hemistich, again, is
clearly anapestic. The seventh hemistich is iambic with an “inverted first
foot,” while the last one contains two consecutive stressed syllables—
“manteau bleu”—rendering the pattern indecisive.
As I suggested earlier, the iambic is more tolerant of deviant stresses,
without falling apart; the anapest more rigidly exerts its will and tends
to suppress linguistic stress that happens to occur in a weak position
(and there are quite a few of them in the translation of this stanza). The
effect of the anapest here is by far less devastating than in Baudelaire’s
“Correspondances” because the content is less elusive; but still, it rein-
forces the sense of control and the atmosphere of relative certainty,
security, and patent purpose generated by manipulating the solid object
“canvas” rather than its thing-free quality into the referring position. To
use Hofstadter’s own words, “the indissoluble fusion of a medium with a
message, the unsunderable wedding of form to content as equal partners,”
has here the wrong effect—the unsunderable wedding of the ternary meter
to the manipulation of a relatively solid object into the referring position
of the phrase. Their indissoluble fusion suggests here a psychological atmo-
sphere of stronger-than-appropriate certainty and security (reinforced by
the internal rhyme at the caesura and the end of the first line, which articu-
lates the verse line into two clear-cut, terse segments). In other words, it is
not enough that form and content are unsunderably wedded; their inter-
play must generate the right kind of atmosphere or perceived quality.
NOTE ON TERMINOLOGY
Finally, I want to emphasize that Scott and I use two of our crucial terms,
rhythm and equivalence, in different ways. From my vantage point, Scott
7. As I have argued elsewhere (and see chapter 6), such constructions of “the
abstract of the concrete” abound in Baroque, Romantic, and Symbolist poetry and
in Whitman’s “meditative catalogue,” with similar effect.
T h e T r a n s l at e d P o e m a s a n A e s t h e t i c O b j e c t [ 183 ]
8
4
1
T h e T r a n s l at e d P o e m a s a n A e s t h e t i c O b j e c t [ 185 ]
8
61
71
8
CHAPTER 8
Figure 8.1 Mbira.
pitch, “high is less” in several important ways: while spatially high objects are
large and higher in quantity, high pitch is small, empty and little in quantity. . . .
Similarly, there is a very high consensus that higher pitches are thinner and
sharper (i.e., of a smaller mass). (Eitan and Timmers 2010: 415)
The consensus in choosing “thin” to represent “low” is high for mass/size
(.93), but the consensus to choose thin to represent “low” in pitch is low (zero).
(Eitan and Timmers 2010: 417)
Higher pitch is positively correlated with higher intensity, and higher
intensity is positively correlated with larger size, but larger size is negatively
correlated with higher pitch. Such intransitive relationships indicate that the
cross-domain mappings of pitch are underlined by several, sometimes conflict-
ing conceptual metaphors. (Eitan and Timmers 2010: 420)
The present chapter explores only this and related issues. I claim that these
findings are puzzling only as long as you view them within the Lakoffian
model of spatial metaphors.2
I propounded the theoretical model used here in Tsur 1992a. There I also
warn that the behaviorist notion of “mediated association” has penetrated
some cognitive discussions, usurping the place of other, more adequate
cognitive explanations. In my later work I elaborated on these notions in a
criticism of mediated association and of Lakoff. In my article “Size-Sound
Symbolism Revisited” (Tsur 2006a) I contrast my position with the
mediated-association conception; in Tsur 2000b, 2002b, I contrast my
position with both mediated association and the Lakoffian conception. The
following discussion draws liberally on these works.
Roger Brown, the great psycholinguist who crucially influenced my early
thoughts on cognitive poetics, puts forward a conception of mediated asso-
ciation with which I disagree: “If the subject is required to guess, he will
call the loud and resonant voice ‘thick.’ This need not be because the voice
shares some inter-sensory quality with the visual or tactile apprehension
of thickness. It could be because the voice is loud and creatures who have
loud voices are usually thick, a mediated association” (1958: 152–153). In
Tsur 2012c I reproduce the photos of a thick soprano and a thin bass singer.
2. I have written on this and related issues in Tsur 1992a: chap. 3. At that time I did
not think of stating my position with reference to Lakoff. So it was not a response to
Lakoff but, rather, mainly to my own perceptions.
M or e I s Up — S o m e of t h e T i m e [ 189 ]
9
01
More importantly, I argue that bass voices are perceived as thicker than
soprano voices not because creatures that have bass voices are usually thick
and heavy but precisely because they share “some inter-sensory quality
with the visual or tactile apprehension of thickness.” Whereas the relation-
ship between thick people and bass voices appears to be quite incidental,
the relationship between thick violin strings and “thick” and “low” sounds
seems to have good physical reasons. Sounds are vibrations of the air or
some other material medium. The thicker the string, other things being
equal, the slower the vibrations and the greater the wavelength. To this one
may add that the “lower” the tones, the greater the number of overtones
within the audible range. Thus, the lower tones have a “thicker” envelope of
overtones than the higher tones.
The notion of “mediated association” is the same as Lakoff’s notion of
“grounding”: “Contemporary theory [of metaphor] postulates that the
more is up metaphor is grounded in experience—in the common experi-
ence of pouring more fluid into a container and seeing the level go up, or
adding more things to a pile and seeing the pile get higher” (1993: 240).
More recently, Lakoff makes an even more extravagant claim: “In a ther-
mometer oriented vertically, the mercury goes up physically as the temper-
ature increases (metaphorically goes up)” (2012: 31). Here one must even
manipulate the physical object in order to make it literally “go up” rather
than, for example, “from left to right.” It would appear that “grounding” is
merely a more cognitive phrasing for what Brown calls “mediated associa-
tion.” To be more precise, it disguises the behaviorist origin of the notion.
It relies, however, more on lifelong conditioning than on cognitive pro-
cesses proper.
The source of Eitan and Timmers’s problem is not so much “that the
cross-domain mappings of pitch are underlined by several, sometimes con-
flicting conceptual metaphors”; rather, it results from the nature of physics,
on the one hand, and our perceptual coping with it, on the other. Briefly,
frequency and wavelength co-vary, but in opposite directions: when fre-
quency is greater, wavelength is smaller, and when wavelength is greater,
frequency is smaller. Thus, faster vibrations are up, as predicted by Lakoff,
but this entails that greater wavelength is down. Here the doctrinaire slo-
gan “more is up” is helpless in the face of the irrevocable facts of phys-
ics. Frequency is a temporal notion, and length, a spatial notion; that is,
in cross-modal mapping the spatial template of pitch is matched with the
temporal, not the spatial aspect of the sound stimulus. In other words, if
you conceive of the phenomenon as perceptual rather than conceptual, you
must perceive (in terms of Eitan and Timmers’s second experiment) or think
(in terms of their first experiment) that the higher the pitch, the thinner
and smaller the sound. Such complexities are beyond the scope of the sys-
tem of conceptual metaphors.
The reasoning underlying this claim is that the sequence of sounds
has a certain phenomenological character that is conveyed by an acoustic
sequence with a different structure (for instance, instead of relatively fast
vibrations and relatively short waves we hear “high” and “thin” sounds).
The minute time periods between vibrations exceed the resolving power
of the human ear. The ear, therefore, fuses the sequence into one continu-
ous whole with a unique phenomenological quality: we become conscious
of the resulting sound as “higher” or “lower.” Such a phenomenological
quality has greater survival value than perceiving and counting the minute
time intervals per second, even if this were possible. In the same way, the
overtones of a sound exceed the resolving power of the ear, so that it is
forced to fuse them into “tone color.” Again, hearing a wooden or metal-
lic noise versus the smooth thump of some jumping feline may make
all the difference and may have greater survival value than perceiving a
multiplicity of overtones. Thus, both tone color and pitch result from the
limitations of the human ear’s resolving power. We attend to the phenom-
enological quality, but at the same time we perceive, subliminally, some of
the acoustic structure.
The foregoing discussion throws some new light on Eitan and Timmers’s
following argument: “Cross-cultural studies of sound symbolism in lan-
guage suggest an association of physical magnitude with pitch, such that
larger size and mass, as well as secondary qualities associated with bodily
magnitude, like slower pace and dominance, are related to lower-pitched
tones in speech and in other vocal utterances” (2010: 407). “Secondary
qualities associated with bodily magnitude, like slower pace,” are, again,
typical mediated associations. In light of the foregoing we need no “medi-
ated association” here: “slower pace” is not among the “secondary qualities
associated with bodily magnitude.” “Slower pace” and “bodily magnitude”
are directly perceived in the stimulus, in the relative rate of vibration and
the relative wavelength.
This conception can also account for the following findings mentioned
in Eitan and Timmers’s general discussion: “ ‘Higher’ loudness and pitch
interact with other dimensions in contrasting ways. Increased loudness, for
instance, is ‘bigger,’ while increased (higher) pitch is ‘smaller’ ” (2010: 419).
These findings are in perfect harmony with my foregoing argument. Both
findings are correlated with wave size: increased pitch with smaller wave-
length, increased loudness with greater wave width.
A similar view underlies at least one cross-cultural study of size-sound
symbolism in language (Ultan 1978). By examining a total of 136 languages,
M or e I s Up — S o m e of t h e T i m e [ 191 ]
921
That which is longer is less short, that which is higher is less low, that
which is faster is less slow, and so forth. The two terms of each pair,
however, are not symmetrical at all, as we shall see. In Tsur 1992a I tried
to solve some of the problems concerning such antonymous adjective
pairs by invoking the cognitive research of Clark and Clark and the
Structuralist notion of “markedness.” At that time I was not concerned
with arguing against Lakoff. Only years later did I compare my analyses
with his theory.
According to the Clarks, in each pair of antonyms small children
first learn the term that denotes the pole that has more (long, high, fast,
more, etc.):
When three year olds are asked about two toy apple trees Does one tree have more
apples on it? or Does one tree have less apples on it?, they can correctly answer yes,
just as adults do. But what happens if they are then asked which tree has more or
which has less? . . . In response to both questions, the children tended to choose
the tree with more apples on it, treating questions with less as if they contained
more. (Clark and Clark 1977: 505)
The preference for the dimension where there is more has good cognitive
reasons behind it. It is easier to see which has more. This seems to be
the source of two facets of the antonymous pairs of adjectives in adult
language:
1. The pole that has more is the unmarked pole, which is selected when
there is no explicit reason to select one or the other pole. We ask, “How
long was the movie?” if we don’t know whether it was long or short for a
movie. “How short was the movie?” would suggest that we already know
that it was very short.
2. The unmarked term is also used to denote the whole scale. We usually
speak of “measures of length”; only in very special cases would we speak
of “measures of shortness.”
M or e I s Up — S o m e of t h e T i m e [ 193 ]
941
Here two crucial questions arise: First, why do we need at all to recode the
sequence of sounds into a spatial template? Second, as we have seen, there
is an inverse relationship between frequency and wavelength: greater fre-
quency goes, necessarily, with shorter wavelength, and vice versa. So why
do we match the unmarked pole of the spatial template precisely with the
unmarked pole of sound frequencies and not with the unmarked pole of
wavelength? In other words, why do we say fast vibration is up rather
than thick sound is up?
To the first question Lakoff would answer “the embodied mind.” I, by
contrast, offer the following answer. The spatial template serves as a facil-
ity for handling noncategorial sensory information. Neisser has noted
that “sounds inform us about events. While vision and touch enable us to
explore stationary environments, hearing tells us only about movement
and change” (1976: 155). Sound is a rapidly changing stream of informa-
tion, consisting of minute stimuli that most accurately signal change. We
can discriminate such rich sensory sound information only as long as it
reverberates in echoic memory, that is, for a few seconds only. In order to
store auditory information for longer periods of time, it must be recoded
in some more “stationary” form that depends less on the niceties of unique
sensory information and is more easily managed by memory. For present
purposes, there appear to be two such ways of recoding: categorization into
a phonetic code and translation into a code of spatial relationships (of a
more stable nature).
The first of these codes uses a system of abstract linguistic categories
that lend themselves to storing for relatively long periods of time, but at
the price of excluding most of the acoustic information. Suppose we ask,
“What did the man say?” and get the answer, “The man said ‘ba.’ ” We are
able to remember this syllable for an hour, or a day, or a month, or even
a year but will be incapable of even reliably noticing the acoustic cues by
which it differs from, for example, “da” or “ga.” Had we received only the
acoustic cues, we could reliably remember them for a few seconds at best.
Consider Figure 8.2. The density of the curves represents frequency. The
distance between the curves represents wavelength, whereas the devia-
tion of the curves from the midline represents amplitude. Obviously, the
distance between the curves (wavelength) co-varies with their density
(frequency): the shorter the wavelength, the greater the density. In the
nonspeech mode we are able to perceive the physical shape of such a sound
wave, but not in the speech mode. If it were possible to perceive such a
sound shape in the speech mode too, we would be hard put to recognize
the syllable ba or even to retrieve the sound sequence after several seconds.
The sound wave in Figure 8.2 gives information about the speaker’s artic-
ulatory gestures, which the listener’s articulatory system decodes as the
intended syllable, “ba.” Subliminally, at the same time, one may sometimes
also perceive some of the sound wave’s physical shape.
The second of these codes exploits the fact that spatial organization is
of a more stable nature (not necessarily determined by our bodily expe-
riences) than the sequence of sounds: it conceives of the relationship
between sounds as spatial relationships. Thus, if we cannot remember the
exact sensory information about a sound, we can still remember the exact
relationship between several sounds—that is, a scale or melody—and it
will be the same even if transposed from one scale to another.
The spatial structural template not only enables us to remember relation-
ships between sounds but also enables us to differentiate better between
them—but at the expense of losing valuable sound information. The
same is true, mutatis mutandis, of phonetic recoding. In both, boundar-
ies between categories can more reliably be discerned than within-category
information. One way to demonstrate this is via so-called categorial per-
ception. It has been experimentally established that listeners discriminate
the same intervals less reliably within categories than at their boundaries,
M or e I s Up — S o m e of t h e T i m e [ 195 ]
961
in both speech and music perception. In other words, those differences that
are required to distinguish between categories are more reliably discerned
than those that are not required. What is more, contrary to commonsense
expectation, professional musicians are less good at within-category dis-
tinctions than some musically naïve listeners. In natural noises there is no
categorial perception.
Categorial perception was long thought to belong only to speech
perception. In his Ph.D. dissertation, however, Mark J. Blechner (1977)
explored categorial perception in music. He used the categories “major”
and “minor chords.” A chord is a group of (typically three or more) notes
sounded together, as a basis of harmony. In a three-note group, the
two kinds of chord differ only in the pitch of the middle note. Blechner
divided the one interval between the two middle notes into nine equal
steps. In each chord, the high and low pitches were always 392 and 262
Hz, respectively. The central tone varied in discrete steps of 2.32 Hz,
from 311 to 329.6 Hz. (The complete set of stimuli is displayed graphi-
cally in Figure 8.3.) Blechner conducted two experiments with three
groups of subjects: professional musicians, nonprofessionals highly
skilled in music (NP-H), and nonprofessionals with low skill (NP-L). In
400
380
360
340
Hz
320
300
280
260
1 2 3 4 5 6 7 8 9
(prototype (prototype
minor) major)
Chord Number
one experiment two consecutive triads were played, and then a third
one repeated one of them. The subjects had to identify which one was
repeated. In the other experiment only two triads were played, either
two consecutive triads or one triad repeated. Subjects had to answer
“same” or “different.” The performance of NP-L subjects was random.
The performance of professional musicians was accurate at the bound-
ary between the two categories; otherwise it was quite poor—they were
subject to categorial perception. Only the results for the NP-H subjects
did not indicate categorial perception.
Of course, one could argue that all our perceptions and categories are
determined by our body or our brain. In other words, the “embodied mind”
may cover the total range of mental possibilities. This may be a sweeping
and valid generalization, but it is of doubtful usefulness. In this case, we
must distinguish different kinds of “embodied mind,” such that they not
only acknowledge an overall notion of the body’s interactions with the envi-
ronment and the ontological assumptions about the world that are built
into the body and the brain but also account for, for example, three different
modes of cognition that prevail one at the expense of the others, according
to immediate needs. If the minute changes in our environment are impor-
tant, we attend to the noises changing from second to second; if we wish
to remember certain stable relationships between sounds, as in a scale or a
melody, we translate them into a stable spatial template, at the expense of
the changing minute sounds that indicate moment-to-moment changes in
the environment. Alternatively, we might wish to translate them into stable
speech sounds. To demonstrate how the same noise can be heard as unde-
fined noise in one context and as a well-defined speech sound in another,
I created a sound file containing five stimuli: a series of noises produced
by clapping my hands, the word suck recorded from the Merriam-Webster
Collegiate Dictionary—Audio Edition, the same word with the [k]replaced by
a clapping sound, an amplified token of the genuine [k], and a token of the
clap excised from the doctored word.3 The same noise is heard as a click with
an elusive echo-like aura in the clap series and as an abrupt [k] in the word.
The present chapter considers evidence that behind the stable catego-
ries of speech and music we subliminally perceive inconstant sound infor-
mation regarding the physical stimulus that escapes categorization. It is
treated as a byproduct of the categories directly used for communication
and is perceived as the thickness, size, color, relative brightness, or pace of
the sound stimulus. As I said, these aspects do not serve communication
directly but are perceptual qualities of the stimulus.
M or e I s Up — S o m e of t h e T i m e [ 197 ]
9
81
CROSS-MODAL MAPPING
John Lyons extracts from localist thinkers an illuminating idea that can
easily account for this but, alas, without giving references (I strongly sus-
pect that these are his own conclusions from reading the localist linguists):
We live and move, normally, on the surface of the earth (rather than in water
or in the air); and we do so, again normally, in an upright position. This gives
us the means of identifying one of the dimensions in a three-dimensional
space; it also gives us a fixed zero-point at ground-level. Directionality in the
vertical dimension—i.e. the difference between upwards and downwards—is
established by our experience of the effects of the force of gravity, by the fact
that, normally, the sky is above us and the ground beneath us and by the asym-
metry of the human body in the vertical dimension. For these, and other rea-
sons, verticality is physically and psychologically the most salient of the spatial
dimensions. . . .
There are two horizontal dimensions, neither of which is fixed, in the way
that verticality is, by the force of gravity or anything comparable. . . . But [man]
is asymmetrical in one of the two horizontal dimensions, and symmetrical in the
other: i.e. he has a front and a back, and two symmetrical sides. He has his prin-
cipal organs of perception directed towards the region in front of him. . . . The
asymmetrical front-back dimension is less salient than the vertical direction,
but more salient than the symmetrical right-left dimension. (1977: 690–691)
The “fixed zero-point at ground-level” may explain why the upward direc-
tion is the unmarked vertical direction, whereas the downward direction,
beneath ground level (as in deep~shallow), is the marked one. What is
upward from ground level is more easily accessible to our senses than what
is downward from it. So, we should rephrase Lakoff’s slogan as more is up
in the unmarked vertical direction and more is down in the marked one.
When there is no explicit reason for preferring any one of them, we have
recourse to the unmarked one.
Thus Lyons creates a markedness scale of the directions of spatial tem-
plates in cross-modal mapping: vertical, front~back, and left~right. When
there is no specific reason to prefer any one of these directions, the least
marked (vertical direction) is chosen, as with music and temperature. Each
later item in this scale is chosen when there is good reason not to prefer the
preceding one(s). The left~right direction is sometimes simply arbitrary, as
in left-wing and right-wing politics. To be sure, the usage “left-wing and
right-wing politics” has good historical reasons, but it is arbitrary in the
sense that it “does not contain in itself the reason why it is so, and not
otherwise”—to use Coleridge’s phrase. If you ask whether the adjectives
in “high and low pitch” or “advanced and backward education policies”
M or e I s Up — S o m e of t h e T i m e [ 199 ]
02
could be reversed, you will most probably get a negative answer. If you ask
whether in “left-wing and right-wing politics” they could be reversed, you
will most probably get a positive answer.
Note, however, that such an analysis explains only why the vertical
direction is preferred as spatial template in cross-modal mapping. But to
account for the matching of the “faster” pole with precisely the “height”
pole in music, or the “more” pole with the “down” pole with reference to
the depth of the sea, you need some cognitive-developmental model like
the Clarks’, combined with the Structuralist notion of “markedness.” At the
same time, to understand why “high” sounds are “thin” and “low” sounds
are “thick” rather than the other way around, one must know a thing or two
about the structure of the physical stimulus.
There seems to be some deep affinity between Lakoff’s ideas and those
of the localists who preceded him by decades or centuries (depending on
whom you ask). Jean-Michel Fortis suggests that some “cognitive linguists
had little knowledge of the past of their own discipline (that is why authors
like Lakoff advertise some of their ideas as new and even as breaking away
from the bonds of tradition)” (2011: 10). But his localist origins did not
prevent him from accounting for the more is up principle using our expe-
rience with liquids in a container or for temperature “going up” using the
mercury in a thermometer oriented vertically.
When I first read Lakoff’s 2008 essay “The Neural Theory of Metaphor”
(see 2012), I hoped that this fresh angle on metaphor would help him over-
come some of the shortcomings of his theory. Later I discovered an ear-
lier article, “The Brain’s Concepts: The Role of the Sensory-Motor System
in Conceptual Knowledge,” by Vittorio Gallese and George Lakoff (2005).
Gallese is an eminent brain scientist, one of the group of scientists who
discovered mirror neurons.
In the first part of his 2008 essay Lakoff gives an outline of his Neural
Theory of Language developed with Jerome Feldman, based on Gallese’s
findings. This in itself is a masterpiece of conciseness and elaborateness
at the same time. In the second part of the essay Lakoff presents his “old
theory.” I will not go into the details of this essay, but only dwell on the
one point that interests us here. As I mentioned above, in this later essay
Lakoff falls back on his earlier theory of grounding: “In a thermometer
oriented vertically, the mercury goes up physically as the temperature
increases (metaphorically goes up).” This he explains as “More Is Up: Our
INCOMPLETE FOSSILIZATION?
I have suggested that the spatial conception of sound (“high” and “low”
sounds) is a convention of our culture but not an arbitrary one: It resulted
from a set of cognitive processes that fossilized into an automatic usage, a
pair of unqualified dictionary entries; when we use them, we don’t neces-
sarily have in mind the cognitive processes that generated them. In a later,
mind-expanding study, “Which Way Is More? Pitch Height, Parametric
Scales, and the Intricacies of Cross-Domain Magnitude Relationships”
(2013), Eitan points out a much more intricate state of affairs in this field,
almost a muddle: there are intuitions conflicting in an interesting way, and,
still, there seems to be method in this madness—to use Polonius’s words.
This would suggest that the cognitive processes involved have not entirely
fossilized.
In this essay, Eitan reviews literature on research on cross-domain map-
pings of pitch. He presents a host of intriguing findings, which he sum-
marizes as follows: “The deceptively simple world of registral pitch thus
M or e I s Up — S o m e of t h e T i m e [ 201 ]
0
2
reveals a truly Carrollian structure: one where ‘more’ and ‘less,’ growing
or shrinking, keep shifting and reversing places and roles, depending on
whether one stands still or moves, on whether one is going up or downhill,
and on what objects of comparison one may happen to observe on the way”
(2013).
In what follows, I will attempt to sort out just a few of those muddles,
hoping that the discussion will light the way to making sense of the other
muddles as well. I draw on the principle underlying this and the preceding
chapter: that both in verbal metaphors and in cross-domain mappings of
pitch, visual and auditory stimuli are treated as bundles of features, each
of which may suggest a mapping, sometimes incompatible with the oth-
ers. Changing circumstances may require us to attend to different features
of the same stimuli or even have recourse to different modes of cognitive
processing. Thus, for reasons suggested in the foregoing discussion, greater
frequency of sounds is matched with the “high” end of the vertical spa-
tial template, so that in this case more is up. Greater frequency entails a
smaller wavelength and a poorer overtone envelope; consequently, “higher”
sounds are perceived as thinner and smaller than “lower” sounds, so that
in this case less is up. At the same time, in the intensity dimension of
the same stimulus, louder sound is associated with greater visual size. In
both cases, the conflicting “visual” sizes are correlated with the size of the
wave: thinner, high sounds are correlated with shorter waves, and louder
sounds are perceived as visually larger—with wider waves. “Higher” (more
frequent) sounds suggest greater speed; this can be accounted for by the
faster vibrations involved. The ingredient shared by all these accounts is
that we perceive a phenomenological quality conveyed by sound informa-
tion of a different structure; it is excluded from consciousness, but we per-
ceive some of it subliminally.
So far I have summarized the various aspects of the sound stimulus
as I have discussed them above. But Eitan points out a radical difference
between static and dynamic sound stimuli, where the mapping of more and
less may be reversed:
Dynamic and static pitch stimuli associate with other domains in different,
sometimes contradictory ways. Consider again, for instance, the ubiquitous
association of pitch with physical size. As noted above, the association of higher
pitch registers with smaller size crosses age and cultural boundaries, affecting
music, speech, and expressive vocalization, and was even associated with the
vocal behavior of diverse non-human species. However, when subjects (of differ-
ent cultures) associate dynamic pitch patterns—ascending or descending melo-
dies or pitch glides, rather than high and low pitch register—with visual stimuli,
they map pitch rise into growing size, and pitch fall into shrinking size—just the
opposite of what static pitch mappings would imply. In a recent study by Kim
and Iwamiya, Japanese participants judged rising pitch patterns as congruent
with expanding (rather than shrinking) animated visual shapes. (2013: 9)
I submit that these differences between static and dynamic stimuli stem
from different modes of cognitive processing. Shape perception and direc-
tion perception are related to different brain centers. In static shape per-
ception we attend to the (unchanging) relationship between the parts of
an object. Regarding moving objects, spatial location becomes salient: we
ignore the relationship between its parts and attend to the changing rela-
tionship between the object and surrounding space. Movement and direc-
tion direct attention away from the internal structure to the relationship
with surrounding space. A sustained “high” pitch suggests a certain physi-
cal structure: short wavelength and relatively few overtones—perceived
as “small” or “thin.” Regarding “ascending,” we attend away from the
structure of the object and focus on its changing relationship to the
surrounding space.
The experiments use explicit demonstrations (in audiovisual matching
or metaphor rating for music and speech sounds) and implicit demonstra-
tions, through a Stroop-like effect in which processing is faster for congru-
ent pitch and visual movement;4 for example, participants judged rising
pitch patterns as congruent with expanding (rather than shrinking) ani-
mated visual shapes, or, when presented with an octave leap, they tended to
term the higher pitch “big” and the lower pitch “small,” in contrast to results
for isolated pitches. In such experiments the experimental design deter-
mines whether pitch change is associated with one moving point or with a
continuously changing shape. What is common to both is that attention is
shifted away from the structure of the wave to movement through space. As
I have argued all along, musical pitch has two coinciding physical correlates,
frequency and wavelength. Accordingly, the same sound may be described
as “high” (spatial location), suggesting a greater extent than a “low” sound,
or as “thin” (perceived size), suggesting a smaller extent than “thick.” When
switching from static to dynamic percepts, we change cognitive processing
modes: we shift attention from perceived size to spatial location.
Eitan explains the finding that in a dynamic context subjects map pitch
rise to growing size and pitch fall to shrinking size as follows: “For dynamic
stimuli the more abstract (and perhaps language-based) analogy between
M or e I s Up — S o m e of t h e T i m e [ 203 ]
4
0
2
pitch ‘rise’ and physical growth may take hold” (2013: 10). If we assume
that we are dealing with sound recoded into a spatial template, as suggested
above, we do not need an “abstract (and perhaps language-based) analogy.”
We simply shift from one mode of cognitive processing to another and
attend to different kinds of relationships in the situation.
As we have seen, Lyons suggests “a fixed zero-point at ground-level” in
the vertical dimension in spatial relationships. In a dynamic context we
shift attention from wave structure to motion away from the ground level;
this, in turn, suggests expanding, whereas moving back to ground level
suggests shrinking distance. Again, the conflicting mappings of the static
and dynamic stimuli result from switching from one aspect to another. The
switch is prompted by the changing needs of the task. This explanation dif-
fers from Eitan’s, since it does not assume linguistic mediation but, rather,
attending to different aspects of the same physical stimulus.
Or consider the following: “Pitch space is directionally asymmetri-
cal: pitch descent is not the opposite of pitch ascent. The two directions of
pitch change may associate with different, not opposite dimensions. While
pitch descent, for instance, implies spatial descent, pitch ascent does not
imply spatial ascent, but rather acceleration” (Eitan 2013: 15). To some read-
ers this sounds puzzling, but Eitan relies on solid experimental evidence.
In my opinion, this may be the truth but not the whole truth—which I will
try to sort out in what follows.
Pitch ascent implying acceleration would be fully consistent with the
foregoing analysis. As we have seen, high pitches are perceived as fast
because vibrations in the higher registers are faster than in the lower regis-
ters. Now, when you have a gradually rising pitch, vibrations become gradu-
ally faster—hence the sense of acceleration.
On the surface, this conforms to my analysis above. In “acceleration” we
focus on (changing) perceived speed; in “spatial descent” we attend away,
for some reason, to spatial location. However, ascending or descending
pitch may imply many things, among them acceleration or spatial ascent
for ascending pitch and spatial descent or coming to a rest for descend-
ing pitch. Moreover, a falling pitch may suggest stability or confidence,
whereas a rising pitch may suggest instability, lack of confidence, or, when
loaded with energy, defiance. In an interrogative sentence it may simply
imply uncertainty. Thus, rising pitch can imply various degrees of instabil-
ity. In music, pitch falling to the tonic, or in speech, intonation falling to
the zero point (at the “ground level”), indicates “coming home,” coming to
a rest. In John Ohala’s (1994) experiments with stripped speech (where
only the intonation contour can be heard, and the words cannot be dis-
cerned), a long-falling intonation contour suggests dominance. Again in
M or e I s Up — S o m e of t h e T i m e [ 205 ]
62
0
nature would be “to advantage dress’d”; with the last one, to disadvan-
tage. You may also create a witty metaphor such as “She has the legs of a
gazelle—not as graceful, not as nimble, not as slender, but as hairy.” Such
a conception may account for the incompatible results produced by the
experiments reported by Eitan and may also suggest that the perception
of sounds and metaphors is governed by a homogeneous set of activities.
SUMMARY
In our culture, rapid sound vibrations are called “high”; slow vibrations,
“low.” In other cultures very different terms are applied to the same distinc-
tion. In this sense, this usage is a convention. It is, however, not entirely
arbitrary; it seems to originate in some active cognitive processes that have
fossilized into rigid dictionary entries. Lakoff attempts to locate these
cognitive processes with the help of his theory of “conceptual metaphor”
and principle of “grounding,” what Roger Brown calls “mediated associa-
tion.” In their impressive intercultural study of this dichotomy, Eitan and
Timmers invoked Lakoff’s theory, epitomized as more is up; but time and
again they came up against the same puzzling findings. “More is up” when
physical space is concerned but not necessarily when pitch is concerned.
Thus, for instance, sounds of “higher” pitch are “thinner,” not “thicker.”
This finding, they think, does not conform to the conceptual-metaphor
theory they espouse. They assume that there is some hidden conflict
between two or more conceptual metaphors. I argued here that “more is
up” only when you discuss upward directions, but in such pairs of adjec-
tives as deep~shallow, more is down and less is up. More importantly,
Eitan and Timmers’s puzzling findings cease to be puzzling if you do not
try to account for them within a theory of conceptual metaphor. Far from
being the result of a hidden conflict between two conceptual metaphors,
they conspicuously conform to the “cognitive-fossils-and-constraints”
approach proposed here, which relies on the feature-deletion theory of
metaphor, cognitive-developmental evidence, the limitations of the
human ear, alternative modes of cognitive processing, and an accurate
physical description of the auditory stimulus. Admittedly, this is less par-
simonious than simply more is up, but it has the advantage of account-
ing for the phenomena discussed. The conceptual-metaphor theory allows
you to point out that many metaphors may mean the same thing; the
feature-deletion theory allows you to point out that one metaphor may
mean many, sometimes unforeseen, things.
M or e I s Up — S o m e of t h e T i m e [ 207 ]
8
0
2
9
0
2
CHAPTER 9
The iambic is the characteristic rhythm of people as they talk. . . . The trochaic rhythm,
again, is too much akin to the comic dance, as may be seen in tetrameter verse, for the
rhythm of tetrameters is light and tripping.
—Aristotle, Rhetoric, 1408b
His rage armed Archilochus with his iambic: comedy and tragedy have adopted it, as being
natural for dialogue, able to drown out the noise of the audience and suited to action.
—Horace, The Art of Poetry
I n chapter 1 I said that this book is devoted to the question of how cogni-
tive processes shape and constrain cultural and literary forms and that
conventions can be used without any awareness of the cognitive processes
that shaped and constrained them or of the perceptions and experiences
related to those processes. In this chapter I will discuss a pair of metrical
conventions of major importance, the iambic and trochaic meters, and the
cognitive processes that shape and constrain their effect.1
1. Much of this chapter was written in 1971 and was published as c hapter 3 of my
1977 A Perception-Oriented Theory of Metre. Recently I realized that metrists are still
puzzled as to the difference between the iambic and the trochaic and the special status
of the iambic, which was pointed out as early as Aristotle and Horace. Since Tsur 1977b
is inaccessible today, I decided to republish this chapter as it appeared in that book,
with some updating and minor corrections.
021
1 2 3 4 5 6 7 8 9 10
(1) shall behóld Gód, and néver tást déaths wóe.
w s w s w s w s w s
THE TROCHAIC
The peculiar nature of the trochaic meter has frequently attracted atten-
tion. Time and time again, critics have felt the need to account for this
peculiarity. In what follows, I add yet another, cognitive item to the list of
would-be explanations and hope to demonstrate its explanatory force. “It
is interesting to note,” Chatman writes, “that lighter syllables seem more
readily ictic in trochaic than in iambic feet”: “The trochaic mode more easily
violates normal prose accentual patterns; it quite insists on dominating the
rhythm. Iambic verse seems not to exert its will so rigidly” (1965: 140–141).
As an explanation for this peculiar character, Chatman quotes Halpern and
then offers to substitute his own explanation:
Halpern’s thesis is that trochaic verse, along with anapestic and dactylic,
is a subspecies of the native Germanic strong-stress verse, which he feels is
both isoaccentual and isochronic. . . . I think his observations about the rela-
tive inflexibility of the trochaic verse are correct, but would suggest another
T r o c h e e s a n d I a m b s a n d T h e i r R e l at i o n s h i p t o O t h e r M e t e r s [ 211 ]
21
cause, namely the comparatively short history of the mode. The sophisticated
smoothness of the iambic verse has been long in developing; trochaic verse,
however, was not taken very seriously in England until the nineteenth century.
(1965: 141n)
this asymmetry between iambic and trochaic lines becomes clear once it is
realized that trochaic verses have stress maxima only at odd positions in the
verse and that an initial iambic foot locates a stress maximum on the second
(i.e., on the even) position in the line, in direct violation of the trochaic prin-
ciple. (1966:199)
This kind of artificiality is absent from Blake’s “The Tyger,” for instance,
“What the anvil, what the chain,” or any other line. This is only one of
the many sources of difference. In actual speech (and according to the
T r o c h e e s a n d I a m b s a n d T h e i r R e l at i o n s h i p t o O t h e r M e t e r s [ 213 ]
4
21
it as an adverbial (in the sense of “no more than”), it bears lexical stress
and must be duly stressed in performance, neutralizing the alleged stress
maximum in the preceding w position (“To líve ′s bút an émpty dréam”). In
s w s w s w s
positions II and III, back to back, we get two stressed syllables, and the
line becomes similar to some perfectly legitimate trochaic lines, as we shall
see. At any rate, the line will be “metrical” under the stress maxima theory
itself. The question is thus not whether a trochaic line can bear “inversion”
of its first foot but, rather, whether but can bear full lexical stress. Hence
this example does not prove that trochaic lines do not tolerate inversion of
their first foot.
Beaver’s transcription “Between them straight ran the pathway” is not
much more felicitous. Between them does not exactly yield a stress maxi-
mum, as Beaver would like to indicate, nor does them straight realize exactly
a trochaic foot (if anything, it has an iambic rise). As a matter of fact, the
first foot of a trochaic line can be “inverted.” To be sure, it is not as frequent
or as easy as in an iambic line. It demands the overarticulation and over-
stressing of the deviant stressed syllable and very emphatic grouping of
the first two feet in order to make it acceptable. But the decisive condition
is, as in the iambic line, that the metric pattern should not be violated in
the second foot. If the first foot of a trochaic line is “inverted” and in the
second the trochaic is genuinely confirmed, there are, back to back, two
stressed syllables, and no question of a stress maximum in the second posi-
tion will arise.
Consider the following three examples from two poems notorious for
their regular rhythms:
The reader may perform these lines rhythmically by grouping the first three
syllables tightly together. In the third position, the stresses of the trochaic
pattern and the linguistic stress pattern emphatically coincide. Thus, the
vigorous, dominant character of the trochee is not the result of its resis-
tance to the “inversion” of its first foot but, rather, its cause. Similarly, Halle
and Keyser are victims of their own mistaken assumption when they sug-
gest the phrase “On first looking into Chapman’s Homer” as an example
of an unmetrical utterance. Suppose we encounter it in an emphatically
trochaic context:
"Áilsa Róck , " " When I have féars, " "To Fánny,"
s w s w s w s w s w
Since most trochaic verse in English is in short lines, and since our impressions
of iambic verse are derived almost entirely from pentameter, it would seem
entirely possible that the issue has been falsely formulated—that the differ-
ences of rhythm encountered are attributed not to the type of foot but to the
length of line in which the foot characteristically appears. And it will be argued
T r o c h e e s a n d I a m b s a n d T h e i r R e l at i o n s h i p t o O t h e r M e t e r s [ 215 ]
621
below that the more regular beat of short-lined verse is accounted for by the
fact that a much higher percentage of positions available for stress maxima are
occupied than is the case in decasyllabic verse. (1968: 314)
And even where reversals are indicated the reader has strong desires to “wrench”
the stress or accent, as in Browning’s:
What there’s nothing in the/mōon note-/wōrthy? (Chatman 1965: 141)
meter the first segment will begin and end with a weak position and the
second segment, with a strong position. In the trochaic meter the converse
will be the case. In the eight-syllable-long line, the caesura will divide it
into segments of equal length (4 + 4) and equal structure: In the iambic
meter both segments begin with a weak position and end with a strong
position; in the trochaic, vice versa. As opposed to the hexameter, in which
the second hemistich also exactly repeats the structure of the first one,
the iambic or trochaic tetrameter does not exceed the limits of short-term
memory: namely, the magical number 7 ± 2. Thus the tetrameter has a
stronger, more rigidly symmetrical shape than the pentameter or the hex-
ameter; that is, it is of a more compelling nature, which more forcefully
determines the character of its parts.
Trochaic meter and short-lined verse reinforce this compelling quality
in each other. But we still have no explanation why the trochaic should
be more compelling than the iambic. Curiously enough, the clue is quoted
in Chatman’s (1965: 26–27) book as well, but he does not connect it to
this specific problem. One would imagine that in an endless series of
equidistant tick-tacks no preference would be given to iambic or trochaic
rhythms; the only distinction would be whether the series began with an
upbeat or a downbeat. Experimental psychology, however, shows that
this is not so. H. Woodrow found in his tick-tack experiments, back in
the 1920s, that in a series 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” (1951: 1233): “Intensity has a group-beginning effect: duration, a
group-ending effect: pitch, neither a group-ending nor a group-beginning
effect” (1911: 77).
Both Chatman and Meyer invoke Woodrow’s experiments; the latter
also indicates at some length their implications for a variety of metric feet:
When time intervals are equal, and every second sound is accented, the rhythm
will appear to be trochaic. If intervals are equal and every third sound is accented,
the rhythm will appear as a dactyl. Thus the trochee and dactyl may be grouped
together in the sense that both are primarily products of intensity differences
rather than durational differences.
Just the opposite is the case with iambic and anapestic rhythms. They are
basically products of durational differences. If we start with a trochaic rhythm
and gradually increase the interval after the louder sound, we arrive at an iambic
rhythm. Similarly, if we begin with a dactylic rhythm and gradually lengthen the
interval after the louder sound, the rhythm tends to become an anapest. Thus,
T r o c h e e s a n d I a m b s a n d T h e i r R e l at i o n s h i p t o O t h e r M e t e r s [ 217 ]
8
21
the greater the relative duration of one tone or beat of a group, the greater the
tendency for it to complete the group, while the greater the relative intensity
of a beat, the greater the tendency for it to begin the rhythmic group. In other
words, durational differences tend to result in “end-accented rhythms,” and
intensity differences tend to result in “beginning-accented rhythms.” (Meyer
1956: 106–107)
Bite in line 4 falls in a strong (odd) position, and it is duly stressed. In line 2
it falls in a weak (even) position and should create metrical complexity. But
the reader prefers to suppress it and, rather, stress the preceding auxiliary
verb (do), which is unemphatic and is introduced mainly for the meter’s
sake. This tendency toward regularity can be reinforced by the repetition
of some notoriously trochaic word or sound pattern in the first line, for
example, “Handy spandy Jack-a-Dandy,” “Tackle, tackle, Mother Goose,” or
T r o c h e e s a n d I a m b s a n d T h e i r R e l at i o n s h i p t o O t h e r M e t e r s [ 219 ]
0
2
“Tyger, tyger, burning bright.” Such a sound pattern will tend to perpetuate
its trochaic cadence in the subsequent lines.
A glance at Henry Carey’s parody on Ambrose Philips’s overly smooth
style (in The Oxford Book of Nursery Rhyme) is revealing:
The pace for both desires is set in the first line. The meter of this line is as
regular as one could expect in a trochaic (with, probably, the need to sup-
press a possible stress on there, used interjectionally). The metrical pattern
is reinforced by the stress pattern coinciding with it, which, in turn, is rein-
forced by a recurring sound pattern: not the simple kind such as “Tackle,
tackle” or “Tyger, tyger” but, rather, a very sophisticated version of the type
“burning bright.” The contrast between the s and w position is heightened
by a sophisticated string of alliterations, emphasizing the s positions, cul-
minating in horr-. It repeats the rs of the first s position, the o of the second,
and the h of the third. Go (preceded by an r) repeats the sounds of GR-R-R,
which is an important keyword. As a nonsense word, it fulfills the function
of nonsense words at the beginning of nursery rhymes: to draw attention
to the sound pattern. As an emotive or onomatopoeic word, it arouses a
strong desire to preserve the rhythms of colloquial language. The first word
of the second line (Water) confirms the trochaic pattern. The next stress,
however, is displaced to a w position. The strong desire “to wrench” the
prose stress here is vigorously counteracted by the colloquial character of
damned. The only option left for readers is to accommodate both patterns
T r o c h e e s a n d I a m b s a n d T h e i r R e l at i o n s h i p t o O t h e r M e t e r s [ 221 ]
2
2. In Tsur 1972 I pointed out that one of Browning’s favorite prosodic devices is to
insert a marked syntactic break before the last syllable of the line. This arouses a strong
desire to complete the verse. So, the last word is highly required and bestows a more
than usually closed shape on the line. This is precisely what happens here at the end of
the second line, with do, gathering momentum for the reader’s (frustrated) desire to
maintain the set of regularly alternating s and w positions. Another keyword is abhor-
rence. It not only contains, as we have seen, the phonetically crucial syllable -horr-; it
also forms an unusually witty “feminine” rhyme with Lawrence, having five(!) sounds
in common, with virtually no common ingredients of meaning, thus emphatically rein-
forcing the quality generated by the trochee (the “antigrammatical” rhyme do–you has
a similar effect).
3. According to the law of Good Continuation, the arrangement of the stimulus pat-
tern that makes the fewest changes or interruptions in it will be preferred as “good
gestalt.”
each stanza. So, we must decide which is more important: the name of the
metric structure or its perceived character.
Beaver writes: “As a matter of fact, many poets have capitalized on this
ambiguity. In ‘To a Skylark,’ Shelley chooses a stanza form which maintains
the ambiguity through the first four lines of each stanza, resolving it in
the iambic hexameter fifth line” (1968: 317). Let us have a look at the first
stanza of the poem:
In fact, only two of the lines (2, 4) can be said to be “ambiguous.” Lines 1 and
3 are unambiguously trochaic, and line 5 is unambiguously iambic. Putting
aside the problem of the fifth line for the time being, one may observe that
the reader of French, German, Russian, Hungarian, and Hebrew poetry
(and possibly that of many other languages) is familiar with the alternation
of “feminine” and “masculine” rhymes that varies rhythm and governs line
grouping without basically changing the meter (see Tsur 2013). In English,
it is more difficult to do this because of the scarcity of “feminine” rhymes.
Everything that happens here is thus an extension of the grouping
principle of performance when irregularities occur. The dissimilarity of
the odd-numbered and even-numbered lines upsets their balance to some
extent, and they demand some further grouping. By grouping line 1 with 2,
and line 3 with 4, the four-line unit is divided into two corresponding
halves, reinforcing the grouping effect of the rhyme scheme. (This is, pre-
cisely, with the necessary changes, what happens in the ballad stanza too.
See discussion of the ballad “Edward,” c hapter 4.) The asymmetry of the
tetrameter and trimeter lines requires grouping them into two symmetri-
cal groups, thus enhancing the unity and simplicity of the stanza.
I am inclined, then, to disagree with Beaver’s more general “solution” to
poems such as Tennyson’s “Lockley Hall”: “An obvious solution to this prob-
lem is to view such poems or portions of poems as metrically ambiguous
in their surface structure, and postulate that they are, in their deep struc-
ture, either iambic with initial position always unoccupied, or trochaic,
with final position always unoccupied” (1968: 317). The problem will not
arise at all if we realize that the obligatory caesura divides the lines of this
poem, or “A Toccata of Galuppi’s,” so that their metric structure is exactly
the same as that of the Browning poem discussed above. The typographical
T r o c h e e s a n d I a m b s a n d T h e i r R e l at i o n s h i p t o O t h e r M e t e r s [ 223 ]
4
2
(12a) Comrades, leave me here a little, while as yet ’tis early morn:
Leave me here, and when you want me, sound upon the bugle-horn.
(12b) Comrades, leave me here a little,
While as yet ’tis early morn:
Leave me here, and when you want me,
Sound upon the bugle-horn.
When preparing this chapter, I felt that I had to modify in one impor-
tant respect earlier positions I took in Tsur 1977b. While the alternation
of feminine and masculine endings enhances grouping and articulateness
in both (13a) and (13b), we do experience the two excerpts differently. In
vocal or subvocal performance, pause and falling intonation are the most
conspicuous cues for discontinuity. In (13b), these cues are equally arrest-
ing at the end of each line. As a result, one typically tends to experience fast
alternations of short perceptual units. In (13a), when some of these dis-
continuities occur in midline at the caesura, the cues are toned down, so as
to subordinate the first hemistich to the whole line. As a result, a broader,
slower rhythm is obtained, suggesting a more serious or dignified attitude.
The “unexpected” appearance of the iambic hexameter at the end of
Shelley’s stanza (11) resolves, in fact, nothing. Rather, it complicates things
further. It does not reveal the “deep structure” of the preceding lines; it devi-
ates from them. According to Barbara Herrnstein Smith (1968), a closural
effect can be achieved by deviating from a previously established structural
principle. This would imply a hierarchy of grouping principles operative
in the stanza: (1) the line, (2) two pairs of lines parallel to each other, and
(3) a fifth line deviating in length and foot, emphatically sealing the stanza.
Whether it actually does so depends on whether the reader realizes the hier-
archy of grouping principles or loses sight of this third, widest one.
Wimsatt and Beardsley seem to be after a far more deeply seated quality
when they remark on the first line of this Shelley poem:
The unquestionably iambic movement following the very strong first syllable
might, if we were desperate, be accounted for by saying that the word “Hail”
breaks into two syllables, “Hay-ul,” with a resultant needed extra syllable and
the familiar opening pattern of iambic inversion. But a much more energetic and
irrefutable assertion of the iamb appears to be in the progressive rise or stress
increase of the three syllables “thée, blit̋he, spír̋it.” (1959: 593–594)
This argument seems invalid to me. Even if we take for granted, for the sake
of the present argument, four degrees of stress in “to thée, blit̋he spír̋it,”
the only thing we need to do to refute this conception of “rising” rhythm
is to say that the energetic iambic rise is followed by a sudden trochaic
fall and that in every trochaic line the final falling movement is preceded
by a rising movement. In fact, the term iambic movement is used here in a
figurative sense: it does not refer to the recurring metrical pattern of feet
but, rather, to the overall rising pattern of performance. In a note, Wimsatt
and Beardsley add: “But many such lines, like the one from Shelley’s
‘Skylark’ . . . , can be shown in one way or another to be in fact iambic. The
shape of the phrases is likely to have much to do with it” (1959: 594).
The line has been suggested, but not shown, to be iambic. Formally, the
line is nothing but trochaic. But what is it that impresses the critics as so
“untrochaic” in this line? It is its complexity. Blithe consists of one stressed
syllable with a diphthong as its nucleus. Neither its stress nor its length
can be sufficiently reduced and fit into a weak position so as to equal to in
the preceding weak position. Thee in the preceding strong position bears no
lexical stress. The regular contrast of intensity, as well as the equal duration
of feet, is upset; this makes the line more like iambic. The stressed syllable
in a weak position can be accommodated only by additional grouping; this
was, indeed, done by Wimsatt and Beardsley, and this makes the line too
complex for an ordinary trochee. But now that we know what is so “untro-
chaic” about the line, we need worry no more, and there seems to be no
reason why we should not continue calling it “trochaic.”
IAMBIC TENSION
The iambic is usually thought to be the most natural meter, the nearest
to ordinary speech rhythms. A striking feature of this measure is that in
languages as different from English as Hebrew and Hungarian, poets use it
T r o c h e e s a n d I a m b s a n d T h e i r R e l at i o n s h i p t o O t h e r M e t e r s [ 225 ]
6
2
in a manner very similar to English blank verse. Note that this is particu-
larly noteworthy in Hungarian, where the end-accented nature of the iamb
generated by duration differences overrides the fact that in Hungarian the
first syllable of a word is invariably stressed. This seems to hold true even
outside the syllabotonic system, as Aristotle and Horace may bear witness,
for whom iambic meant a “quantitative” measure, alternating longer and
shorter syllables:
The iambic is the characteristic rhythm of people as they talk. . . . The tro-
chaic rhythm, again, is too much akin to the comic dance, as may be seen in
tetrameter verse, for the rhythm of tetrameters is light and tripping. (Aristotle
1932: 3.8, 1408b)
His rage armed Archilochus with his iambic: comedy and tragedy have
adopted it, as being natural for dialogue, able to drown out the noise of the audi-
ence and suited to action. (Horace 1951: 117)
It is difficult to know now exactly what Horace meant by the last two
phrases (“able to drown out the noise of the audience and suited to action”),
but the fact that the iambic was found to be “natural for dialogue” remains.
As we have seen, Aristotle agrees: “The iambic is the characteristic rhythm
of people as they talk.” Our explanation of its nature will have to take this
fact into account.
At the end of his “ ‘Prose Rhythm’ and Metre,” Roger Fowler comes to
the following conclusion: “This, paradoxically, may help to explain why the
iambic measure is felt to be suited to English: not because its pattern cor-
responds 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). Surprising as this may be, the great vari-
ety of languages to which the iambic measure seems to be “suited” (even
Hungarian!) supports this conclusion. I would, however, prefer tolerates to
necessitates. Verbal necessity alone cannot account for syncopation; it can-
not explain why Pope should resort to syncopation less frequently than
Milton (as Fowler himself asserts) or why Shakespeare, in his earlier work,
should resort to it less frequently than in his later work. It seems, rather,
that syncopation is related to wider issues and that it takes deliberate dar-
ing to abandon the security of established strong shapes. Far from being
a necessity, syncopation is a powerful achievement; it is a delicate balance
between prose rhythm and meter. As Fowler states, “There can be a situa-
tion (Hopkins’s ‘counterpoint’) where the prose rhythm makes itself felt as
something playing against the ostensible meter. But extreme lack of fit of
words with feet results in the total assertion of prose rhythm” (1966: 94);
and “in Donne, who, according to Chatman, has a much higher propor-
tion of reversed feet [than Pope], the tension is less because the metrical
pattern is obscured” (1966: 93). In trying to achieve ambiguous, complex
rhythms, there is always a possibility that the meter will be lost sight of,
which is a risk some poets seem to be unwilling to run.
I submit, then, the following two explanations for Fowler’s generaliza-
tion about the iambic. First, as we have seen, end-accented meters allow
for greater flexibility in manipulating the time factor in performance than
beginning-accented meters. Second, bisyllabic feet have “stronger gestalt”
than trisyllabic feet; they can be more effectively maintained when con-
flicting with prose rhythm. When performing trisyllabic feet, the reader
is inclined to entirely suppress prose rhythm when it conflicts with prose
rhythm; otherwise the meter of the poem is in danger of disintegration.
(This may probably account for the fact that in “quantitative” measures as
well, the iambic was found to be most “natural for dialogue.”)
TERNARY METERS
We have seen that critics are inclined to link the general character of ter-
nary meters with that of the trochaic. We are now in a position to point
out the source of their common character. They resemble each other in that
they are all unlike the iambic. We have seen that the iambic is “natural for
dialogue,” not in that its pattern is similar to the stress patterns of ordinary
speech but in that it allows the greatest tension between prose rhythm and
meter. It has a stronger gestalt than the ternary meters, and therefore, the
feel of it can be preserved even when the lexical stress pattern strongly
deviates from it. In order to preserve the integrity of the larger feet of three
syllables, one has to suppress prose rhythm whenever it deviates from a
ternary foot. This is the common factor to ternary meters and the trochaic.
We have seen that in the trochaic too, although for different reasons, in
order to preserve its peculiar character, one feels inclined to suppress in
prose rhythm whatever deviates from perfect regularity. The greater regu-
larity of ternary meters and the trochaic renders them of a more compel-
ling, though less complex nature.
Finally, I wish to clarify the negative relationship between the iambic
and the ternary meters by briefly commenting upon prevailing practice
in some languages other than English. Unlike in English, where ternary
T r o c h e e s a n d I a m b s a n d T h e i r R e l at i o n s h i p t o O t h e r M e t e r s [ 227 ]
8
2
4. During World War II and several decades afterward, one of the greatest Hebrew
modernist poets, Nathan Alterman, published a weekly column commenting on cur-
rent events. He used the prosody of his serious poetry.
CONCLUSION
T r o c h e e s a n d I a m b s a n d T h e i r R e l at i o n s h i p t o O t h e r M e t e r s [ 229 ]
0
32
231
CHAPTER 10
(a) v voiced
b
m
p
f unvoiced
(b) voiced
n d
t θ unvoiced
1. Kenneth Burke notes the same linguistic mechanism in the following para-
graph: “ ‘In Xanadu did Kubla Khan’ is found, by reason of the cognate relationship
between n and d, to be much more closely knit, on the phonetic basis, than would oth-
erwise be supposed. One might make this apparent by imagining himself pronouncing
the line with a head cold, thus: ‘Id Xadadu did Kubla Khad’ ” (1957: 300). The speech
error “Lemadon” would suggest that Burke is underscoring a phenomenon that does
have psychological reality.
defined by linguists are not merely theoretical constructs but real units
with an independent existence in mental grammar. In my view this is the
first step in a much more complex process.
Fromkin collected a great variety of speech errors. I will concentrate only
on those errors in which some symmetrical exchange of linguistic units
takes place. Some of these errors are called spoonerisms, that is, the trans-
position of initial or other sounds of words, usually by accident, attributed
to the Oxford don the Reverend W. A. Spooner, as in “a blushing crow” for
“a crushing blow,” “our queer old dean” for “our dear old queen,” or, chiding
one of his students, “You have hissed all my mystery lectures. I saw you
fight a liar in the back quad; indeed, you have tasted the whole worm.” As
we shall see below, there may also be switches of whole words at the one
end and, as we have seen above, even symmetrical switches of distinctive
features at the other.
Most genuine speech errors of this kind result in nonsense words.
Meaningless speech errors are, no doubt, funny. Bergson defined the
comic as “something mechanical encrusted on the living” (Bergson n.d.).
The ensuing nonsensical speech errors collected by Fromkin are, in fact,
a highly extreme case of this: the initial speech sounds of words are
mechanically switched, irrespective of meaning. Since the shift results
in a meaningless utterance, there is little more in it than “something
mechanical encrusted on the living.” The observer, in turn, may have—in
Muecke’s (1970: 36–37) words—a feeling of superiority, freedom, and
amusement in perceiving the failure of the other. And when the switch
does yield a meaningful expression, the meaning may be anomalous, as in
“a blushing crow.”
Molière, for instance, employed mechanical reversal of phrases to comic
effect. Consider the following dialogue in his L’Avare (The Miser):
This process is governed by the principle of the jigsaw puzzle. The shift of
letters or distinctive features undermines the meaning and generates cha-
otic utterances just as in the above speech errors. But then, in a split sec-
ond, the changed parts are restructured, so as to fit into a new whole. In
jigsaw puzzles there are sometimes thousands of pieces forming a chaotic
jumble; with sustained, patient work one can put the pieces together into
a coherent whole. This gives a sense of satisfaction caused by the transi-
tion from chaos to order and coherence. In spoonerisms and slips of the
tongue the transition from chaos to coherence occurs in split seconds
and appears to consciousness as an insight. Köhler’s discussion of insight
seems apropos here:
Those European psychologists, myself once included, sometimes went a bit too
far. Very much impressed by the essential rôle of insight in productive thinking,
they often said that the solution of problems is brought about by insight—as
though nothing else counted. Now this statement is not entirely correct for
the following reason. Insight is insight into relations that emerge when cer-
tain parts of a situation are inspected. . . . In the solution of a problem . . . we
suddenly become aware of new relations, but these new relations appear only
after we have mentally changed, amplified, or restructured the given material.
(1972: 152–153)
According to this analysis, what we usually call insight is the unique con-
scious quality that suddenly emerges when the objects of mentation are
restructured. In terms employed in cognitive poetics, it is not the cause but
the perceived quality of this sudden emergence. The jigsaw-puzzle effect
thus takes an interesting twist. Köhler (1972: 163) refers to the three Bs,
namely, “the Bus, the Bath, and the Bed,” where some of the greatest sci-
entific discoveries have been made (recall Archimedes and Kékulé!2). As for
the various insights reached in this way, “they all agree on one point. After
periods during which one has actively tried to solve a problem, but has not
succeeded, the sudden right organization of the situation, and with it the
solution, tend to occur at moments of extreme mental passivity” (Köhler
1972: 160).
The solution suddenly occurs at a time when, in Lowes’s phrase, “all con-
scious imaginative control is for some reason in abeyance” (Lowes 1927).
This is when a restructuring of the situation can take place. In piecing
together a jigsaw puzzle the element of suddenness is missing. Insight is
also preceded by a long period of sustained effort, but the sudden occurrence
of the solution renders it more dramatic and creates something new. In an
important way, this is exactly what happens in spoonerisms and slips of
the tongue as well. They suddenly restructure aspects of both the phonetic
component and the meaning of the utterance into a new whole. That is how
spoonerism and related phenomena become verbal imitations of insight.
One might add that in Freudian slips of the tongue, at least, long-repressed
or conflicting wishes emerge when “conscious . . . control is for some reason
in abeyance.” If such switches of linguistic units (ranging from distinctive
features, through phonemes, through clusters of phonemes, through mor-
phemes, to whole words) are embedded in a poetic text, they can amplify
and sometimes generate the poetic effects of the text.
2. “There I sat and wrote my Lehrbuch, but it did not proceed well, my mind was else-
where. I turned the chair to the fireplace and fell half asleep. Again the atoms gamboled
before my eyes. Smaller groups this time kept modestly to the background. My mind’s
eyes, trained by visions of a similar kind, now distinguished larger formations of vari-
ous shapes. Long rows, in many ways more densely joined; everything in movement,
winding and turning like snakes. And look, what was that? One snake grabbed its own
tail, and mockingly the shape whirled before my eyes. As if struck by lightning I awoke.
This time again I spent the rest of the night working out the consequences” (http://
www.famousscientists.org/friedrich-august-kekule/).
The first couplet rhymes [+ noun] with [+ noun] involving the same affix,
whereas the latter two rhyme [± verb]. Jakobson termed such these rhymes
“grammatical” and “antigrammatical,” respectively. Wimsatt described
them as “tame” and “vigorous” rhymes. I have discussed this issue elsewhere
(see, e.g., Tsur 2008a: 277–278). Wimsatt only refers to different parts of
speech, but I have shown that the reversal of the features [± concrete]
and [± animate] can be very effective in generating vigorous rhymes. Much
of this musical’s wit can be attributed to the vigorous, antigrammatical
Les sanglots longs
Des violons
De l’Automne,
Blessent mon coeur
D’une langueur
Monotonne.
[The long sobbings
of Autumn’s
violins
wound my heart
with a languor
monotone.]. (my translation)
Here I draw attention to one point that rarely receives attention, which
makes use of a dynamic discussed above in relation to speech errors. As
I have pointed out (Tsur 1992a), the vowel of the first syllable of sanglots
([sã]) in the first line is nasal, and the second syllable is oral. The last word
of the line repeats the preceding syllable ([lõ]), but the [+ nasal] feature
from the second-to-last syllable is transferred to it.
In the last stanza there is another (rather complex) poetic spoonerism:
Et je m’en vais
Au vent mauvais
[And I am walking
In the bad wind]
The word m’en in the foregoing excerpt is involved in three sound patterns.
First, in the sequence “m’en v–” the two labials [m]and [v] colliterate: they
are contrasted by the features [± nasal] and [± fricative]. Second,
“vent m–” in the next line repeats this pattern, except that the features
[+ nasal] and [+ fricative] are reversed. Furthermore, m’en and vent are
near-homonyms, in that only the same two distinctive features have been
reversed in them. Third, the two rhyming pairs consist of the sequence
“m—vais.” In the first member of the rhyme pair, m’en vais, the dash stands
for a nasal vowel, whereas in the second rhyme fellow, mauvais, the feature
[+ nasal] is deleted. In this way, the activity of the sound texture is ampli-
fied at a more fine-grained level than on the phoneme level; by the same
token, its diffuse quality is enhanced, so that not only the sound structure
and will not, therefore, account for the appearance of such SS words as sex-
tet and sexton” (Brown 1970: 293).
In the tot case the first retrieval must include a card with the definition of
sextant entered on it but with the word itself incompletely entered. The card
might, for instance, have the following information about the word: two syl-
lables, initial s, final t. “The entry would be a punch card equivalent of s_ _t.
Perhaps an incomplete entry of this sort is James’s ‘singularly definite gap’
and the basis of generic recall.” Subjects with a correct definition matching the
input and an incomplete word entry will know that they know the word and
will feel that they almost have it, that it is on the tip of their tongue. If they
are asked to guess the number of syllables and the initial letter they should
be able to do so. They should also be able to produce SS words. Subjects are
frequently able to tell even which syllable of the missing word is stressed.
When SM and SS words are “matched” with the missing word, “the match
brings out the missing parts the way the heat brings out anything written
in lemon juice,” in both the semantic and the phonetic dimensions. Such
features seem to be available not only in the definition but in the missing
word as well, even when the word itself is not available. Had they been
extracted only from the definition, one could not produce SS words. Thus
the semantic features and the phonetic information seem to be intensely
active even when the word itself cannot be recalled. The psychological real-
ity of such semantic and phonetic information is further supported, as we
have seen, by Fromkin’s findings with speech errors.
I have conjectured that both in the tot state and in poetic language the sup-
pressed semantic and phonetic features constitute some intangible and invisible
condensed mass that induces a thick atmosphere. There is, however, a serious
problem of communication here: How can one convey such information to stu-
dents, for instance? As to the tot state, everybody has experienced it. Someone
who hasn’t will never understand what we are talking about. As to poetic lan-
guage, we can only recommend what to look for and how to look at it and sug-
gest that what we are looking for is, in an important sense, something not
unlike the tot experience. Let us consider briefly stanza XLIX of Keats’s Isabella:
the senses that have a poorer vocabulary, rather than vice versa. Ullmann
has no satisfactory explanation for this. I argue that language is a highly
differentiated tool and is, apparently, unsuitable to convey these less differ-
entiated, elusive, mysterious meanings it is sometimes supposed to convey
in poetry. Speaking of the more differentiated sense in terms of the less
differentiated one is one of the ways to do this. The above stanza is a fine
example.
“An abstraction that has no stable visual shape” alludes to my concep-
tion derived from Gestalt theory that the smooth fusion of sensory infor-
mation is inhibited across clear-cut contours (for further discussion, see
Tsur 2012c). Ullmann is puzzled by the following line: “The same bright
face I tasted in my sleep” (Keats, Endymion, I, line 895). He (1957: 287) says
that it is a strange phrase but cannot account for this strangeness. Here,
too, the poet speaks of the visual sense in terms of taste, that is, an upward
transfer. However, based on the above reasoning, the stable characteristic
visual shape of “face” inhibits the smooth fusion of the senses.
Thus, the slip-of-the-tongue and tip-of-the-tongue phenomena provide
empirical evidence that words are bundles of semantic and phonetic fea-
tures. The feature-deletion theory of metaphor, too, conceives of words
as bundles of semantic features. Metaphoric contradiction deletes incom-
patible semantic features and foregrounds others, thus performing what
Ehrenzweig called “thing-destruction” and generating “thing-free” quali-
ties (by the same token it can account for humans’ ability to understand
novel metaphors to which they have not been exposed before). In poetry,
just as in the tot state, a mass of unexpressed features generates a thick,
intangible, invisible, and inaudible undifferentiated atmosphere. This is
because suppressed meanings in poetry do not mean deleted features but,
rather, that the remaining undeleted features that convey crucial informa-
tion are not explicitly mentioned.
Thus I take issue with Brown and McNeill’s suggestion that “the entry
would be a punch card equivalent of s_ _t. Perhaps an incomplete entry of
this sort is James’s ‘singularly definite gap’ and the basis of generic recall.”
The “singularly definite gap” is generated not only by the incomplete pho-
netic entry but by the suppressed mass of phonetic and semantic features.
Moreover, while this model provides an exceptionally illuminating expla-
nation of the process underlying the phenomenon, the solid object, the
punch card, grossly interferes with its phenomenological quality, which we
tend to experience as a condensed mass of engulfing diffuse particles sus-
pended in the air.
I have said that subjects in the tot state sometimes have exact informa-
tion on the number of syllables and the placement of stress in the missing
word. It would appear at first sight that this piece of information has little
relevance to our inquiry. In poetry the thick atmosphere is not generated
by the semantic features of one word that failed to grow together but is
abstracted from several words in a context, and there is no uniquely identi-
fiable word to be retrieved. One can thus have no knowledge of the number
of syllables or the location of primary stress. On second thought, however,
it is precisely these features of language that are systematized into syllabo-
tonic meter, independently of the specific words.
A subject trying to find the word ambergris thinks of Seagram. The usual sorts of
resemblance that constitute the main effects of the experiment (short strings of
identical letters at the beginning and ends of the words) are absent. Still there
is a resemblance that can hardly be accidental. All the letters [= speech sounds]
of Seagram are contained in ambergris. The word found seems to utilize the same
letter-stock or sound as the word sought without regard for order. This is a fasci-
nating outcome because it corresponds with one kind of rather common reading
mistake and, together with the reading mistakes, it suggests that order may be a
feature of a word that is stored independently of letters. (1970: 275)4
4. As to reading mistakes, more recently, drawing on a long series of rigorous neu-
ropsychological experiments, Dehaene gives a meticulously detailed account of the
reading process in the brain: “The adaptation process in the letterbox area reveals sev-
eral successive levels of letter coding, organized hierarchically from the back toward
the front of the occipito-temporal cortex, with an increasing degree of abstraction”
(2009: 91). At one stage, the system treats anagrams such as anger and range or cen-
ter and recent as the same; at the next stage it treats them as different (Dehaene
2009: 91–92; cf. chapter 2 above). This suggests that the reading mistake results from
stopping short in the cognitive chain. Sudden awareness of these subliminal stages
may account for pleasure in the scrambling of letters and speech sounds in witty or
emotional poetry. According to Freudian theory, one of the possible causes of pleasure
is regression to some earlier stage of functioning: “A part of the pleasure derives . . .
from the relation to infantile life”; “in all play with words, in puns as well as nonsense
talk, there is a renewal of the child’s pleasure when it just learns to master language”
(Kris 1965: 174). Thus, anagrams may serve as a publicly sanctioned mode of regression
to an earlier stage of information processing.
{
Ana- MARY gram
ARMY }
How well her name an Army doth present,
In whom the Lord of Hosts did pitch his tent!
I could not manage to write down the combinations of letters [which auto-
matically spurted out of my pen], and if there had been ten people present
they would not have been able to write down so many combinations as
came to me during the influx” (in Scholem 1961: 150–151) (see chapter 5).
In Hellenistic or Renaissance anagrams, the letters rather than the speech
sounds are manipulated, and the enormous charge of emotional energy
is eliminated. This process is nearer to wit than to ecstasy: the expressive
means are frozen into frivolous word games. I argue in Tsur 2003 that in
Herbert’s anagram this is taken one step further, where the frozen literary
device is revived by literary manipulations.
It is illuminating to compare the perceived effect of scrambling in
Coleridge’s “A damsel with a dulcimer” and Herbert’s “Mary” and “Army.”
The former sounds musical, smooth, emotional, “Romantic”; the latter,
rather sharp and witty, “Mannerist.” According to Jakobson (1960), the
distinctive characteristic of poetic language is that it compels the reader to
attend back to the linguistic signifier from the extralinguistic signified. In
chapter 6 I go one step further: I conceive of Mannerist and non-Mannerist
styles in terms of different sign relationships. As long as the signified pre-
serves its dominance to some extent, the emphasis on the signifier is more
natural, less “marked.” When the relative weights of the signifier and the
signified are balanced or the former overrides the latter, the process is per-
ceived as less natural, more artificial, more “marked.”
Jakobson (1968) distinguishes two stages in child language: the bab-
bling period, when the infant explores the similarity of speech sounds, and
the stage when the child learns to use speech sounds referentially. In this
stage, the organizing principle is contiguity. Similarity and contiguity are
organizing principles on all levels of language (Jakobson 1956). The dif-
ferentia of poetic language is that similarity is superimposed on contiguity
(Jakobson 1960). In our case, on the phonological level, in both “A damsel
with a dulcimer” and Mary and Army the contiguous speech sounds add
up to words and larger syntactic units. At the same time, the same speech
sounds form pairs of similar phoneme clusters. Syntactically, “with a dul-
cimer” is a restrictive attribute of “A damsel” and as such contiguous with
it. Thus, the referential function of language based on contiguity is in the
foreground; the similarity of the sounds is perceived as a smooth musical
background. The relationship between Mary and Army is neither restrictive
nor contiguous syntactically, but they are parallel in both their collitera-
tive and figurative structure, foregrounding the linguistic signifier at the
expense of the referent. This is what Joseph Addison (quoted in chapter 6)
called similarity of words rather than similarity of ideas. To be precise, both
instances display the same similarity of words, but in Herbert’s case the
two words refer to two dissimilar ideas, whereas Coleridge’s two phrases
join forces to designate one idea.
There is an additional aspect that may illuminate the difference
between the two examples, precisely where they deviate from regularity.
Both examples involve two similar clusters of signs: the same units, in
a different order, except one deviation. We have in dulcimer an /s/ and
in damsel a /z/, which is a voiced /s/. In Mary and Army the letters m,
a, r, and y are repeated, but the letter a refers to different vowels in the
two words. Consequently, the component shared by /s/ and /z/ may
fuse smoothly in perception, whereas the vowels signified by the letter
a in Mary and Army cannot. It is the deviant item that resolves whether
we are dealing with the patterning of speech sounds or graphemes: in
Coleridge the patterning of speech sounds, in Herbert the patterning of
graphemes. In chapter 6 (as well as previously in this chapter) I argued
that stable visual shapes interfere with the smooth fusion of elements
in poetry, yielding a sharp, witty, split focus, whereas sound gives infor-
mation about change and may fuse smoothly, yielding a soft integrated
focus. Consequently, the patterning of graphemes in Herbert, reinforced
by the genre “picture poetry,” results in a witty, “Mannerist” quality; the
patterning of speech sounds in Coleridge, in a smooth musical, emotional,
“Romantic” quality. The upshot of this discussion is as follows: the same
psycholinguistic mechanism underlies scrambling in certain speech and
reading errors, on the one hand, and in certain poetic devices, on the
other. What is more, in the poetic devices the same mechanism yields very
different stylistic effects, determined by aesthetic principles discussed
in chapter 6.
it deals with a piece of poetry, it also provides insights into the mutual
interferences between psychopathological processes and poetic language.
It is illuminating to note that Keats himself consciously relied, as we shall
see, on the same rules in revising his poem “La Belle Dame sans Merci” for
stylistic purposes as the unconscious relied on in misquoting Keats’s “Ode
to Apollo” for psychopathological purposes.
A young woman misquoted four lines from Keats’s “Ode to Apollo” as
The correct lines read as follows (the words forgotten and replaced by oth-
ers being italicized):
Significantly, the young woman initially could not recall when she memo-
rized these lines. But it came back after Brill suggested: “Judging by the
conversation, it would seem that this poem is intimately associated with
the idea of over-estimation of personality of one in love. Have you per-
haps memorized this poem when you were in such a state?” The upshot
of the ensuing story is as follows. Everything went well for a few months
until the woman suddenly received word that her Apollo, for whom she had
memorized these lines, had eloped with and married a very wealthy young
woman. A few years later she heard that he was living in a Western city,
where he was taking care of his father-in-law’s interests.
The misquoted lines are now quite plain. Brill’s comment about the
overestimation of personality among lovers unconsciously reminded the
woman of a disagreeable experience, when she herself overestimated
the personality of the man she loved. She thought that he was a god, but
he turned out to be even worse than the average mortal. The episode could
not come to the surface, because it was accompanied by very disagreeable
and painful thoughts, but the unconscious variations in the poem plainly
showed her present mental state. The poetic expressions not only were
changed to prosaic ones but clearly alluded to the whole episode (Freud
1953: 19–20).
For a literate society the function of verse is primarily esthetic, but for preliter-
ate societies, verse is a means of transmitting verbal information of cultural
importance with a minimum of paraphrase. The rules of verse are, in effect, an
addition to phonology which requires that recalled material not only should pre-
serve the semantic values of the original, but should also conform to a specific,
rule-determined phonetic pattern. (1972: 325)
An individual does not normally take such a situation detail by detail and metic-
ulously build up the whole. In all ordinary instances he has an overmastering
tendency simply to get a general impression of the whole; and, on the basis of
this, he constructs the probable detail. . . . The construction that is effected is the
sort of construction that would justify the observer’s “attitude.” (1932: 206–207;
cf. Tsur 2008a: 17–18)
Here the unconscious mind had its way by interfering with the cognitive
process at the “attitude” level. By modifying the young woman’s attitude
toward the poem, the construction of the probable detail resulted in an
utterly different fourth line. “Prosaic truths that came too late” expresses
the young woman’s attitude toward the occasion when she memorized
Keats’s lines, but here it also affected the attitude that served the recon-
struction of the lines themselves from memory. In other words, the two
kinds of attitude were not kept properly separated. Here as well the line is
constructed in accordance with the prosodic rule memorized: The fourth
line deviates from the preceding lines in being iambic; it begins with a tri-
syllabic word and ends with the sounds -ate. The substitution of where for
when before this can only be explained by the assumption that once her
unconscious removed the poetic restrictions on the recoding of the text,
the woman needed to reconstruct the probable detail by rule.
CONCLUSION
* * *
It would appear that the artwork as cognitive fossil has been with us for the
past eighteen thousand years or so. The great neuropsychologist Stanislas
Dehaene begins his book on consciousness and the brain with the follow-
ing example:
Deep inside the Lascaux cave, past the world-renowned Great Hall of the Bulls,
where Paleolithic artists painted a colorful menagerie of horses, deer, and bulls,
starts a lesser-known corridor known as the Apse. There, at the bottom of a
sixteen-foot pit, next to fine drawings of a wounded bison and a rhinoceros, lies
one of the rare depictions of a human being in prehistoric art [Figure 10.3]. The
man is lying flat on his back, palms up and arms extended. Next to him stands
a bird perched on a stick. Nearby lies a broken spear that was probably used to
disembowel the bison, whose intestines are hanging out.
The person is clearly a man, for his penis is fully erect. And this, according
to the sleep researcher Michel Jouvet, illuminates the drawing’s meaning: it
depicts a dreamer and his dream. As Jouvet and his team discovered, dreaming
occurs primarily during a specific phase of sleep, which they dubbed “paradoxi-
cal” because it does not look like sleep; during this period, the brain is almost
as active as it is in wakefulness, and the eyes ceaselessly move around. In males,
this phase is invariably accompanied by a strong erection (even when the dream
is devoid of sexual content). Although this weird physiological fact became
Figure 10.3 A dreamer and his dream: painting in the Lascaux cave, circa eighteen thou-
sand years old.
known to science only in the twentieth century, Jouvet wittily remarks that
our ancestors would easily have noticed it. And the bird seems the most natural
metaphor for the dreamer’s soul: during dreams, the mind flies to distant places
and ancient times, free as a sparrow. (2014: 1–2)
REFERENCES
Abd Razak, Ratna Roshida. 2007. “Arabic Poet Al-Mutanabbi: A Maslovian Humanistic
Approach.” PsyArt. Available online: http://psyartjournal.com/article/show/
roshida_abd_razak-arabic_poet_al_mutanabbi_a_maslovian_hum.
Addison, Joseph. 1951 [1711–1712]. “Spectator Papers.” In The Great Critics, ed. James
Harry Smith and Edd Winfield Parks, pp. 819–825. New York: Norton.
Aristotle. 1932. The “Rhetoric” of Aristotle. Trans. Lane Cooper. New York: Appleton
Century.
Aristotle. 1951. Aristotle, “Poetics.” Trans. S. H. Butcher. In The Great Critics, ed. James
Harry Smith and Edd Winfield Parks, pp. 28–61. New York: Norton.
Arnheim, Rudolf. 1957. Art and Visual Perception. London: Faber and Faber.
Balaban, Victor. 1999. “Religious Metaphor and Cognitive Linguistics: Perceptual
Metaphors for Knowledge at a Marian Apparition Site.” In Metaphor and
God-Talk, ed. Lieven Boeve and Kurt Feyaerts, pp. 253–271. Bern: Peter Lang.
Bartlett, F. C. 1920. “Some Experiments on the Reproduction of Folk Stories.”
Folk-Lore 31: 30–47. Available online: http://home.iscte-iul.pt/~fgvs/
Bartlett,%20Experiments.pdf.
Bartlett, Frederic C. 1932. Remembering. Cambridge: Cambridge University Press.
Beardsley, Monroe C. 1958. Aesthetics: Problems in the Philosophy of Criticism.
New York: Harcourt, Brace and World.
Beaver, Joseph C. 1968. “A Grammar of Prosody.” College English 29: 310–321.
Beer, J. B. 1959. Coleridge the Visionary. London: Chatto and Windus.
Bergson, Henri. N.d. Laughter: an Essay on the Meaning of the Comic. The Project
Gutenberg EBook.
Berliner, Paul. 1978. The Soul of Mbira: Music and Traditions of the Shona People of
Zimbabwe. Berkeley: University of California Press.
Blechner, Mark J. 1977. “Musical Skill and the Categorical Perception of Harmonic
Mode.” Haskins Laboratories: Status Report on Speech Research SR-51/52.
Available online: http://www.haskins.yale.edu/sr/SR051/SR051_12.pdf.
Bownds, M. Deric. 1999. The Biology of Mind—Origins and Structures of Mind, Brain,
and Consciousness. Bethesda, MD: Fitzgerald Science Press.
Bradburn, Beth. “Interview with Reuven Tsur.” Literature, Cognition & the Brain.
Available online: http://www2.bc.edu/~richarad/lcb/fea/tsurin/tsurmain.html.
Brann, Ross. 1991. The Compunctious Poet: Cultural Ambiguity and Hebrew Poetry in
Muslim Spain. Baltimore: Johns Hopkins University Press.
Brooke-Rose, Christine. 1958. A Grammar of Metaphor. London: Secker and Warburg.
2
6
[ 262 ] References
3
6
2
References [ 263 ]
4
6
2
Goldstein, K., and M. Scheerer. 1941. “Abstract and Concrete Behavior.” Psychological
Monographs 53 (2): 1–10.
Goldwasser, Orly. 2010. “How the Alphabet Was Born from Hieroglyphs.” Biblical
Archeological Review 36 (2): 40–53. Available online: http://members.bib-arch.
org/publication.asp?PubID=BSBA&Volume=36&Issue=02&ArticleID=06.
Golston, Chris, and Tomas Riad. 1997. “The Phonology of Classical Arabic Meter.”
Linguistics 35: 111–132. Available online: http://zimmer.csufresno.edu/~chrisg/
index_files/ArabicMeter.pdf.
Golston, Chris, and Tomas Riad. 2010. “Empirical Arabic Metrics: Implications for a
General Model.” Paper presented at “Metrical Studies Today: Topics, Tools and
Theories: An International Conference,” June 16–18, Halden, Norway.
Gombrich, E. H. 1954. Visual Metaphors of Value in Art. New York: Harper and Bros.
Gombrich, E. H. 1985. Meditations on a Hobby Horse. Chicago: University of
Chicago Press.
Granville-Barker, Harley. 1957 [1936]. Preface to Hamlet. New York: Hill and Wang.
Greene, Roland, ed. 2013. The Princeton Encyclopedia of Poetry and Poetics, 4th ed.
Princeton: Princeton University Press.
Halle, Morris, and Samuel Jay Keyser. 1966. “Chaucer and the Study of Prosody.”
College English 28: 187–219.
Halle, Morris, and Samuel Jay Keyser. 1971. English Stress: Its Form, Its Growth, and Its
Role in Verse. New York: Harper and Row.
Halpern, Martin. 1962. “On the Two Chief Metrical Modes in English.” PMLA
77: 177–186.
Halevy, Yehuda. 1994. “Treatise on Meter, appendix to Tova Rozen “The Strophic
movement in Yehuda Halevy’s Poetry—An Alternative Poetics?” In Essays
in Honour of Israel Levin, eds. Reuven Tsur and Tova Rozen, pp. 315–328. Tel
Aviv: Tel Aviv University (in Hebrew).
Halevy, Yehuda. 2009. The Kuzari: In Defense of the Despised Faith.
New York: Feldheim.
Hayes, Bruce. 1985. “Iambic and Trochaic Rhythm in Stress Rules.” In Proceedings of
BLS 11: Parasession on Poetics, Metrics, and Prosody, ed. N. Niepokuj, M. VanClay,
V. Nikiforidou, and D. Jeder, pp. 429–446. Berkeley: Berkeley Linguistic
Society.
Hayes, Bruce. 1995. Metrical Stress Theory: Principles and Case Studies.
Chicago: University of Chicago Press.
Herrnstein Smith, Barbara. 1968. Poetic Closure. Chicago: University of Chicago Press.
Hofstadter, Douglas R. 1997. Le Ton beau de Marot: In Praise of the Music of Language.
London: Bloomsbury.
Horace [Quintus Horatius Flaccus]. 1995. “The Art of Poetry (Ars Poetica).” Trans. Leon
Golden. In Horace for Students of Literature: The Ars Poetica and Its Tradition.
Gainesville: University Presses of Florida. Available online: http://www.baylor.
edu/theatre/index.php?id=89823.
Ibn Gabirol, Solomon. 1971. The Liturgical Poems. Ed. Dov Yarden. Jerusalem.
Jakobson, Roman. 1956. “Two Aspects of Language and Two Types of Aphasic
Disturbances.” In Fundamentals of Language, by Roman Jakobson and Morris
Halle, pp. 55–82. The Hague: Mouton.
Jakobson, Roman. 1960. “Closing Statement: Linguistics and Poetics.” In Style in
Language, ed. Thomas A. Sebeok, pp. 350–377. Cambridge: MIT Press.
Jakobson, Roman. 1968. Child Language, Aphasia, and Phonological Universals. The
Hague: Mouton.
[ 264 ] References
5
6
2
References [ 265 ]
62
[ 266 ] References
7
6
2
References [ 267 ]
8
62
Schramm, Wilbur L. 1935. Approaches to the Science of English Verse. Iowa City: Iowa
University Press.
Scott, Clive. 2011. “Free Verse and the Translation of Rhythm.” Thinking Verse
1: 67–101.
Shoshany, Ronit. 2009. “The Original Purpose of Biblical Accentuation” (in Hebrew).
Masʾat Aharon: Linguistic Studies Presented to Aron Dotan, ed. Moshe Bar-Asher
and Chaim E. Cohen, pp. 469–486. Jerusalem: Bialik Institute.
Siciliano, Italo. 1934. François Villon et les thèmes poétiques du Moyen Age.
Paris: Librarie Armand Colin.
Skoyles, John. 1985. “Did Ancient People Read with Their Right Hemispheres?
A Study in Neuropalaeographology.” New Ideas in Psychology 3: 243–252.
Skoyles, John. 1988. “Right Hemisphere Literacy.” In The Alphabet and the Brain: The
Lateralisation of Writing, ed. Derrick de Kerchhove and Charles J. Lumsden,
pp. 362–380. Berlin: Springer-Verlag.
Skoyles, John. 1997. “Ancient Readers: Were They Right Hemisphere Literates?”
Available online: http://www.users.globalnet.co.uk/~skoyles/index.htm.
Smith, Hallett. 1968. Elizabethan Poetry: A Study in Conventions, Meaning, and
Expression. Ann Arbor: University of Michigan Press.
Smith, James. 1933. “Metaphysical Poetry.” Scrutiny 2: 222–238.
Snyder, E. D. 1930. Hypnotic Poetry: A Study of Trance-Inducing Techniques in
Certain Poems and Its Literary Significance. Philadelphia: University of
Pennsylvania Press.
Sternberg, Meir. 1976. “Temporal Ordering, Modes of Expositional Distribution, and 3
Models of Rhetorical Control in the Narrative Test—Faulkner, Balzac and Austen.”
PTL: A Journal for Descriptive Poetics and Theory of Literature 1 (2): 295–316.
Stevenson, Robert Louis. 1882. Familiar Studies of Men and Books. Available
online: http://www.fullbooks.com/Familiar-Studies-of-Men--Books.html.
Stroop, Ridley J. 1935. “Studies of Interference in Serial Verbal Reactions.” Journal of
Experimental Psychology 18: 643–662. Available online: http://psychclassics.
yorku.ca/Stroop/.
Sidney, Sir Philip. “The Defense of Poesy.” In English Essays: Sidney to Macaulay. The
Harvard Classics. 1909–14. http://www.bartleby.com/27/1.html.
Sypher, Wylie. 1955. Four Stages of Renaissance Style. Garden City, NY: Anchor.
Szerb, Antal. 1943. A Világirodalom Története. Budapest: Révai.
Taylor, Jane H. M. 2001. The Poetry of François Villon: Text and Context.
Cambridge: Cambridge University Press.
Thomson, Philip. 1972. The Grotesque. London: Methuen.
Tsur, Reuven. 1969. Studies in Mediaeval Hebrew Poetry (in Hebrew). Tel Aviv: Daga.
Tsur, Reuven. 1972. “Articulateness and Requiredness in Iambic Verse.” Style 6: 123–148.
Tsur, Reuven. 1974. “Poem, Prayer and Meditation: An Exercise in Literary
Semantics.” Style 8: 405–425.
Tsur, Reuven. 1975. “Two Critical Attitudes: Quest for Certitude and Negative
Capability.” College English 36: 776–788.
Tsur, Reuven. 1977a. A Perception-Oriented Theory of Metre. Tel Aviv: Porter Institute
for Poetics and Semiotics.
Tsur, Reuven. 1977b. “Some Remarks on the Nature of Trochees and Iambs and
Their Relationship to Other Metres.” In A Perception-Oriented Theory of Metre,
pp. 83–96. Tel Aviv: Porter Institute for Poetics and Semiotics. Available
online: http://www.tau.ac.il/~tsurxx/Trochee/bit_of_personal_history2.html.
[ 268 ] References
9
6
2
References [ 269 ]
0
72
[ 270 ] References
271
INDEX
[ 272 ] Index
273
defense mechanism, 14, 66, 85, 101 Fowler, Roger, 174, 183, 226–227
Dehaene, Stanislas, 28, 42–49, 115, 248, Frazer, Sir James, 7–8
258–259 Frenkel-Brunswik, Else, 96
depth-psychological, 2, 5, 20, 232–233 Freud, Sigmund, 20–21, 66, 69, 94, 101,
devotional poetry, 62, 127 231–234, 239, 244, 248, 252–253,
Dewey, John, 163 257–258
distinctive features, 164, 169, 183, 185, Fromkin, Victoria, 235–238, 245, 249, 257
233, 235–239, 241–243 Fry, D. B., 26, 218
Dolman, John, 177 fusiform, 48
Donne, John, 20, 62, 134, 142–143, 151,
154–155, 210, 227 Gallese, Vittorio, 200–201
double syntax, 122, 128 garden descriptions, 59, 145
Drayton, Michael, 20, 70, 73–75, 77 generation by rule, 254
Dresher, Elan B., 52 generative metrists, 172, 183–184
drinking poems, 59 generic recall, 245, 247
Dunash Ben Labrat, 28 genre, 16, 52, 57, 59, 64, 84, 108
Gestalt, 13, 22, 24, 30, 40, 100–102, 104,
echoic memory, 194 107, 128, 149, 172, 179, 211, 216,
Ehrenzweig, Anton, 3, 5, 14, 16–18, 63, 222, 227, 229, 246–247
66, 84–85, 87, 100–102, 105, 126, gestalt-free, 100, 104, 107, 149, 246
128, 232, 243, 247 Gestaltist, 10, 22, 100–101
Einat-Nov, Idit, 99 Gil, David, 52
Eitan,Zohar and Renée Timmers, glissandi, 16, 18
188–191, 206–207 Goethe, Johann Wolfgang von, 75
electrical perturbations, 117–118, 129 Goldstein K. and M. Scheerer, 96
elegant solution to a problem, 157, 160, Goldwasser, Orly, 44–46
163, 165–166, 174, 184 Golston, Krus and Tomas Riad, 26
Elior, Rachel, 101, 105, 111, 129 Gombrich, Ernst Hans, 3, 17–18
Elizabethan drama, 4, 36–37, 40 Good Continuation, 222
ellipsis, 15, 31, 88 Granville-Barker, Harley, 36, 39
Emanuel of Rome, 29 graphic primitives, 47–48
Empson, William, 122 grounding, 190, 200, 206
erotic symbolism, 14 grouping, 13, 26, 172, 174–176, 210–211,
Eskimo story, 33–34, 40 214–215, 218–219, 221–225, 229
Even-Shoshan, Abraham, 124
Even-Zohar, Itamar, 7 Halle, Morris and Samuel J. Keyser, 9,
event time and micro-time, 198 212–215, 218, 220, 229
evolution, 7, 14, 42–43, 47, 55, 84, 154 Halpern, Martin, 211, 216
evolutionary tree, 7, 55, 154 Hamlet, 21, 36–40, 159, 178, 185
Ezekiel, 103, 110 Harvey, William, 57–58
Hayes, Bruce, 218
fairy tales, 20, 66, 85 Hekhalot, 103, 105, 108, 110–111, 129
false wit, 133–134, 153 Herbert, George, 38, 134, 249–252
Father William, 83 Herrnstein Smith, Barbara, 22, 224
feature reversal, 239 hieroglyphs, 44–47
feature-deletion, 206, 247 Hofstadter, Douglas, 160, 173, 180–182
Feldman, Jerome, 200 Holy Living Creatures, 99, 103, 110
Fleischer, Ezra, 108–110 hominid, 7
Folk-Lore in the Old Testament, 7 Homo erectus, 7
Fortis, Jean-Michel, 200 Homo sapiens, 7, 115
I n de x [ 273 ]
274
[ 274 ] Index
275
love poetry, 20, 59, 63, 139, 153 More, Sir Thomas, 177
Lowes, John Livingston, 239 Morgenstern, Christian, 152
Luchins, 76 Moses Ibn Ezra, 59–60, 84, 145
Lyons, John, 199, 204 Mozart, Leopold, 162
Lyotard, Jean-François, 180 Muecke, D. C., 237
multiple relationship, 241
magical number 7 ± 2, 13, 22–23, 179, 217 multivalence, 241
mannerism, 18, 131, 133–135, 138, 143, muwwashaḥ, 59
151, 153 mysterium tremendum, 108, 129
markedness, 9–12, 43, 51, 100, 133, 143,
150, 153, 192, 199–200, 205, 214, natural selection, 28, 40, 49, 54
222, 235, 251 Neanderthals, 7
Marranos, 4 Neapolitan songs, 18
Maslow, Abraham H., 57, 63–64, 68, 84 Neisser, Ulric, 21, 150, 194
Mastroiani, Marcello, 83 Neoplatonic, 140, 144
mbira, 187–188 Neural Theory of Metaphor, 200, 207
McNeill, David, 244, 247 neurological fallacy, 207
mediated association, 187, 189 neuronal recycling, 42, 46–47
meditative catalogue, 104, 182 neurons, 43, 46–47, 200–201
Merkabah, 99–100, 102–103, 105–106 Newberg, Andre, Eugene D’Aquili, and
Merkabah hymn, 99–100, 102, 105 Vincent Rause, 116
Merkabah mysticism, 99, 102 noncategorial sounds and recoding, 194
meta-awareness, 150 Norman, Donald. A., and David E.
metaphor, 54, 61, 75, 104, 137, 139, 146, Rumelhart, 32
154, 183, 188–191, 200, 202–203, Noy, Pinchas, 234
205–207, 232, 247, 259 Numinous Hymns, 104
Metaphysical, 54, 100, 105, 129, 132, nursery rhymes, 212, 221
134–135, 138–140, 142–143, 145,
149–155 Occam’s Razor, 22
Metaphysical conceit, 54, 132, 139–140, oceanic dedifferentiation
142, 151, 153 (undifferentiation), 101–102,
metaphysical intuition, 100 106–107
Metaphysical style, 135, 143 Oedipal situation, 93, 96
metaphysical wit, 134, 145, 149, 152 Oedipus complex, 96–97
metonymy, 80, 91, 93, 110, 138, 143, 155, Oedipus Rex, 62
165, 169, 259 Ohala, John, 204
metrical set, 210, 222 onomatopoeia, 161
Meyer, Leonard B., 13, 24, 26, 173, 211, ophanim, 108
217–218 organized violence against language,
migration, 1–2, 19–20, 53–55 231, 258
Miller, George A., and Philip N. orientation mechanism, 103, 107, 150
Johnson-Laird, 150 origins of language, 3
Miller, George, 13, 22, 179 ornament, 3, 5, 14, 16–18, 63–64,
Milner, Marion, 101 85, 87, 101–102, 108, 118,
Milton, John, 11–12, 106, 115, 143, 176, 128–129, 135
180, 226 Ornstein, Robert E., 38
mirror neurons, 200 Otto, Rudolf, 104, 108, 129, 232
Mirrour for Magistrates, 61, 177 overarticulation, 210–211, 214, 221–222
misquotation, 231, 252, 254–256, 258 overstressing, 210, 214, 221
Molière, Jean-Baptiste Poquelin, 237 oxymoron, 87
I n de x [ 275 ]
276
Pagis, Dan, 29, 60, 83, 135, 233–236 Psychopathology of Everyday Life, 231,
panchronistic tendencies in 252, 257
synesthesia, 246 Ptolemaic world picture, 143
panegyrics, 58–60 Pure Attribution, 146, 153
Paradise Lost, 11, 106, 143, 180
paradox, 20–21, 69–70, 73–76, 80, qassida, 59
86–87, 125, 152 Quintilian, Marcus Fabius, 97
parody, 37, 78, 80, 82–83, 86, 220
paronomasia, 137 Racine, Jean, 173
parsimony, 21–23, 31–32, 198, 206 Radnóti, Miklós, 27
passing notes, 18 rahitim, 108
Pears, David, 151 Ransom, John Crowe, 137
perception, 2, 4, 6, 16, 22, 26, 30–31, 34, rapid closure, 96
37–40, 43, 69, 93, 96, 125, 135, regional quality, 10, 99–100
139, 146–147, 149, 151, 159, 177, repeated social transmission, 1, 4–5, 8,
187, 189, 195–199, 203, 206, 209, 19–21, 26–27, 30, 34, 36, 39–43,
218, 222, 229, 232, 252, 256 45, 47–49, 53–55, 63–64, 67, 84,
peripeteia, 91 154, 259
Persinger, Michael, 117 retrieval, 233–234, 244–245
personal creator, 140, 144 Revithiadou, Anthi, 218
personality style, 6 rhyme, 29, 68, 78–80, 82–83, 110–111,
Petrarch, Francesco, 20, 64, 67–69, 75, 113–114, 132–134, 158, 165,
138, 154 168–169, 172, 177, 182, 184–185,
Philips, Ambrose, 220 212, 220–223, 239–242, 255
Phoenician script, 45–46 Rice, Curt, 26, 218
phonetic code, 194 Richard III, 174, 176
phonetic recoding, 195 riddles, 20, 68–69
pigeonholing, 87–88, 96–97 right hemisphere, 15, 38, 40, 48, 67
Pinkernell, Gert, 82–83 Rimbaud, Arthur, 104
pitch ascent, 204 Romanticism, 11, 59, 78, 104, 134–135,
pitch descent, 204 138, 147, 149–150, 153, 182, 232,
Platonic, 106, 135, 137–138, 141, 243, 246, 251–252, 257
144–147, 149 Ronsard, Pierre de, 20, 64–65, 69
Platonic dialogue, 144 Rorschach inkblot test, 128
Platonic poetry, 137 Rosen, Tova, 144
Poems of Complaints, 60 Rumelhart, David E., 31–32
Poems of Contemplation, 59 Russian school of verse study, 8
Polányi, Michael, 47
Pope, Alexander, 11–12, 154, 176, Sacks, Oliver., 117
226–227 saturation, 111, 228
possible world, 115, 159 Sayf al-Dawlah, 58–59
Précieux, 135, 138, 145 Schank, Roger, 31–33, 254
primacy and recency effects, 76 Schank and Abelson, 31–33
Principle of Marginal Control, 47 Scheindlin, Ray, 144
principle of maximum contrast, 198 Schema, 25, 30–36, 40–41, 47, 55, 68,
prosody, 9, 49, 59, 144, 210, 228 85, 150, 210, 256
Proto-Sinaitic, 45–47 Schirmann, Jefim, 28–29
proverbs, 78, 82–83 Scholem, Gershom G., 100, 102–105,
psychic distance, 18 107, 115–116, 249, 251
psychological reality, 234–236, 245 Schramm, Wilbur, 218
[ 276 ] Index
277
Scott, Clive, 157–160, 164–165, 172, static and dynamic sound stimuli, 202
174–176, 180, 182–184 Sternberg, Meir, 76–77
scrambling, 248, 250–252, 257 Stevenson, Robert Louis, 75
script, 32–33, 45–46 stock phrases, 108, 110
secondary elaboration, 5, 14 stress maxima in weak positions, 11
self-specifying information, 103, 107 stress maximum, 11, 212–215, 11, 212–216
semantic features, 111–112, 233, stress-timed language, 54
240–241, 244–248 Stroop, Ridley, 67–68, 203
semantic primitives, 254 Stroop effect, 67, 203
Serbian folk epic, 23, 49 sublime, 99, 105
Shakespeare, William, 4, 11, 37, 61, 154, Sumerian-Babylonian-biblical influences, 8
174, 176–177, 212, 226 syllable-timed language, 28, 54
sharpened, 16–17, 21, 69–70 syllepsis, 87
Shekhinah, 103–104 Symbolist, 104, 170, 182, 232, 243, 257
Shelley, Percy, Bysshe, 10, 123, 158, 176, syncopation, 226, 228
222–225 Sypher, Wylie, 131, 135, 143, 151
Shlomo Ibn Gabirol, 60, 102, 108, 110, Szabó, Lőrinc, 165, 167–169
112, 116, 118, 121, 123, 125–128, Szerb, Antal, 242
136, 138, 140–145, 152, 154–155
Shlonsky, Abraham, 146, 178 Taylor, Jane, 78, 82–83
Shmuel Hanagid, 136 temporal lobe transients, 117
short-term memory, 12–13, 22, 24, 30, Tennyson, Lord Alfred, 102, 223
52, 179, 217 The Golden Bough, 7
Shoshany, Ronit, 1, 52 thematized predicate, 104
Siciliano, Italo, 61 Theurgy, 115
Sidney, Sir Philip, 62 thing-destruction, 104, 247
sign relationships, 131, 251 thing-free, 100, 104, 149, 181–182, 247
significant variation, 92 Thomson, Philip, 152
signified, 114–115, 132–138, 151, 184, Tiberian system, 53
251–252 tip-of-the-tongue, 231–232, 247, 257
signifier, 114, 132–138, 151, 184, 251 topicalized attribute, 104
silukim, 108 tot experience, 243, 245
Silvestre, Armand, 173, 181 tot experiment, 232
slip-of-the-tongue, 231–232, 247, 257 tot phenomenon, 231, 243
Smith, Hallett, 22, 61–62, 69, 139, 224 Tóth, Árpád, 158–159, 165, 167–169
Smith, James, 139 translation, 12, 27, 53, 59, 65–66, 73–74,
Snyder, E.D., 106 79, 82–83, 105, 136–138, 140–141,
sonnet, 29, 61–62, 64–65, 67–69, 73, 75, 145–146, 148–149, 157–161,
77, 142, 177, 250 164–175, 177–178, 180–182,
Southey, Robert, 83 184–185, 194, 228, 242
spatial template, 190, 194–195, 197–200, trochee, 9, 27, 50, 52–54, 172, 174, 183,
202, 204, 207 209, 211–223, 225–229, 255
speaking in tongues, 117 troubadours, 20
speech errors, 235–238, 241–242, 245, true wit, 133–134
249, 257 Tsur, Reuven, 6, 9, 11–13, 24, 28, 30,
Spenser, Edmund, 256 38–39, 41–42, 46, 50–52, 54, 62,
Spooner, William Archibald, 237 73, 96, 99, 104, 106–107, 135,
spoonerism, 237–239, 242–243, 257 145, 166–167, 176, 179, 189, 192,
stages in the development of artistic 207, 209, 222–224, 240–243, 247,
devices, 16, 101 249–252, 254–256
I n de x [ 277 ]
278
[ 278 ] Index
279
280
281
282