(The Terry Lectures Series) Mary Douglas - Thinking in Circles - An Essay On Ring Composition-Yale University Press (2007)

the terry lectures
THINKING IN CIRCLES
ALSO IN THE TERRY LECTURE SERIES
Becoming: Basic Considerations for a Psychology of Personality

Gordon W. Allport
A Common Faith John Dewey
Psychoanalysis and Religion Erich Fromm
Master Control Genes in Development and Evolution
Walter J. Gehring
Israelis and the Jewish Tradition David Hartman
The Divine Relativity Charles Hartshorne
Psychology and Religion Carl G. Jung
Freud and the Problem of God Hans Küng translated by Edward Quinn
Education at the Crossroads Jacques Maritain
Exorcism and Enlightenment H. C. Erik Midelfort
Belief in God in an Age of Science John Polkinghorne
Freud and Philosophy: An Essay on Interpretation
Paul Ricoeur translated by Denis Savage
The Meaning of Evolution George Gaylord Simpson
One World Peter Singer
The Courage to Be Paul Tillich
The Empirical Stance Bas C. van Fraassen
MARY DOUGLAS
Thinking in Circles
An Essay on Ring Composition
YALE UNIVERSITY PRESS ≤ NEW HAVEN AND LONDON

Copyright ∫ 2007 by Mary Douglas.
All rights reserved.
This book may not be reproduced, in whole or in part,
including illustrations, in any form (beyond that copying
permitted by Sections 107 and 108 of the U.S. Copyright
Law and except by reviewers for the public press), without
written permission from the publishers.
Printed in the United States of America.
Library of Congress Cataloging-in-Publication Data

Douglas, Mary, 1921–
Thinking in circles : an essay on ring composition / Mary
Douglas.
p. cm. — (The Terry lectures)
Includes bibliographical references and index.
isbn-13: 978-0-300-11762-2 (cloth : alk. paper)
isbn-10: 0-300-11762-0
1. Narration (Rhetoric) 2. Homer, Iliad. 3. Bible. O.T.
Numbers—Criticism, interpretation, etc. 4. Sterne,
Laurence, 1713–1768. Life and opinions of Tristram
Shandy, gentleman. I. Title.
pn212.d68 2007
808—dc22
2006020410
A catalogue record for this book is available from the

British Library.
The paper in this book meets the guidelines for perma-

nence and durability of the Committee on Production
Guidelines for Book Longevity of the Council on Library
Resources.
10 9 8 7 6 5 4 3 2 1
THE DWIGHT HARRINGTON TERRY
FOUNDATION LECTURES ON
RELIGION IN THE
LIGHT OF SCIENCE AND PHILOSOPHY
The deed of gift declares that ‘‘the object of this foundation is

not the promotion of scientific investigation and discovery, but
rather the assimilation and interpretation of that which has
been or shall be hereafter discovered, and its application to
human welfare, especially by the building of the truths of sci-
ence and philosophy into the structure of a broadened and pu-
rified religion. The founder believes that such a religion will
greatly stimulate intelligent effort for the improvement of hu-
man conditions and the advancement of the race in strength and
excellence of character. To this end it is desired that a series of
lectures be given by men eminent in their respective depart-
ments, on ethics, the history of civilization and religion, biblical
research, all sciences and branches of knowledge which have an
important bearing on the subject, all the great laws of nature,
especially of evolution . . . also such interpretations of literature
and sociology as are in accord with the spirit of this foundation,
to the end that the Christian spirit may be nurtured in the
fullest light of the world’s knowledge and that mankind may be
helped to attain its highest possible welfare and happiness upon
this earth.’’ The present work constitutes the latest volume pub-
lished on this foundation.
CONTENTS
Preface ix
1 Ancient Rings Worldwide 1
2 Modes and Genres 17
3 How to Construct and Recognize a Ring 31
4 Alternating Bands: Numbers 43
5 The Central Place: Numbers 58
6 Modern, Not-Quite Rings 72
7 Tristram Shandy: Testing for Ring Shape 85
8 Two Central Places, Two Rings: The Iliad 101
9 Alternating Nights and Days: The Iliad 115
10 The Ending: How to Complete a Ring 125
11 The Latch: Jakobson’s Conundrum 139
Notes 149
Index 161
CONTENTS IN A RING
1. Ancient Rings Worldwide

2. Modes and Genres — 11. The Latch: Jakobson’s
Conundrum
3. How to Construct — 10. The Ending:
and Recognize How to Complete
a Ring — a Ring
4. Alternating Bands: — 9. Alternating
Numbers — Nights and
Days: The Iliad
5. The Central Place: — 8. Two Central
Numbers — Places, Two Rings:
The Iliad
6. Modern, Not-Quite Rings

7. Tristram Shandy: Testing
for Ring Shape
PREFACE
To publish at last an essay on ring composition is a great personal

satisfaction. I first read the book of Numbers in 1987 and was sur-
prised to find that it has been much depreciated by commentators for
disorderly writing. I later found that it is not disorderly but well
organized as an elegant ring.
After seeing how badly that great book has fared at the hands of
qualified commentators, I could not get the topic off my mind. I was,
and still am, convinced that there is a lot that ought to be better
known about ring composition. Something should be done to end the
unworthy slurs on great authors and ancient texts. I put off making a
serious study because I was not nearly qualified (and I still am not).
The topic is arcane; the ground has been dug over many times by
specialist scholars commanding many languages. I must regard my-
self as a visitor to the field of biblical scholarship. To have the courage
to make a strong complaint I would have had to focus on the task
exclusively and put everything else aside for at least a decade. Given
these obstacles, I would never have made a start but for the honor of
an invitation from King’s College London in 2002 to give the F. D.
Maurice Lectures. There was absolutely no restriction on topic, and
no requirement to publish. Instead of the profound and scholarly
review that the occasion deserved, I settled for just waving a flag. I
wanted to make the lectures a call for help. In effect, help was given so
generously that when I had a second invitation from Yale University
to give the Terry Lectures the next year my diffidence had been
dispelled.
The audience at King’s College had been stimulating and almost
too polite. I jumped at the chance to invite more criticism and discus-
sion. The audience at Yale was just as polite and full of stimulating
suggestions, most of which I have tried to follow in these pages. I
remember gratefully discussions with Geoffrey Hartman and David
ix
PREFACE
Apter, and the strong support of Dianne Witte, who organized every-
thing. By the time I had finished I realized that the project was even
more formidable than I had originally feared. I am grateful that Yale
University Press offered to publish the second set of lectures on ring
composition.
Ring composition is found all over the world, not just in a few
places stemming from the Middle East, so it is a worldwide method of
writing. It is a construction of parallelisms that must open a theme,
develop it, and round it off by bringing the conclusion back to the
beginning. It sounds simple, but, paradoxically, ring composition is
extremely difficult for Westerners to recognize. To me this is myste-
rious. Apparently, when Western scholars perceive the texts to be
muddled and class the authors as simpletons, it is because they do not
recognize the unfamiliar method of construction.
Friends ask me, what does it matter? Why is it important to know
the construction? This leads to another point: in a ring composition
the meaning is located in the middle. A reader who reads a ring as if
it were a straight linear composition will miss the meaning. Surely
that matters! The text is seriously misunderstood, the composition is
classed as lacking in syntax, and the author dismissed with disdain.
Surely, misinterpretation does matter.
The anthropologists’ standard criticism of attempts to interpret
mythology apply to this venture. A typical gibe is to accuse the would-
be myth analyst of giving free rein to her imagination. Friends have
said, ‘‘Ring composition is a loose and fuzzy concept, Mary will al-
ways be able to find a ring form if she looks hard enough, in a laundry
list, sports news, or whatever. Rings are everywhere.’’ This lethal
criticism I must rebut. Fear of it was one of the reasons why I was not
too disappointed to find that Leviticus is not an example of con-
struction in a ring.
I had very much hoped to reveal the ring form of that book; I tried
hard and I failed. The consolation was to discover that Leviticus
conformed to another famous type of composition, figure poetry,
which I was not expecting at all. So the method of enquiry was justi-
fied even though the results, as far as ring composition is concerned,
x
PREFACE
were negative. Healthy respect for the same criticism from Bible
scholars accounts for my interest in identifying the rules that the ring
authors have been following, and accounts also for some rather heavy
treatment.
The chapters in this volume start with describing and analyzing
ring compositions so that the reader has the tools for discovering
them, and for appreciating the causes of misreading. Antique ring
compositions are a precious heritage. There exist many more than I
have described, all liable to the rejection that I have noted. It is neces-
sary to work out what the maligned authors were originally saying. If
it is only to rehabilitate them the task is worthy.
I have tried to sum up a few compositional rules, but I know they
only apply to certain types of ring composition. If any two students
were put in non-communicating rooms, each furnished with a set of
these seven rules and a copy of a given ring document, I hope they
would come out agreeing on the pattern they had found in it. It is an
experiment that I admit I have never had the chance to try.
Pursuing this topic I discovered many new friends. In 1989 I had
the temerity to lecture on the book of Numbers. That year the Gif-
ford Lectures were held in Edinburgh in the Divinity School, and I
had the good fortune to have a chivalrous Bible scholar, Graeme
Auld, as my host. He is an indefatigable pattern spotter; from his
analyses of one biblical story plotted upon another story from an-
other Bible book, he has forged a powerful exegetical tool for chal-
lenging received ideas about the history and order of the biblical
texts. It is thanks to his learned example and his personal support that
I have gone on adventuring among ring compositions.
We need to reread them systematically with a view to making a
typology of the different kinds of rings, the places and periods they
flourished in. In this thought I have been inspired by the late Yehuda
Radday, who did so much of the basic work on biblical chiasmus. He
told me (in private correspondence) that he hoped that a typological
comparison of ring styles would be useful background for the contro-
versies on dating of Bible writings, a hope I share.
When I was at Northwestern University I had the privilege of
xi
PREFACE
hearing Wolfgang Roth’s lecture on the pattern of the miracles in

Mark’s gospel. It made me remember and never again forget that pat-
tern perception is one of the basic skills of anthropology. At that time I
had not started to work on the anthropology of the Bible. Back in
London after 1988, I gratefully acknowledge encouragement from
Mark Geller, head of the Institute of Jewish Studies in University
College. Directly on the topic of ring composition, by a stroke of good
luck I met Simon Weightman, of the School of Oriental and African
Studies. He himself was working with Aditya Behl on early Persian
poetry, both experienced in formal analysis. I have referred to their
exciting studies in this volume. And for invaluable help on figured
poetry (or picture poems) I am immensely grateful to Jeremy Adler.
Most of the friends who helped me with this book are colleagues
of long standing whose help is too pervasive to be picked out for
this specialized topic—for example, Wendy Doniger, who has talked
about it with me several times. But some are friendships made specifi-
cally on account of research into ring composition. Joan Pittock, for
example, I met by sheer luck, when I visited the University of Aber-
deen for a conference on the Bible. She is researching in eighteenth-
century literature, focusing particularly on the history of the Oxford
Chair of Poetry. She introduced me to the background of Robert
Lowth’s famous lectures on Hebrew poetry, whose influence on the
study of biblical literary styles, including ring composition, cannot
be overestimated. Again it was by luck that I first met my friend,
Milena Dolezelova, the Czech sinologist, also very interested in chi-
astic structures. She it was who introduced me to Chinese literary
conventions, including novels composed as rings. Geoffrey Lloyd
also helped me by suggesting that I should think of Chinese divina-
tion in connection with parallelism.
I am indebted to Alan Griffiths, Leonard Muellner, and Malcolm
Willcock for valuable advice on the chapters on the Iliad. My biggest
debt is to Simon Hornblower, who was, when we met at UCL in
1997, researching on ring composition in Herodotus and Thucydi-
des. He introduced me to John Myres’s account of ‘‘pedimental’’ writ-
ing. ‘‘Pedimental’’ means writing that goes up to a central point,
xii
PREFACE
makes a turn, then comes down step by step on the other side, like
wide-angled pediments on doorways. ‘‘Pedimental,’’ as another name
for a chiasmus, is usually applied to short pieces of writing, whereas I
am using the words ‘‘ring composition’’ for much longer texts. It is
like the difference between the decoration on a porch and the struc-
ture of the house.
The idea of a pedimental composition is clearly seen in Jacob Mil-
grom’s design of a biblical ring that embraces in its scope the whole
of the Pentateuch and the Book of Joshua. He has drawn a steeply
angled mountain labeled ‘‘The Theological-Literary Structure of the
Hexateuch’’ (Fig. 1). The left side, ascending from Genesis, is la-
beled with an arrow ‘‘From Slavery’’; on the peak, referring to Exo-
dus, is ‘‘Theophany, My Presence.’’ Coming down from the top on
the right-hand side the arrow points ‘‘To Freedom,’’ referring to
Leviticus, Numbers, Deuteronomy, and Joshua (four books are roll-
ing down the mountainside much more quickly than the first two
climbed up).
The ring is closed by the reference at the end to ‘‘Land promise
fulfilled,’’ Joshua (13–24), which matches ‘‘Land promised’’ at the
beginning (Genesis 12–50). It shows how the circle does what writing
in rings can do: raise the level of understanding.
It is a pleasure to record my debt in this book to Jacob Milgrom,
specially for help on the chapters on Numbers. Without his learned
support sustained over twenty years I never would have persevered
with the study of the Bible as literature.
Ending is different from completion, as I have explained in the last
chapter; the first is difficult, and the second impossible. I thank these
people for their inestimable help. It is such a pleasure to remember
them that I could go on for pages, as there are many more debts to
acknowledge.
I would not forget to thank my sister, Pat Novy, for her line draw-
ings of Abraham and Isaac, for the drawing of the scroll whose mean-
ing is in the middle, and for her ring diagrams that vindicate such
disparate materials being brought together under the one head of
‘‘Ring Composition.’’ I also thank Tom Fardon for taking time off
xiii
history
Primeval
Gen. 1–11
Fig. 1. A pedimental composition; the theological structure of the Hexateuch. Adapted from Jacob Milgrom, The JPS Torah
promised
Land
A Gen. 12–50
Fr
“Bones of Joseph”
om
circumcision
Sla
judged B
Egypt
A
pesah
ve
B Exod. 1–12:36
ry
“Put off your shoes...holy”
Egypt
from
Exodus
C Exod. 12:37–15:21 splitting the sea
Commentary: Numbers (Philadelphia: Jewish Publication Society, 1990), xix.
C
three days
wandering D
Wilderness
manna, quail, water
D Exod. 15:22–18:27
fire
regulated
ratified and E
Covenant
encroaching
E Exod. 19–24
sacred architecture
planned F
Sanctuary
F Exod. 25–31
Sabbath law
broken G
Covenant
G Exod. 32
Theophany
MOUNT
SINAI
X Exod. 33 “My Presence”
X
G' renewed
Covenant
G' Exod. 34
F' built
Sabbath law
Sanctuary
F' Exod. 35–40

E' regulations
Covenant
sacred architecture
E' Lev. 1–Num. 10:10

D' wandering
encroaching
fire
Wilderness
D' Num. 10:11–36:13

three days
manna, quail, water
Covenant
renewed
Deuteronomy
C' into
To
Canaan
Entry
splitting of the Jordan

C' Josh. 1–4
Fr
ed e
B' judged
om
“Put off your shoes...holy”

Canaan
B' Josh. 5–12 pesah

circumcision
A' promise
fulfilled
Land
A' Josh. 13–24

“Bones of Joseph”
PREFACE
from his school leaving examinations and for giving invaluable skill in
the tricky work of preparing the typescript for the publishers. I also
extend thanks to my friendly neighbor, Colin Donne, for computer
diagrams. Of course I gratefully acknowledge the publishers them-
selves, notably Jean Black and her team, Laura Davulis for her pa-
tience in the work of turning this typescript into a book, and Joyce
Ippolito for her extreme care and attention to the text.
xv
one
ANCIENT RINGS WORLDWIDE
A new interest in ring composition has lately arisen. This antique

literary form is being discovered in documents that the scholars have
known for centuries and have translated without recognizing that
they have any formal structure. Many fine old texts have been dis-
dained and disrespectfully mauled in the effort to get to the sense.
What a shame, and what dull and trivial interpretations have been
piously accepted in default! And how ready the commentators were to
lay the perceived incoherences to the door of weak writing skills, or
even weak intelligence. Writings that used to baffle and dismay un-
prepared readers, when read correctly, turn out to be marvelously
controlled and complex compositions. Learning how they were con-
structed is like a revelation, with something of the excitement of
hidden treasure. Now is a good moment for the effort of rereading.
Various disciplines are taking up the task.
Glenn Most has pointed out a paradox in the reading of Pindar. In
antiquity no one thought his poetry was particularly ‘‘difficult,’’ but in
modern times he is seen as ‘‘the very paradigm of poetic difficulty . . .
an esoteric poet in whom clarity of thought . . . is mantled over by so
thorough an obscurity of expression that the meaning of individual
phrases is already often impenetrable and the organization of poems
as wholes scarcely imaginable.’’∞ The stories of misreadings that I
could tell are nearly as shocking.
The minimum criterion for a ring composition is for the ending to
join up with the beginning. Chapter 2 gives more detail. A ring is
a framing device. The linking up of starting point and end creates
an envelope that contains everything between the opening phrases
and the conclusion. The rule for closing the ring endows the work
with unity; it also causes all the problems that another set of rules
has been designed to solve. It takes skill to compose a polished speci-
men. There has to be a well-marked point at which the ring turns,
1
preparatory to working back to the beginning, and the whole series

of stanzas from the beginning to the middle should be in paral-
lel with the other series going from the middle back to the start.
Each section on the second side of the ring corresponds to a match-
ing section on the first side. The form is well known. It is basically
the chiastic structure, ABBA, or ABCBA, a form that pervades the
Bible and other famous archaic texts. It comes in many sizes, from
a few lines to a whole book enclosed in its macro-envelope, ar-
ranged throughout in intricately corresponding parallelisms.≤ It is
called either inclusio, emphasizing the bracketing into one unit of
everything from the start to the end, or chiasmus, emphasizing the
inverted word order.
Jacob Milgrom takes a fine example of phrase inversion from the
Book of Numbers.≥ On the march of the Israelites to the Jordan, two
tribes see land where they would like to settle. They ask Moses, but he
refuses to let them take any land for themselves until the whole jour-
ney is accomplished; he requires the warriors of Israel to fight side by
side all the way to the final victory. First they ask Moses, saying:
A We will build here sheepfolds for our flocks
B and cities for our little ones. (Num. 32.16)
He expostulates, but when they promise to stay fighting with their

comrades until victory, he concedes their wishes, reverting to their
own words, but in reversed order.
B% Build cities for your little ones
A% And sheepfolds for your flocks. (Num. 32.24)
These are very short examples. What I mainly have in mind when
I refer to ring composition is the large-scale, blown-up version of
the same structure. A single ring is inclusive enough to comprise
the whole of the book of Numbers; other rings include equally emi-
nent and even lengthier texts. The small version can depend on re-
peated key word clusters for recognizing the structure. The macro-
composition has to organize whole verses and paragraphs first in one
ordering and then again in reverse order. This form is readily appreci-
2
ated and understood by scholars when it includes only a few sen-

tences. The large compositions constructed in this way, however,
have had a bad reception in the West. The ring form has had a history
of misunderstanding and disregard.
Holding the Oxford Chair of Poetry, Bishop Robert Lowth gave
thirty-four lectures in Latin on the sacred poetry of the Bible.∂ He
distinguished four types of Hebrew parallelism, and named the fol-
lowing varieties:
Synonymous parallelism: poems chanted by two groups, repeating back
and forth:
When Israel went out of Egypt

The house of Jacob from a strange people. [Psalm 116]
AB/A%B%
(The words are not the same, but they are synonymous: Jacob’s other
name was Israel, Egypt was indeed a strange people to Israel.)
A If only we had died

B in the land of Egypt.
B% or in this wilderness
A% If only we had died. (Num. 14.2)
Antithetic parallelism: an image illustrated by its contrary, in the same

exactly repeated pattern, but sentiments opposed to sentiments, words
to words, singulars to singulars, plurals to plurals.
Example from Proverbs 27.6–7:
The blows of a friend are faithful, but the kisses of an enemy are
treacherous.
The cloyed will trample on a honeycomb, but to the hungry every
bitter thing is sweet.
Synthetic or constructive parallelism: the parallel lies only in the con-

struction, for example, using numbers, triplets, and detailed balancing
of syllable with syllable.
Example from Deuteronomy 32.2:

My doctrine shall drop, as the rain.
My word shall distill as the dew.
3
Or antithetic and synthetic parallelism can be combined.
Example from Isaiah 9:

The bricks are fallen, but we will rebuild them with stone.
The sycamores are cut down, but we will replace them with cedars.
ABC B% A&.
Bible scholars do not use Lowth’s terminology nowadays, but they

do write about parallelism, introverted parallelism, and chiasmus or
chiastic structure, in which a pair of items reverses itself, yielding the
pattern ABB%A%. Milgrom calls it ‘‘chiasm in subsequent repetition.’’∑
After Lowth had spoken, a tacit presumption grew up to the effect
that chiastic structures are essentially part of the Semitic literary heri-
tage. They were not explicitly assimilated to ring composition in the
Greek classics until W. A. A. van Otterlo’s systematic study of ring
composition in Homer. There was soon as much to say about Greek
ring form as about Semitic chiasmus. The two fields, biblical and
classical, continued on their separate ways, which was very reasonable
as each had to do with evolved and regionally distinct literary con-
ventions, each having its own history. Different or additional terms
were used by Greek scholars; paratactic contrasted with syntactic, the
parallelism with the punch at the end is called ‘‘priamel.’’ For a bibli-
cal example:
Saul has smote his thousands,
And David his ten thousands. (1 Sam. 18.7)
Saul was very angry when he heard the women singing this.
The problems and the need for specialized terms arise when we
discover that this literary form is not confined to two cultural regions
but is universally present in archaic literary forms as well as in con-
temporary folkloric recitals. Roman Jakobson, the great philologist,
surveyed the forms of parallelism among the peoples of the Ural-
Atlantic area, where Finnish oral poetry offers the classic case. He said
he was astonished to realize that the strong presence of parallelism
‘‘seemed hardly to interest the specialists in Russian folklore.’’∏ Fol-
lowing in his footsteps, the anthropologist James Fox has collected
4
further evidence of it in major collections of texts from the Aus-

tronesian peoples, central Rotinese, the Celebes, the islands of east-
ern Indonesia.π Parallelism appears in millennia-old Chinese poetry
and in the ancient Hawaiian creation chants. Speakers of Papuan
languages in Timor and New Guinea use it. Parallelism remains dom-
inant in Chinese literature, it is a common literary form in Borneo and
Madagascar and in Vietnamese parallel poetry, and similar traditions
exist among the Burmese and Thai.
Roman Jakobson described parallelism broadly as ‘‘a system of
steady correspondences in composition and order of elements on
many different levels: syntactic constructions, grammatical forms and
grammatical categories, lexical synonyms and total lexical identities,
and finally combinations of sounds and prosodic schemes. This sys-
tem confers upon the lines connected through parallelism both clear
uniformity and great diversity. Against the background of the integral
matrix, the effect of the variations of phonic, grammatical, and lexical
forms and meanings appears particularly eloquent.’’∫
In the same chapter Jakobson quoted with approval Gerard Manley
Hopkins saying in 1865: ‘‘The artificial part of poetry, perhaps we
shall be right to say all artifice, reduces itself to the principle of paral-
lelism.’’Ω For his own example of parallelism Roman Jakobson chose
Edgar Allan Poe’s lines:
To One in Paradise:
And all my days are trances,
And all my nightly dreams
Are where thy dark eye glances
And where thy footstep gleams.∞≠
In these lines, day and night are complementary, and so are eye and
foot, and dark and gleaming. There are some delicate transpositions
when darkness is attributed to the dancer’s eye, and the gleaming to
her foot (presumably she has jewels on her toes). These poetic exam-
ples are pleasing parallels; they are not rings. There is no crossing
over halfway through the series and no matching of the beginning
with the ending, no ring form.
5
Ring composition is parallelism with an important difference. It is

based on parallelism in the straightforward sense that one section has
to be read in connection with another that is parallel because it covers
similar or antithetical situations, and some of the same vocabulary acts
as cues to the pairing. But the parallel sections are not juxtaposed in the
texts. They must be placed opposite each other, one on each side of the
ring. The structure is chiastic; it depends on the ‘‘crossing over’’ or
change of direction of the movement at the middle point. A ring is
much more interesting than a succession of simple parallels that are
simply laid out consecutively. If ring composition were just a local lit-
erary form, like the sonnet adopted as a favored convention, or like a
genre used for particular kinds of expression, elegiac, comic, or tragic,
there would not be any big puzzle. Jakobson has launched a challenge
by describing thinking in parallelisms as a faculty inherent in the rela-
tion among language, grammar, and the brain. If this is so, it would ex-
plain why the literary form based on it is so widespread. We should ex-
pect it to rise up anywhere, at any time, and it does, with many variants.
The local variations should not disguise the fact that the ring form
appears to be extraordinarily stable over time. That should count as a
point supporting Jakobson’s theory about parallelism, by giving evi-
dence that the same formal structure governs very ancient composi-
tions. Zoroastrian literature includes seventeen poems (collectively
known as the Gathas) that are attributed to Zoroaster himself.∞∞ They
are found to correspond in their formal structure with other better
known compositions, such as the Iliad and the book of Numbers,
dating probably from between the eighth to the sixth centuries bce.
Greater antiquity is indicated by the forms the Gathas have in com-
mon with the Rig Veda. Sections of the composition are ordered in
parallel stanzas, laid out in a single sequence until at a well-marked
midpoint the sequence stops, turns around, and reversing the order of
the stanzas returns the way it came, making an inverted parallel with
the first sequence. So the simple chiastic pattern sustained through
the centuries is AB C BA.
It is thanks to the relatively recent analyses of Martin Schwartz
(who thoroughly systematized and greatly elaborated Hanns-Peter
6
Table 1. Zoroastrian hymns or gathas from the thirteenth

to the tenth centuries bce.
(I a) (I b) (I c)
1—11 1—11 1—12

2—10 2—10 2—11
3—9 3—9 3—10
4—8 4—8 4—9
\5 — 7/ \5 — 7/ 5—8
\6 / \6 / 6—7
Schmidt’s seminal observations of elements of ring composition in

the Gathas of Zoroaster (Zarathushtra) that we can now appreciate
the great technical complexity of the ring structures of this ancient
poetry.∞≤ Verbal markers based on line and syllable count make links
not only at line endings but systematically through all the stanzas of a
composition.
Verses are put into correspondence with one another, echoing
words and themes like rhyming structures in poetry, exploiting dou-
ble meanings and puns. More complex patterns draw attention to the
mid-turn stanza by nesting it in a threefold or even fourfold chiasmus
of its own. Even more complex patterns link poems with each other in
the same way, so that the poems once regarded as separate units can
now be read as connected sequences. In 2003 Martin Schwartz re-
called: ‘‘By about a decade ago I realized that in every Gathic poem all
the stanzas are concatenated concentrically in accord with one of
several patterns of symmetry, and that for each poem the central
stanza(s), which are thematically significant, concatenate(s) with the
immediately preceding and following stanzas, and, more importantly,
with the first and last stanza.’’∞≥ In these words he has named the three
basic principles of ring composition. Of these, the most important is
the loading of the main message onto the central stanzas. Evidently in
this literary convention, the interest lies in the elaborate word play.
The semantic content of the poems seems limited and simple by
contrast. In that period technical virtuosity counted for much more.
7
We can go even further back in time to discover ancient Chinese

forms of parallelism.∞∂ In China in the eleventh millennium before
Christ a form of divination was practiced that involved reading from
marks on the shell of a tortoise (or turtle, either word).∞∑ Parallelism
has always been prominent in Chinese poetry, philosophy, and litera-
ture.∞∏ Vandermeersch relates the dominance of parallelism in Chi-
nese thought to the great importance of tortoise divination from early
times.∞π Later it came to be absorbed into the cosmological pattern of
the Yin and the Yang, which allocates every matter and its paired
opposite into a philosophical dichotomy. The tortoise is taken to be
a natural symbol of the cosmos; its carapace is round like heaven,
underneath it is flat like the earth. The tortoise’s extreme longevity
puts it apart from other creatures. It has strongly marked right and
left sides (Fig. 2).
The five horizontal lines represent the five elements: water, iron,
earth, fire, and wood.∞∫ The diviners have systematically divided up
five areas of the whole shell: the head end, the tail end, and three areas
between the head and tail on each side, each with its own cosmic
meanings. The diviners scorch the underside of the shell with a red-
hot poker, making the surface crackle. During a consultation they
would scrutinize the patterns of crackle marks on the shell with two
written statements set up as binary contradiction. Crackles on the
right side of the tortoise shell were read as positive responses, those
on the left side as negative, in relation to the written statements. The
patterns affirm or deny. It is sometimes a divination of a future event,
sometimes a statement about the past or about the moral law, always
highly technical and systematic, enjoying very great prestige.
The medieval Chinese novel was commonly dismissed by Chinese
critics for lack of structure. To those unfamiliar with the tradition, a
long story about highwaymen killing and plundering gave the im-
pression of being the same story monotonously repeated over and
over again. Hua I. Wu says: ‘‘many modern readers are disappointed
and even annoyed by such repetitions and apparent clumsiness and
until very recently the prevalent opinion was that The Water Margins
and novels like it are episodic and lacking a coherent construction.’’∞Ω
8
[To view this image, refer to

the print version of this title.]
Fig. 2. Ancient Chinese divination on the tortoise shell.

Adapted from Léon Vandermeersch, ‘‘Les origines divinatoires
de la tradition chinoise du parallélisme littéraire,’’ Extrème-
Orient-Extrème-Occident 11 (1989): 25.
‘‘Repetition,’’ ‘‘clumsiness,’’ ‘‘episodic’’! The complaint of incoher-

ence will soon begin to sound familiar to my readers. In the seven-
teenth century the construction of the Chinese novel was revealed; it
turns out to be based on elaborately entwined parallelisms. So the
previous negative opinion was revised.
A fine survey of pre-modern Chinese literature summarizes its very
old tradition of literary theory, starting in the thirteenth century in
the form of theoretical prefaces to plays and novels, much influenced
by an even older traditional theory of art, painting, and music.≤≠ The
theory of literature seems to have developed in the seventeenth cen-
tury as an important genre. It generated canons of poetry and elegant
writing and instigated intense debates on the relation of art to mor-
als and politics. Jin Shentang, a well-known dramatist and critic of
the time, particularly emphasized parallelism as a structuring princi-
ple. He and his colleagues developed the thesis ‘‘that the novel is a
finely constructed whole . . . a structure made up of components—
themes, episodes, characters, sentences, words—that mutually reso-
nate on both the syntagmatic and the paradigmatic levels of the text,
9
endowing it with rhythmic dynamism and dramatic tension. The

novel is thus a highly complex organization of a stratified character
with multiple relationships whose patterns are set up at the text’s
beginning by a small unit (for example, a prologue or the first chapter)
that is repeated, with textual variations, in larger units throughout the
whole text.’’≤∞ This is well worth quoting because it anticipates what I
want to say about ring composition in general, and in particular about
the concept of repletion that I develop in Chapter 10.
There are many instances of misunderstanding, of a scholar despis-
ing or rejecting an antique text because of its alleged lack of order and
syntax. In this volume I quote several examples of negative judgments
made by learned commentators on ancient texts. The various cases of
error are either due to the critic having missed altogether the internal
structure of alternating parallels or due to their having missed the
central place where the keys to the main theme are gathered together.
The Literary Guide to the Bible is a sophisticated book and a great
achievement on practically every level.≤≤ But it is a pity that only a few
of the contributors to a book with this title are interested in the Bible’s
dominant literary construction, parallelism. In a very fine chapter at
its end, on ‘‘The Characteristics of Ancient Hebrew Poetry,’’ Rob-
ert Alter makes a close analysis of parallelism in parts of particular
poems, sometimes the psalms, and also short poetic structures, as
from the book of Job or Samuel. Joel Rosenberg’s impressive analysis
of the prophets Jeremiah and Ezekiel is another exception.≤≥ He de-
ploys the symmetries and parallels of the prophetic discourses to
reveal the meanings of famously puzzling texts. He even summarizes
the formal pattern of the whole book of Jeremiah in a clear chiastic
structure. I am not able to judge the soundness of the analysis, but I
regard the attempt as a major step toward replacing our own with
contemporary literary judgments. Most of the other chapters apply
modern literary standards, focusing on plots, themes, and person-
alities. They rarely show interest in the macro-structure of a long
text. When they do say something about parallelism it is concerned
with very short pieces, a few lines.
10
As I see it, the tracing of the total scheme from beginning to end
and the concern for a coherent pattern would correspond to the bibli-
cal authors’ own preoccupations, and implicitly to their criteria of
literary excellence. I am particularly glad to read Rosenberg’s re-
marks on earlier commentators who have found the texts elusive:
‘‘The Structure of Jeremiah, and especially of its apparently cha-
otic chronology, has proved elusive to critical investigators, many of
whom have declared the text to be in disarray and have attempted a
reconstruction of an ‘original.’ ’’≤∂ The terms disarray and chaotic, to-
gether with disordered, clumsy, and other pejoratives, will crop up very
often as we proceed.
To show that these negative judgments on structure are not solely
due to a blind spot in Western literary criticism, I quote the Persians’
comments on the structure of the chef d’oeuvre of their own great
mystic, Rumi. Simon Weightman has recently described the scholarly
reception of Rumi’s masterpiece, the Mathnawi.≤∑ The response to
this famous mystical writing exactly echoes the response of certain
biblical commentaries—for example, Martin Noth’s misreading of
the book of Numbers. Rumi’s book is renowned, the commentators
revere it, but they find fault with the author for failing ‘‘to keep the
discourse on orderly lines.’’ They complain that it consists of uncon-
nected anecdotes, it is disjointed, and it is full of long digressions.
Weightman mentions it being called ‘‘rambling,’’ ‘‘a trackless ocean,’’
lacking in ‘‘any firm framework.’’ But in the case of religious mysti-
cism, to discount the structure is not a criticism, and these remarks
are not meant to be disparaging. On the contrary, the disorderly style,
as the critics take it to be, is all the more esteemed because it is
supposed to indicate a spontaneous flow of inspiration.
Whenever I read criticism of dire editorial confusion, my pulse
quickens; I scent a hidden structure, probably a ring composition. In
the case of Rumi’s Mathnawi my gut response is vindicated. A doctoral
student at the School of Oriental and African Studies, Seyed Ghahre-
man Savi-Homani, has closely examined the structure of Rumi’s text.
He demonstrates that it is not at all disorderly. It obeys very precisely
11
the rules of a well-conceived and well-executed ring composition as I

describe them in the next chapter.≤∏
Why is ring composition practiced all over the world? What is
it for? So many people! So many epochs! They could not have all
learned it from one another. Its robustness over thousands of years
supports the theory that something in the brain preserves it, and yet
we know that it can fade out so completely that new readers miss it
altogether.
Among speculations about the origin of this complicated rhetorical
system, a common one is that it comes from pre-literate times where
it was a necessary help to the memory of the bards. It is true of course
that literacy comes later. It is also possible that reciting or writing in
parallels may be good for memorizing. Another theory suggests that
it is based on very ancient ‘‘finger rhymes.’’ These are games and
verses using the five fingers of one hand to make the statement in five
steps, and the fingers of the other hand for elaborating or balancing it,
and bringing the two hands together at the end.≤π Another suggestion
is that it has a common origin in very early times with acrostic com-
position. Whatever its pre-literate origins, ring composition has of
course been transformed with the advent of writing. For my own
purposes, the question of an origin in oral literature can be left to one
side. It does not need to be settled for considering the main questions
about ring composition.
The arrival of literacy does not make parallelism atrophy and dis-
appear. The prolific biblical examples of parallelism come from a very
literate civilization. Albert Lord’s pioneering fieldwork among illiter-
ate Yugo-Slav peasants showed that trained bards had astonishingly
long and accurate memories, though they were not interested in per-
fect word-for-word recall.≤∫ So far from requiring a mechanical mem-
orizing of lines, even if the singer thinks he is repeating a story ver-
batim, the singing of it is essentially a creative interaction between the
current performer and the traditional song.
Jack Goody’s theory about memorizing is relevant. He associates
learning by heart with early literacy; once literacy begins to be wide-
12
spread it changes the whole basis for social life by redistributing

power in favor of the literate. One of the effects is the establishment
of schools for learning literate skills and for memorizing important
texts; consequently, learning by rote is a characteristic of literate so-
ciety. Evidence exists for Egyptian schools as early as the third mil-
lennium bce. Scholars would have spent their time reading, writing,
and reciting.≤Ω
I am more concerned to emphasize ring composition’s exegetical
function. It controls meaning, it restricts what is said, and in doing so
it expands meanings along channels it has dug. Though it never com-
pletely escapes ambiguity, writing in a ring puts various strategies at
the writer’s disposal; when he chooses one path or another he ties the
meaning into a recognizable, restricting context. Ring composition is
not poetry but it puts syntactic-like restrictions on the writer. It is
worth pausing to reflect on why restraints are necessary.
It is not just a matter of syntax, though syntax is confusing enough.
In our London street a notice board tells us that ‘‘dog-fouling’’ is an
offense. Usually, in English, when a noun qualifies a participle the
noun is the object of the verb, as in wife-beating, trout-fishing, mak-
ing it easy to use ‘‘wife-beater’’ or ‘‘trout-fisher’’ as the agent. Nor-
mal syntax implies that we should be looking out for some malefactor
who is fouling our dogs. The trouble with words is that they do not
stay the same. ‘‘It was vulgar at the beginning of the nineteenth cen-
tury to mention the word ‘handkerchief ’ on the French tragic stage.
An arbitrary convention had decreed that tragic personages inhabit a
world in which noses only exist to distinguish the noble Roman nose
from the Greeks and Hebrews, never to be blown.’’≥≠ Words inter-
penetrate, as Ferdinand de Saussure liked to show. Meanings never
stay the same, analogy is wild and willful, ambiguity thrives as the
words tumble vertiginously through the sounds; they mix, they riddle
and jump.≥∞
Written words lack the support of body language, voice, and physi-
cal context. The ring convention does something to fill the interpreta-
tive gap by virtue of its symmetry, its completeness, and its patterned
13
cross-referencing. A citation from Douglas Hofstadter brings home

the problems solved by the ring form: ‘‘A word being the name of a
concept, and a concept being a class of items linked by analogy, and
people by nature being creative and ever finding new analogies, a
word’s connotations are consequently oozing continually outwards to
form an ever-larger and blurrier nebula as more and more analogies
are cognized as legitimate and welcomed by the culture. A table thus
acquires legs, a mountain acquires a foot, ships venture into space,
sopranos sing high and basses low, books have jackets, families have
trees, computers have memories, salad is dressed, wine breathes, cars
run, hearts dance, a storm threatens . . . ,’’ and so on.≥≤
Analogies are endless; as a pattern of analogies a ring composition
constrains the multiple meanings of words. It does so by giving each
stanza or section its parallel pair; the members of a pair are placed on
opposite sides of the ring so that each faces the other; each indicates
its pair by verbal correspondences. Thus two constructed likenesses
have been selected and polished and carefully matched so as to guide
the range of interpretation. You can compare the functions of ring
composition to syntax: it tames wild words and firmly binds their
meanings to its frame. Another function is greatly to deepen the
range of reference by playing on the double meanings of words. An-
other of its benefits is that it is a form of play; it gives the pleasure of a
game to the composer and the reader.
Some examples will show how deceptively complex a seemingly
simple ring can be. I will start with a very familiar story from the book
of Genesis, the Garden of Eden, which Cassuto identified as one of
the little rings of which Genesis is composed.≥≥ The more the story is
already familiar, the more the sense of surprise at the density of mean-
ing that is packed into this form. The story seems to flow so smoothly
that the strong structure of its background environment is unsus-
pected. It starts in an amorphous, misty, barren place that contrasts
with the charming garden full of trees that God willed into being.
The story begins from before the point at which God has created
man and woman (1.26–31). At first, there was no man to till the
14
Gn 2:5 There was no ādām to till the ădāmah.

2:7 God formed the ādām from the dust of the ădāmah,
2:15 to guard and till the garden of Eden.
2:17 The ādām was not to eat of the tree of knowledge.
2:21–24 To allay his loneliness God brought a partner out of the ādām.
The ādām is now ı̄sh and ishshah, united in perfect harmony.
2:25 They are naked (‘ărūmı̄m) and free from shame.
3:1–6 But the subtle (‘ărūm) serpent tempts them to disobey.
3:7–10 They become aware of their nakedness and feel shame.
3:16 Their harmony is spoilt: hereafter ı̄sh will dominate ishshah.
3:19 The ādām must return to the ădāmah from which he was taken.
3:22 The ādām may not eat of the tree of life.
3:23 The ādām must till the ădāmah but with toil and sweat, no longer in Eden;
3:24 Eden now has a new guardian to keep the ādām from the tree of life.
Fig. 3. The tragedy of Adam (human being) formed from the adamah (soil).
Courtesy of Robert Murray.
ground. God made a man from the earth, breathed life into him,
and set him in a garden with four rivers in and around it (2.4–3.24).
The concept of water channeled into rivers contrasts with the ini-
tial undifferentiated swampiness. The theme seems to be very sim-
ple in English, but there is a punning play on the Hebrew word for
‘‘soil’’ (adamah) and on the Hebrew word for a human being named
‘‘Adam.’’
Robert Murray has selected the resonances between these two
words and shown how they underpin the progress of the story (Fig.
3).≥∂ The panel presentation shows up the central place very clearly,
verses 2.25 to 3.7–10. It is a lesson to teach us to expect subtle sophis-
tication and to realize how much we miss in the Bible when we try to
read the apparently simple stories without knowing the sounds of the
Hebrew words. Note that the subject matter, the theme of return, the
return of Adam to the soil, is exemplified in the ring convention itself,
the end returns to the beginning. Adam was made from the soil.
Except for a slight displacement from 3.19 on, this is a clear
15
example of a pattern that occurs in a number of places in the Bible;

perfect form would be:
a
b
c
d
e
d*
c*
b*
a*
The effect is to give special emphasis to the pivotal central point. In
the Eden story, this place is occupied by the intervention of the ser-
pent. The serpent’s malice is highlighted even more by the play on
two words quite different in meaning but sounding very like each
other, which come in adjacent verses (2.25 and 3.1—chapter divisions
are medieval, not original, in the Hebrew Bible). These words refer
respectively to the nakedness of the human pair and the slyness of
the serpent. But note (very important): Hebrew has two words for
‘‘naked’’: one is used in contexts of sexual activity, the other, used in
2.25, always connotes vulnerability—of children and other helpless
persons.
This is a superb example of a very short ring, a small ring inside a
larger one. It is a blessing to have its subtleties displayed by a Hebrew
scholar. I hope it whets the appetite for the macro-composition of a
very large ring, such as that of Genesis or Numbers.
16
two
MODES AND GENRES
Literature is institutional; institutions establish stereotyped forms

of behavior, and literature itself contributes to the selection and ste-
reotyping process. It is useful to be able to think of genres as dis-
tinctive institutional forms within literature, among other ‘‘literary
kinds.’’∞ They coordinate with non-literary institutional forms, set up
ideals for behavior, and justify them. The readers, themselves embed-
ded in institutional life, need to recognize the genre they are reading,
so it has to make itself distinctive. Assigning a work to a genre gives
the readers expectations about a literary piece, and they learn to judge
whether it is a good sample.
Frank Kermode would modify the sharpness of definition that
would follow from calling a genre a ‘‘literary kind’’: ‘‘A genre is not
what used to be known as a kind, with rules prescribed by institutional
authority; it is a context of expectation, ‘an internalized probabil-
ity system,’ ’’ he says, quoting a nice phrase of Leonard Meyer’s.≤ A
genre, Kermode goes on to say, ‘‘is the set of expectations which
enables us to follow a sentence as it is spoken.’’≥
There are rules. They are not imposed from outside of the literary
work. They are not there first. They emerge from the first completed
works. New genres are always in the making; the breach of one rule is
enough to introduce a new variant. They overlap, and mix; some last a
long time, some die out. It was only half a century ago that the Dutch
scholar W. Otterlo produced the first systematic analysis of ring com-
position.∂ He noted that ring composition has passed unnoticed over
the centuries until our time. His book was focused on Homeric stud-
ies, to which I have occasion to return in Chapter 8. He counts ring
composition as a late archaic style that finishes around the middle of
the fifth century bce. This view is endorsed by G. B. Gray, who
thought that it had been ousted by metric forms.∑ Ring composition
17
MODES AND GENRES
did become obsolete. How it came to be ousted is one of the central

(not wholly answered) questions of this study.
Otterlo was interested in the variety of circular structures in the
Iliad. The Iliad has rings with only one ‘‘member,’’ and rings with two
or three ‘‘members.’’ His classification of structures is chiefly con-
cerned with the relations of the minor rings to the major ring in
which they are enclosed. In the latter a main ring keeps track of the
thread of the narration, picking it up again after a digression. Otterlo
cites as an example the travels of Odysseus interrupted by such digres-
sions as the hunt of Autolycus, a clearly circumscribed event, after
which the poet returns to the main ring and resumes the tale of the
journey.∏ Another structure is a main ring that includes minor rings
containing an argument about or explanation of what is narrated.
These interludes are not digressions but supplements. He goes on to
identify different kinds of rings within rings.
Classification can be extended to the difference between the struc-
ture of the external enclosing ring and the structure of the rings
within rings, which are predictably much shorter. I would suspect that
it might be easier to make a little ring in which clusters of key words
will be enough to indicate the matching parallel sections and to bring
them into position. But a very elegant small ring, like that in the
Garden of Eden, which I discussed in Chapter 1, cannot be easy at all.
At least we can say that long pieces need more structure and more
conspicuous signaling of the structure. In a short piece the word-for-
word echoing is enough to match the ending to the beginning. A long
composition often has to add something more to knot the ending
firmly to the start. It often has an epilogue (or ‘‘latch’’) added after all
the paralleling is done.
It is time to give examples. First, for a small ring, the heart-
wrenching story of the binding of Isaac will illustrate Otterlo’s inter-
est in the ring within a ring. The story of Isaac’s binding is replete
with meaning, moral, theological, and emotional, and has been con-
troversial through the ages. The literal text carries various ambi-
guities, especially concerning the intentions of God in commanding
18
MODES AND GENRES
Abraham to sacrifice his own son. In secular understanding both God

and Abraham have very bad coverage. How can God be good if he
makes such a grotesque demand? How can Abraham be good if he
obeys it? Jonathan Magonet puts it as crudely as I could dare: ‘‘If we
read chapter 22 alone, as unfortunately too often we do, then, if we
are honest, we must conclude that God is as mad as Abraham who
would obey such a God.’’π
A huge statue of Abraham, blade in hand, standing grimly over his
young son, stands on the campus of Princeton University. Created by
the sculptor George Segal, it is called ‘‘In Memory of May 4th, 1970,
Kent State: Abraham and Isaac.’’ Kent State is a university in Bowl-
ing Green, Ohio. When Richard Nixon announced the bombing of
Cambodia in April 1970, student protests immediately followed. The
National Guard was called out at Kent State, and on May 4 guards
shot and killed four students.∫ That Abraham should have been cast in
this brutal role in the memorial to the students’ revolt testifies to the
common lay interpretation (misinterpretation) of the story of Isaac. It
is utterly inept: Isaac never revolted; he was to be a consenting victim,
and Abraham never killed him.
The text for the binding of Isaac (Genesis 22) is summarized and
arranged as a chiastic structure in Figure 4. Interpretation must focus
on the question of how much Abraham understood when making
preparations to obey. Many Western commentators expatiate on the
anguish in the old man’s heart and on the cruelty of a God who could
lay such a brutal command on a loving father.Ω The characters of
both Abraham and God show in a quite different light if Abraham
understood what was to happen. If Abraham did know that he would
never have to carry out the command, then literary, psychological,
and theological speculations about his anguish are beside the point.
Jewish traditions take other lines. Abraham’s total commitment to
God is taken for granted. Some emphasize his unquestioning obe-
dience. There is also a tradition that emphasizes Isaac’s own consent,
even his heroic desire to be the victim in the sacrifice of atonement.∞≠
Sebastian Brock presents a Syriac text from late antiquity that focuses
19
1.
God tested
Abraham & said to
him: ‘Abraham!’
and he said: ‘Here
G I am’
2. ‘Take your son, your only son, 15–18. The Angel of the Lord
Isaac, and offer him as a burnt called to Abraham a second time
offering on one of the mountains from heaven & said: ‘Because you
of which I shall tell you. . . ’ have done this and have not
withheld your son, your only son, I
will bless you indeed. . . .’
3. Preparations for going to the
place of which God had told
him. . . the Ass, two young men,
the wood for the burnt offering
and Isaac. 14. Abraham called the name of
9 4. Abraham lifted up his eyes and the place ‘God will provide.’
saw the place afar off.
5. . . . told the young men to
stay. . . .
6. Took the wood for the burnt 13. Abraham took the ram . . .
offering, laid it on Isaac, took in offered it as a burnt offering,
his hand the fire & the knife. instead of his son.
7. So they went on both of them
together:
Isaac: ‘Father’
Abraham: ‘Here I am my son’
‘Behold the fire & the wood,
where is the lamb for a burnt 7
offering?’
8. ‘God will see to the lamb for a 13. Abraham lifted up his eyes
burnt offering, my son.’ and behind him was a ram.
So they went on both of them
together.
9. When they came to the place of
which God had told him,
Abraham built the altar, laid on
the wood, bound Isaac, laid him 12. ‘Do not raise your hand
on the altar and the wood. against the lad . . . for now I know
10. Abraham raised his hand, that you fear God, you have not
took the knife to slay his son. withheld your only son.’
The Angel of the G
Lord called to him
from heaven:
‘Abraham,
Abraham!’
‘Here I am’
11.
Fig. 4. The binding of Isaac, Genesis 22.1–18.

MODES AND GENRES
on Isaac’s mother, Sarah.∞∞ As he is setting out, she knows Abraham’s

intention and actually gives her assent, but she does not believe for a
moment he will kill his son. At parting, she tells Isaac,
If God desires you for life, he will give orders that you live; he who is
the immortal Lord will not kill you.
Now I shall boast: having offered you as a gift from my womb to him
who gave you to me, I shall be blessed.
Go then, my child, be a sacrifice to God, go with your father—or
rather your slayer. But I have faith that your father will not become
your slayer, for the saviour of our souls alone is God.∞≤
All the ambiguities of the story are caught by this speech.

On my reading of the ring, this is why and how Abraham earned
the blessing he gained for his response to God’s command, not for
blind obedience but for unswerving confidence in God. Differing
with respect from Kierkegaard’s interpretation, it was not at all ab-
surd to expect that God would never go back on his given word; the
contrary would be absurd. God had unequivocally told Abraham that
Isaac would be the progenitor of a people who would belong to him
and whom he would protect. All the promises to Abraham of a num-
berless progeny from Sarah depended on this child’s survival because
it was already a miracle that at her advanced age she should have given
birth at all, and clearly years had passed since then and she was not
going to have any more sons. Should any disaster overtake Isaac, it
would make nonsense of God’s covenant with Abraham, which is at
the very center of the religion.
A lesser man than Abraham on receiving this command to sacrifice
the boy might suppose that God had decided to cancel plan A and was
now preparing for plan B by getting rid of Isaac. But a man of such
faith as Abraham’s could not have forgotten the promises whose ful-
fillment required Isaac to survive, and he believed in a God who was
true to his word.
Against the view that Abraham knew that he would not be required
to sacrifice his son, some scholars would argue that nothing in the text
of Genesis 22 advises the reader that Isaac was not going to be killed,
21
MODES AND GENRES
and that the text does nothing to show that Abraham knew it. A close
look at the ring structure quenches this criticism, however, and the
ambiguity and uncertainty fade away. First, note how strongly the
ring syntax puts the story in a frame of loving fathers and sons. It sets
the mood for a touching scene. God refers to Abraham’s beloved son;
Abraham’s filial relation to God is pointed by the echoing of the reply,
‘‘Here am I,’’ twice, in verses 1, 11, when God called him; and Abra-
ham uses the same words in verse 7, in his reply to Isaac’s calling him
‘‘Father.’’ In ring composition repetitions are markers of structure.
These repeated answers have made a parallelism between two sons:
Abraham becomes beloved son to God, and Isaac is Abraham’s be-
loved son. The repeated double emphasis on paternal affection, di-
vine and human, tells the reader to anticipate the happy outcome. It
draws a clear correspondence between God as father to Abraham and
Abraham as father to Isaac.
Remember also Abraham’s reasons for trust. Isaac was the miracle
child that his mother (too old to conceive and bear) was promised a
year before his birth: ‘‘Sarah, your wife, shall bear you a son, and you
shall call his name Isaac. I will establish my covenant with him as an
everlasting covenant for his descendants after him’’ (Genesis 17.19).
The covenant is the background of the story and ratifies the interpre-
tation that fits with the established reputations of God and Abraham.
This panel form is one of the usual ways of presenting a chiastic
structure. Chapter 11 is the center (ignore the asymmetry of bulk).
This is the point at which the recital starts its return to the beginning.
I personally do not like this presentation, partly because it is difficult
to see the match between items, but mainly because the design ob-
scures the link between beginning and ending by placing them visu-
ally as far apart as possible. I prefer to display it as a ring, so that the
ending visibly joins up with the beginning, and the matching of corre-
sponding stanzas on each side is easy to see. This would be a ring with
twelve points of reference (Fig. 5).
To present this short version of the story I have picked out obvi-
ous verbal concordances and overlooked some that did not seem to
mean anything in this pattern, such as Abraham twice raising his eyes.
22
MODES AND GENRES
v.1. God called

‘‘Abraham!’’ ‘‘Here I am’’
v. 2. ‘‘Take your son, your v. 15–18 ‘‘You have not

only son’’ withheld your son,
your only son.’’
v. 3–6. Going to the place. v. 14. Called the name of the

Abraham saw the place place, ‘‘God will provide’’
afar off
v. 7–8. So they went on together. v. 13. Abraham saw the ram,

‘‘Father,’’ ‘‘Here I am, my son,’’ took it. Offered it as a burnt
offering
‘‘Where is the lamb for the burnt
offering?’’
‘‘God will see to it, my son.’’

So they went on together.
Fig. 5. The Akedah in a ring, showing the pattern of key words.
Clearly this version can be improved and elaborated. It is here to

introduce the point about a ring within the ring, verses 7 and 8, which
are parallel to verse 13.
Reading the binding of Isaac (Genesis 22.1–18) as a ring, we need
to notice the opening phrase and its repetitions. The story opens
abruptly with God calling from Heaven, ‘‘Abraham!’’ Then, at the
most dramatic moment in verse 11, he repeats it: ‘‘Abraham!’’ The
call is answered each time by ‘‘Here am I!’’ at verse 1 and verse 11, in
the same words, thus making the start correspond to the midpoint.
The repetition splits the ring down the middle, as in the ring diagram
in Figure 5. The split into two halves is the first big clue to reading it
as a ring.
At verses 15–18 we know we have come to the end because the
heavenly call ‘‘Abraham!’’ is repeated, and also because ‘‘You have not
23
MODES AND GENRES
withheld your only son’’ concords with verse 2. The ending has met
the beginning. In between the three heavenly interventions, which
open and conclude the whole and mark the middle, Abraham is pre-
paring to obey the command. In all respects it is a standard ring form,
where the ring is divided vertically by the two ‘‘Here am I’’ phrases.
There is a clear mid-turn. No further elaboration connects the mid-
dle with the beginning and end; they are connected directly by the
spoken words. This is so short a story, and it is so compact, no more
needs to be said. In the big rings the mid-turn is often elaborated at
length and forms another a ring in itself.
This story is one of the literary examples chosen by Erich Auerbach
to illustrate the power and authority of biblical narrative, in spite of its
‘‘only rudimentary syntax.’’∞≥ He generally compares the Bible nega-
tively with Homer’s Odyssey, replete with syntax, rich in psychological
information and sense of place and time. However, we note that both
are ring structures, though he is comparing short with long rings. We
can expect that for both long and short forms the ring itself provides
syntax for the composition. It explains the power and authority of
the biblical narration, which Auerbach extols. It also controls the
interpretation.
We said that the story is split down the middle by the repeated call
‘‘Abraham!’’ and the answer, ‘‘Here I am!’’ (in verses 1 and 11). There
is a first section in which the constituent verses 2 through 10 are
paired with their corresponding verses in the second section, 12 to 18,
on the other side of the mid-turn. Notice how the author has placed a
little ring within the bigger ring. The words ‘‘So they went on, both
of them together’’ are repeated to make an envelope, or inclusio:
A So they went on both of them together. [v. 6]

B Isaac: ‘‘My Father,’’
C Abraham: ‘‘Here am I, my son.’’ [v. 7]
B% ‘‘Behold, the fire and the wood. Where is the lamb for a burnt
offering?’’
C% ‘‘God will see to the lamb for a burnt offering, my son.’’ [v. 8]
A% So they went on, both of them together.
24
MODES AND GENRES
The four lines inside the bracketing words of A and A% are the dia-
logue between father and son, two sentences for each speaker.∞∂ At the
C and C% lines Abraham repeats the words ‘‘my son’’ and affirms his
confidence in God. Father and son walking along, ‘‘both of them
together,’’ conveys intimacy and affectionate trust but also much
more. By repeating the words ‘‘Here am I’’ but adding ‘‘my son,’’ the
loving relation between Abraham and Isaac makes a parallel with the
relation of Abraham and God: it indicates that Abraham is like a
beloved son to God. This spreads the trusting intimacy of the re-
peated phrase ‘‘They went on, both of them together.’’ The affection-
ate wording of the internal ring sets the mood of the piece. I maintain
that it tells us that Abraham knows what it is all about and that he has
nothing to fear for his son.
Isaac knows he is unique and beloved, and he almost certainly has
known from infancy that God’s plan for a covenanted people depends
entirely on his own survival. The double ring tells us eventually that
there is no reason to suppose they are speaking to each other in
anguish. It is like a family performance. Father and son support each
other in demonstrating their perfect confidence in God’s love and
goodness. Abraham knows full well that God has set all his plans for a
people to worship him on this one boy’s progeny. Isaac is destined to
grow up and found a lineage.
The clinching point is that the positioning of the internal ring is a
pointer: it is placed exactly opposite the position of the ram caught by
his horns (see Fig. 4). Outside of the literary frame Isaac’s question
‘‘Where is the lamb?’’ might seem to be devastating. But when Abra-
ham, unperturbed, gives his cool answer, ‘‘God will provide it,’’ and
the parallel points across the diagram to Abraham’s finding the ram
trapped in the thicket, his answer is validated. As the bard sings verses
13 and 14 the listeners will catch the words ‘‘for a burnt offering’’
repeated, and will thrill with recognition.
The double-ring form has set a mood that is not anguished or
suspenseful; Abraham is not in agony, God is not unkind, the child is
not afraid (Fig. 6). God is content that Abraham has demonstrated
25
MODES AND GENRES

Fig. 6. Father and son confident together. Drawn by Pat Novy.
that he knew what would happen; he trusted in God’s keeping his

word, and he was right to do so, and so was Sarah.
There are also cross-references and smaller chiastic twists inside
the big ring. For example, verses 11 through 16 start and end with
God’s call, so enclosing the whole of the second section as a second
ring, or a subset of the first. Verses 7 and 13, and 4 and 13 are also
elegantly cross-linked, creating a tightly integrated whole. I hope that
these short examples show how richly worked the small rings can be.
Placing them here will help me in the next chapter to describe how a
ring composition, large or small, is constructed.
26
MODES AND GENRES
After Otterlo and Vladimir Propp, no one should be surprised to

find sophistication and subtlety in early literature, as well as obser-
vance of elaborate rhetorical conventions.∞∑ The strongly knit com-
position explains why the ring does not easily yield its meaning at the
first reading and perhaps explains why it has fallen out of use. (More
of this in Chapter 10.)
A literary form is a resource; it can be used to declare whatever the
writers want to publish abroad. It comes into fashion and goes out.
The society itself, not the instruments of expression it has produced,
puts constraints on the thought of its members, if only by the sheer
need to live together in peace. The ring is obviously a demanding
construction. The initial question for this study is why people whose
life is relatively simple would want to write in such a complex way.
Part of the answer is that however simple the form of community, the
members normally need to distinguish a high style from the low-style
speech that serves for work or domestic life. Ring composition is used
for ceremonial speeches, victory odes, funeral orations, and joyful
celebrations. It is also the common form for solemnly reciting myths
of origin, or for entertainment by wandering bards. It is very old, and
it is extant still in different parts of the world.
Complex conventions of writing or speaking fulfill a role in validat-
ing a message. Religious doctrine tends always to be under challenge.
Skeptics or rivals will be alert to contest it. No wonder that religious
themes should inspire very elegant writing and, in their high periods,
unsurpassed literary and artistic technique. A well-crafted composi-
tion is its own authentication. The elaboration is not just for fun; it is
the way to say that something is important, something serious needs
to be said, there is a message that must be heard. It must be unam-
biguous, so it is surrounded on all sides with compatible materials and
packed deep into a big container.
We need to recall how any community may be aware of its own
fragility. So much work has to go into its maintenance, and dissen-
sions arise so easily. Angry departures threaten to split the group, and
quiet defections happen all the time. Even at the basic level of keeping
up its numbers, a community is vulnerable to strife. Disagreement
27
MODES AND GENRES
about facts tears it apart. But facts by themselves will not force a
decisive verdict. Interpretation is open in principle, but the members
who want the community to stay together, and at peace, will try to
achieve agreement.
Those ancient peoples whose writings puzzle us have struggled
hard against intellectual confusion by establishing their own founda-
tions of knowledge. Strong taboos helped to maintain their cosmol-
ogy. The mere fact that their civilization has persisted shows that they
have held disruptive intellectual challenge in check. Blank areas of
knowledge protected by rules of silence uphold the coherence of their
intercourse. To preserve their old certainties they would have had to
control the young and keep foreigners out. The better they are insu-
lated from the outside, the easier it is for an isolated community to
hold to their common knowledge of the world’s regularities. Spared
challenge from beyond their walls, they can lay their cosmologies on
logical foundations and entrench them by regular practice. Their
taboos and purity rules would hold the foundations steady.
But these solutions to social problems imply that a stranger’s claims
to hospitality are inevitably subject to scrutiny. He needs to give
evidence of his claimed identity. To do so, he may have to recite a long
kinship genealogy, in which the least mistake will damage his claims.
If he is to pass as a sage, he may be required to name the sages under
whom he has studied. Display of literary skills is an additional kind of
validation.∞∏ For the traveler needing friends abroad, a show of liter-
ary authority may be as good as a letter of introduction. I suggest that
this may be one context for explaining why ring composition would
be highly elaborate and difficult to master.
Competence in literary skills would be one of the proofs of worth.
Simply by virtue of its symmetry and intricate completeness, the ring
form conveys authority and prestige. This would be one of the advan-
tages of a dominant literary form in societies with simple technology
and weak coordination. Durkheim’s argument was that these are pre-
cisely the kinds of communities that need to have shared meanings to
give a basis to their otherwise fragile solidarity.∞π Such communities
absolutely need to settle on some agreed meanings.
28
MODES AND GENRES
Competition for a following is one of the ways this is done. A large

following gives prestige and authority to the winners.∞∫ The very
complexity of the conventions of ring composition suggests com-
petition among the poets, teachers, and songwriters. Competition
for prestige is one of the drives that hold society together. For an
example of competition in poetic competence, consider the Somali,
people of a warrior nation who compose and passionately love poetry.
They greatly admire an accomplished poet; they deride an unskilled
one. Their judgments are based on elaborate poetic forms deploying
formidably complex rules of alliteration. ‘‘In every hemistich of a
poem at least one word has to begin with a chosen consonant or with
a vowel . . . only identical initial consonants are regarded as allit-
erative . . . with one another and no substitution by similar words is
admissible. All initial vowels count as alliterative with each other, and
again, this principle is most strictly observed. The same alliteration is
maintained throughout the poem . . . if the alliterative sound of the
poem is the consonant g, in every hemistich there is one word begin-
ning with g.’’∞Ω
A man aspiring to leadership among the Somali can win support
only if he can demonstrate his mastery of these fiendishly difficult lit-
erary exercises. This is a nation where politicians must compete as
poets. By their poetry they earn esteem and attract followers. Their
political hustings are great poetry sessions where poets recite to
highly critical audiences.
Language is not self-authenticating. As we listen we consider
whether to believe the words we hear, whether we have understood
them right, whether to act on them. What is true is not always said
convincingly. In the summer of 2003, as I was writing this, the Brit-
ish government was acutely embarrassed by the charge of having
spoken deceiving words about the need to go to war. Scrutinizing the
words does not help; they can bear two meanings: as a warning of
danger, desperately urgent, or as merely cautionary. Written words
are not supported by body language, voice, and physical context.
They lack that foundation. This is where poetry can help: a show of
literary authority may be as good as a show of independence or valor.
29
MODES AND GENRES
This would be one of the advantages of a high-style literary form in

societies with simple technology and weak coordination. All commu-
nities need to have shared meanings. A set of literary conventions
(the more complex the better) is necessary for the itinerant teacher or
holy man.
30
three
HOW TO CONSTRUCT AND RECOGNIZE A RING
If ring composition is really rooted in our universal mental heri-

tage, why do we have to have all this explained? Why do we ourselves
not compose ring structures all the time? and everywhere? Parallel-
ism has had a universal distribution over the globe, so why does
it feebly give way to other compositional forms? Why do the old-
fashioned rings no longer make sense? More directly to my theme,
why do ring compositions get so badly treated by Western scholars?
And going back a step further, why are ring compositions so difficult
for us to read?
Moving now to show how a ring is composed will help to lay the
basis for answering this question. I have put off this chapter too long.
We will now consider some of the technical problems in composing
in rings. When the sequence of a composition seems to be jumbled,
the question is, ‘‘What is this piece doing here?’’ One satisfying an-
swer to this is: ‘‘It is here to complete the pattern of the book.’’ Or,
‘‘Its position just here cues the reader to see a correspondence across
the book.’’ Or else: ‘‘The positioning of this piece enriches the mean-
ing by pointing to analogies.’’
It is time to be more precise about what makes a text into a ring.
Definition is frustrated by the great variety of ring structures; some
are quite loose and free, some are very strict. We know already that a
ring composition is known by the ending coming back to match the
beginning. From this feature the name ‘‘ring’’ derives, but there are
other features of ring composition that follow from it.
Chiasmus is a structuring device that inverts the ordering of words.
A major ring is a triumph of chiastic ordering. ‘‘The phenomenon of
inverted word orders,’’ as Welch and McKinlay’s impressive Chiasmus
Bibliography demonstrates, is one of the prime indicators of the ring.∞
The other prime test of a well-turned ring is the loading of meaning
on the center and the connections made between the center and the
31
HOW TO CONSTRUCT A RING
beginning; in other words, the center of a polished ring integrates the

whole. Welch’s concluding chapter lists fifteen criteria that can be
used to measure the strength or weakness of the chiastic structure in a
given text. The most interesting of these (and the most debatable) is
his effort to distinguish a chiastic ordering that just happened, with-
out any authorial intent, from one that has been deliberately planned.
He warns against the ‘‘intentional fallacy,’’ the idea that any discern-
ible pattern in the text must have been devised intentionally.≤
To this I demur, finding it hard to believe that a large poem could
be chiastically ordered without anyone having knowingly created the
structure. To my mind it is deliberate, which invites us to ask, what
this form of literature is meant to achieve when it is used for long
literary exercises? Are the restrictions onerous? Why are they ac-
cepted and the rules obeyed?
I can think of very short chiastic forms arising spontaneously in
compositions or in snatches of conversation. In my childhood we
knew a countryman who used to speak chiastically. Waiting for him to
work through an idea was slow. In the following speech the first and
last lines correspond; the order of words is inverted. Like a true ring
the message is cupped in the center, C, and framed by parallels, AA%,
and BB%, and the first and last lines are virtually the same.
A These young plants don’t want too much water;

BDon’t water them every day,
C Water them every other day. [Here is the kernel of the
message.]
B% If you water them on Monday, do nowt on Tuesday, water
them on Wednesday.
A% Too much water isn’t good for these young plants.
This was quite spontaneous. Welch mentions the story of a queen

who answered twelve questions, starting with the twelfth and working
her way back to the first. It seems a natural way to remember a list.
Even longer chiastic passages are probably not too difficult to pro-
duce off the cuff. They are found in the Bible, plentifully scattered
through most of the books of the Pentateuch and histories and also
32
through the New Testament. Roland Meynet usefully confirms that

this style makes reading of the Bible difficult for the modern reader.≥
They have been much discussed, following in the path of Lund’s
classic Chiasmus in the New Testament, which gives a survey of earlier
work.∂ But most of these examples are content to pick out little paral-
lelisms and forbear to test the structure of a whole book.
Attempts to identify the macro-compositional form that controls
a long poem or a whole book by giving it an all-embracing frame
are fewer. It is a massive undertaking in which Cassuto led the way
with looking for ring forms in the beginning of Genesis.∑ I note also
Jonathan Magonet’s analysis of the book of Jonah as a system of
parallelisms.∏ An outstanding exception is Gary Rendsburg in The
Redaction of Genesis. He reveals an all-inclusive chiastic macro-
structure that incorporates the whole book of Genesis. He displays
the ring structure while systematically checking on the observance of
the conventions for ring composition that I am about to list.π
This analysis has been frequently applied to the New Testament.
Some of the analyses are a bit thin and scrappy. There are outstanding
exceptions. One of the finest is John Bligh’s commentary on Paul’s
letter to the Galatians. Bligh is convinced that the whole composition
is arranged in chiastic form, and he demonstrates the case with metic-
ulous scholarship.∫
The author of a major ring needs to know how to expand the
structure of a few bare lines into the structure of a long poem or of a
book. Following Welch’s general invitation to go further with the
work he has begun, and following his example of a list of fifteen
criteria, I have found that the following seven indicators are enough
to show how ring composers identify and signal the literary kind they
are embarked upon. The minimum seven rules are just too many to be
of the spontaneous, unintended kind of inverted sequence.
To summarize, the first indication of a ring composition is that the
end corresponds to the beginning. The link between start and finish
signals completion. If there is no such tie-up, the composition is not
in a ring. The correspondences are indicated by key words, as we saw
in the story of Isaac. The key words or word clusters indicate thematic
33
parallels that appear in both items of a pair. Part of the strategy of

construction is to divide the whole piece into two parallel halves that
will be chiastically related. The pattern is ABC for the first sequence
and CBA for the second. Reversing the order is the technique for
bringing the ending back to the beginning. The second part can be
said to have turned around, or crossed over. The mid-turn is written
so that it makes correspondences with the prologue and with the
ending. When the whole poem or book is gathered together in the
middle, and referred to again at the end, the result is a well-integrated
composition.
The formal closure that makes it a ring would be an easy require-
ment, if that were all. The turning and closure by return to the start is
the commonest rule for almost any composition—a beginning, mid-
dle, and an end. But, as I have just said, closure is not the only, or even
the principal, condition for a ring. Essentially, ring composition is a
double sequence of analogies. First a sequence is laid down, then at a
certain point the sequence stops and the series turns around and a
new sequence works its way backward, step by step toward the begin-
ning. This puts each member of the new series parallel to its opposite
number in the first series, so the return journey reverses the order
of the outgoing journey. The longer ring forms tend to embellish
the mid-turn with an elaborate commentary. A well-marked turning
point is a sign of a well-designed ring composition (Fig. 7). Some-
times it takes the form of a minor ring. Sometimes it is so long as to
mislead the reader about its place in a larger structure.
It is not difficult to recognize a mid-turn if you are alert to the func-
tions it performs for the piece as a whole. As in any well-developed
literary form, the conventions answer to specific technical problems.
Other bards or rival writers will judge the work by its skill in resolving
these problems. They will expect symmetry and balance, and they will
judge how well the ending slots on to the start. To bring the pre-
ordained ending elegantly back to the beginning is not so easy as it
may sound. One of the special literary merits of a ring is to anticipate
its own form of closure from the beginning.
34

Fig. 7. The scroll. Drawn by Pat Novy.
I have abstracted the following seven rules or conventions from

long ring compositions. They are not rules in the sense of there being
something hard and fast about them. Breach carries no penalties, but
insofar as they are commonly observed they are like rules. They are
responses to the technical problems of coming back gracefully to the
start. Other technical problems arise out of the solution adopted for
circumventing the first.
35
1. Exposition or prologue: There is generally an introductory section

that states the theme and introduces the main characters. You can call
it a prologue. It sets the stage, sometimes the time and the place.
Usually its tone is bland and somewhat enigmatic. It tells of a di-
lemma that has to be faced, a command to be obeyed, or a doubt to be
allayed. Above all, it is laid out so as to anticipate the mid-turn and the
ending that will eventually respond to it.
2. Split into two halves: If the end is going to join the beginning the
composition will at some point need to make a turn toward the start.
The convention draws an imaginary line between the middle and the
beginning, which divides the work into two halves, the first, outgoing,
the second, returning. In a long text it is important to accentuate the
turn lest the hasty reader miss it, in which case the rest of the carefully
balanced correspondences will also be missed.
3. Parallel sections: After the mid-turn the next challenge for the
composer of a ring is to arrange the two sides in parallel. This is done
by making separate sections that are placed in parallel across the
central dividing line. Each section on one side has to be matched by
its corresponding pair on the other side. In practice the matching of
sections often contains surprises; items are put into concordance that
had not previously been seen to be similar. Parallelism gives the artist
opportunities of taking the text to deeper levels of analogy. When the
reader finds two pages set in parallel that seem quite disparate, the
challenge is to ask what they may have in common, not to surmise
that the editor got muddled.
4. Indicators to mark individual sections: Some method for marking
the consecutive units of structure is technically necessary. The pri-
mary problem is to make clear to the reader or listener where one
section stops and the next begins. Otherwise the pattern fades out.
There are various methods. Key words always carry a lot of the weight
of marking the sections. In a long composition the author will also
have resort to specific signals to indicate beginnings or endings of the
sections. Only when these have been found can the meanings that
have been packed together be sorted out. One method is to close off
each section by repeating a refrain, like the chorus line of a folk song.
36
Another method is to use alternation to mark the beginnings and

endings of sections. The book of Numbers solves the problem of
marking individual sections by using a very strong principle of alter-
nation. The whole book is organized as a system of strictly alternating
narratives and laws. It thus boldly solves the technical problem: the
general question of how to recognize where a section starts and where
it ends becomes the question of how to recognize a narrative, and
where it starts and stops, and how to recognize a legal section. In the
Iliad Homer organizes the Trojan War by alternating nights and days.
The strong patterning works to guide the interpretation.
5. Central loading: The turning point of the ring is equivalent to the
middle term, C, that is the middle term of a chiasmus, AB / C / BA.
Consequently, much of the rest of the structure depends on a well-
marked turning point that should be unmistakable. One clue that the
middle has been reached is that it uses some of the same key word
clusters that were found in the exposition. As the ending also accords
with the exposition, the mid-turn tends to be in concordance with
them both. Then the whole piece is densely interconnected.
6: Rings within rings: As Otterlo pointed out, the major ring may be
internally structured by little rings. Some rings emphasize the divi-
sion into two halves by making each half a ring (as demonstrated in
Genesis 22, the binding of Isaac). A large book often contains many
small rings. They may come from different sources, times, and au-
thors. One large ring can be composed entirely from minor rings
strung together in groups. This practice makes the ring form ideal for
incorporating old materials, as in the Bible. For examples of rings
inside a ring, refer to the stories of the shield of Achilles in the Iliad,
Balaam in the book of Numbers, and Isaac’s binding in Genesis.
7: Closure at two levels: By joining up with the beginning, the ending
unequivocally signals completion. It is recognizably a fulfillment of
the initial promise. Just arriving at the beginning by the process of
inverted ordering is not enough to produce a firm closure. The final
section signals its arrival at the end by using some conspicuous key
words from the exposition. Verbal repetitions indicate that the first
and the last section match in other ways. Most importantly, there also
37
has to be thematic correspondence: the original mission turns out to

have been successful, or it has failed; the setting forth is matched by
the journey ended; the command to go into battle at the start is
completed by news of the battle being finally won, or lost. The ex-
position will have been designed to correspond to the ending. When
it comes the reader can recognize it as the ending that was anticipated
in the exposition. In my opinion this sort of double literary closure is
not going to happen accidentally.
The seven conventions are drawn from the style of large ring form
prevalent in the literature of the Mediterranean eastern hinterland in
the eighth to the fourth centuries. These have been for me the most
accessible. But this is not the only form of ring; many different re-
gional styles have developed over time. For example, the pattern of
the Indian Sufi romance Madhumālatı̄ (by Manjhan, a.d. 1545) be-
longs to a different set of conventions (Fig. 8). Madhumālatı̄ is a
mystical romance built on four triangles. It is certainly a long and a
complex ring. It demonstrates how a system of parallels can hold the
composition together by adding great depth and range to the meaning
of the words as their mutual echoing draws distant contexts together.
To return to the seven conventions, when a piece conforms to them
it is easy to recognize the central place and the parallel arrangement
of the sections on either side. Then the interpretation is safe. It
should be difficult to misinterpret the fully integrated ring or to dis-
agree on what it is about; the parallels hold the meanings in place.
The method is good for laying emphasis. As a kind of syntax, the ring
form brings ambiguity under control and reduces confusion. And
knowledge of the construction usually changes prior interpretations.
However, famous epics or speeches in ring form seem to have had a
tendency to grow; minor rings are so easy to incorporate. The tale
within the tale can be elaborated so as to sharpen the point of the
whole composition, bringing an added elegance and enrichment. But
too many minor rings may distract the reader’s attention from the
main movement along the major (inclusive) ring. The clarity of what
Leonard Muellner calls the metonymic or the ‘‘syntagmatic axis’’ may
be clouded.Ω In the Homeric examples that he considers, the relation
38
Manohar and
Madhumalatï
eternally together
Leave-taking and King longing for
1/540
the second son
marriage 495 45
Full union of Madhumalatï’s
Manohar and divine beauty
Madhumalatï
90
450
Madhumalatï
longing for
135 The united
Manohar 405
couple
Madhumalatï
As a bird: 180 Manohar
Taracand 360 shipwrecked: Pema
225 Manohar’s story and

315 Madhumalatï longing for Madhumalatï
and Manohar
reunited 270
Battle with the demon
Fig. 8. The Four Triangles. From Aditya Behl and Simon Weightman,
trans., Madhumālatı̄: An Indian Sufi Romance
(Oxford: Oxford University Press, 2001). By permission of Oxford
University Press.
of one episode to the next is like that of one stanza to the rest in
nursery rhymes, the last episode incorporating the whole sequence, as
in ‘‘The House That Jack Built.’’ I am not sure that this is always the
case. But it is true that the outer shell of the major ring, the frame
made by the largest inclusio, has a strong incorporative effect.
By following the formal ring structure the narrator can take the
opportunity of a differentiated shaping of the theme. There is scope
for mood change. The mood may change in keeping with the twofold
structure of the ring. The first series may develop a problem, show a
tragedy, or present a puzzle, the central place may be the site of a
major crisis, then the second series may deepen the mood or lighten it
until the final denouement is reached.
39
Such a structure is very clear in the book of Numbers. In the story in

section V the people of Israel are in the desert, complaining about
God and Moses. They dare to compare their present hardships with
the good food they enjoyed in Egypt (11.1). God in anger answers
them with a great fire. Nonetheless, they complain a second time, ‘‘O
that we had meat to eat! We remember the fish we ate in Egypt for
nothing, the cucumbers, the melons, the leeks, the onions and the
garlic’’ (11.5), and also complain about the manna (11.6–9). Their
complaints made God angry again (11.10), but he brought them quails
(11.31–32), and then as they ate so lustfully he was angry and smote
them with plague (11.33). This all takes place in chapter 11 on the first
side of the mid-turn. On the other side of the mid-turn, in chapter
20.2–5, which is placed in parallel to 11, the people complain again,
this time for water. This time the mood is kinder; God is not angry; he
makes Moses strike water for them out of the rock (20.7–11).
When it is engaged in reading a structure of parallels the brain is
doing its everyday exercise. Can the same be said of the return of the
ending to the beginning? The words we use in everyday speech for
describing the process are suggestive. When the ring has gone from
the middle to the beginning it has achieved a formal completion. The
task of composing the ring has been finished. As I have said, it is
usually accompanied by semantic completion, such as tying up the
loose themes of a narrative. The ending section has to refer explicitly
to what was announced in the opening passages and to what has come
to pass in the course of the poem. It brings matters to a close by
distributing their just deserts to the characters or by reconciling them
by fulfilling curses or promises, making prophecies come true, point-
ing out that the original mission has been accomplished or harmony
restored. This is indirect support for Jakobson’s argument that paral-
lelism (and the chiastic structure) is hardwired in the brain.
Does the closing of the ring imply a closure of the narrative events?
If so, is the closure final? The series of rings may carry the sense
of being caught in successive rings without end. All the prospects may
be fixed at the end, as they were foretold at the beginning. In some
cultures it may be usual for the story to have shown transformations
40
along the way, so that after the end more transformations may be
expected.
For example, the prologue of the book of Numbers shows the
people of Israel preparing to set out for the Promised Land. As we will
see in the next chapters, it anticipates fighting and promises land at
the end. By the end of Numbers many subplots have opened, devel-
oped, and ended. At the very end, the people of Israel have reached
the Jordan and partitioned the land among themselves, tribe by tribe.
At this point Moses reports that God has said (three times): ‘‘each of
the tribes of the people of Israel shall cleave to its own inheritance’’
(36.7–10). This ending is an opening on the new life, which is to be
described in Joshua and Chronicles. Likewise, the ending of the Iliad
is conciliatory, as we shall see in due course.
The ring structure itself may suggest a cosmology of eternal return,
or it could suggest ending and renewal. We can but look to observe
whether the concluding mood is hopeful or grim. The point is that
the rhetorical form does not impose any particular mood for the
ending. The general impression is that the ring is a literary form that
is good for reflecting on, and for establishing a long view.
When we come to chapter 7 of the Iliad we must be struck by
how closely Homer’s poem conforms to these rules of composition.
This confirms that these rules are not a form of regulation that would
block the creative freedom of the performing poet. Gregory Nagy
insists on the element of spontaneity. He comments on bardic perfor-
mance, saying: ‘‘Song is inherently recurrent and recomposed, much
as every new spring is a joyous event of inherent recurrence and
recomposition.’’∞≠ There is no contradiction so long as we know that
the prescriptions of poetic form are a stimulus, not a constraint. As
Alastair Fowler says of the rules of genre, ‘‘The very mention of
prescriptive genres will have raised specters of inhibited creativity. . . .
Great writers have found a challenge in genre rules, while minor or
invertebrate talents have been positively supported by them, as by
armatures.’’∞∞
Douglas Hofstadter’s big book on problems of translation, Le Ton
Beau de Marot, offers a discussion of the relation of poetic form to
41
poetic metaphor.∞≤ He trounces the idea that the essence of a poem is

the sequence of metaphors; the form is essential. He lambasts Vladi-
mir Nabakov for trying to translate Pushkin without representing the
original structure of the poetry. However difficult it may be to capture
the Russian poetic form in English, Hofstadter insists that nothing
less will do justice to the task. And moreover, the constraints of rhyme
and scansion do not restrict but rather spur the creative impulse. He
quotes approvingly another translator of Pushkin, James Falen, on
the topic of formal structure: ‘‘a translator positions himself, in a
sense, on the work’s home ground and imposes upon himself a useful
discipline for the journey. Furthermore, he is thereby constrained, as
was the poet himself, to seek solutions without self-indulgence, to
find variety within oneness, and to earn a freedom within the bondage
of the form. The very rigidity of the stanzaic structure can bring at
times a fruitful tension to the words with which the form is made
manifest, and the economy of expression it enforces upon the transla-
tor will sometimes reward him with an unexpected gift.’’∞≥
Greatly to my purpose is Falen’s tiny poem that Hofstadter re-
ceived in a letter, called an ‘‘odelet’’ to constraints:
Every task involves constraint,
Solve the thing without complaint;
There are magic links and chains
Forged to loose our rigid brains.
Structures, strictures, though they bind,
Strangely liberate the mind.∞∂
We should add the sense of liberation, or of intellectual excitement

and play, to the list of good reasons that the antique poets had for
following the conventions of ring composition.
42
four
ALTERNATING BANDS
N U M B E R S
The book of Numbers has the reputation of a disorderly, unstruc-

tured book. If the reader thinks that all the books of the Penta-
teuch enjoy equal esteem, it will be a surprise to learn that the great
nineteenth-century commentator Julius Wellhausen regarded the
book of Numbers as a kind of attic used for storing biblical materials
that did not fit into the other books. It was a junk room for the rest of
the Pentateuch. As an anthropologist, my own reading is the exact
contrary. This book turns out to be another example of what Glenn
Most has described as the ‘‘Pindar problem’’ for Western Greek clas-
sicists: the misinterpretation of the text due to a misunderstanding of
its structure.∞ Numbers’ problem is the same: a poet highly esteemed
in his period is found to be quite impenetrable in modern times. I
maintain that Numbers is not a muddle; it is a highly structured ring
composition (Box 1). The book only seems disorderly because mod-
erns are not used to reading in rings.
According to the list of conventions given in the previous chapter, a
ring composition needs to have an exposition, a split into two halves, a
central place or mid-turn matched to the exposition, identifiable par-
allel series, and an ending. After the ending there may be added a
‘‘latch.’’ The piece may also contain smaller rings. Because it takes
work to make a ring we can assume that the author intends the lis-
tener or reader to appreciate the structure. To do that, we first need to
identify its building blocks. If the reader or audience is unable to
distinguish one section from the next the whole thing falls into a
muddle.
Numbers uses the principle of alternation very simply. A section on
laws is followed by a section of narratives, then laws, then narratives
all the way round. (My book In the Wilderness expounds the arrange-
ment of Numbers so I hope there is no need to go into detail here.)≤
43
Box 1. Numbers conforms to the seven conventions
of ring composition.
Convention 1. The Exposition (or Prologue)

Chapters 1 to 4 clearly qualify as an exposition. They lay
out the three main themes that will be developed through the
rest of the book and will turn up again at the end. The opening
theme is the order of the twelve tribes descended from the
twelve sons of Jacob. The Lord tells Moses to number the
tribes of Israel so that they can set out in marching formation
on the journey to the banks of the Jordan. As this book follows
on Exodus in the canonical order, it is obvious that they are
summoned to fight their way from Sinai to the Promised
Land, where they will find their inheritance (Exodus 3.8,17).
The exposition defines the status of the Levites. They will
have to be counted separately. They are relegated to the role
of temple servants. In this dignified but subservient destiny
God has ordained that they must always be under the com-
mand of the priests, Aaron and his sons (Numbers 3.9, 10).
The third and most important theme is the sanctity of the
tabernacle and the need to protect it from all profane con-
tact. The people are warned that anyone who encroaches on
the tabernacle will die. The narrative will stick faithfully to
the elaboration of these initial themes.
This exposition has a grim tone; danger is in the air, dan-
ger from God’s anger at encroachment on the tabernacle.
The Lord tells Moses not to let the Levite family of
Kohathites be destroyed (4.17), a point that will culminate in
high drama in the middle section when Korah, a member of
that tribe, is killed dramatically (but his brethren survive). I
will say no more about this point here because the match in
Numbers between exposition and mid-turn will be the main
theme of the next chapter.
Convention 2. Split into Two Halves
In Numbers the mid-turn is at chapters 16 and 17 (ring
section 7 in Fig. 9). After that point the text starts to work
back to the beginning. This is the device that enables the
sections of the first half to relate to their matching pair in the
second half.
Convention 3. Identifiable Sections

We have noted that in a short poem the attentive audience
can begin after the midpoint to hear repeated clusters of
words that had been heard in the first half. The repetitions
are clues to finding matched verses that indicate the ABCBA
form. Simple verbal clues by themselves, however, are not
reliable for finding the structure of a long book. Numbers
makes a special point of ending each section with strongly
chiastic verses.
Convention 4. Indicators
In addition to verbal clues a long book like Numbers also
needs to use other, more discriminating devices to make
units of structure abundantly clear. Numbers is built on the
principle of alternating sections; law and narrative alternate
throughout in clearly identifiable ways.
Convention 5. Central Place

Numbers makes a big climactic moment at its central
place, chapters 16 and 7. This is where the Levites stage their
revolt against Moses and Aaron, the joint authority under
which they have been ceremonially placed in the exposition.
This is where Moses’ authority is justified by God. When
Moses asks for a sign of divine support, the ringleaders are
swallowed up by an earthquake. And this is where the
authority of Aaron is asserted by the miracle of the flowering
rod. In the book of Numbers the rule of loading the meaning
onto the central place is fully observed. Here it gives a very
political meaning to the whole book that is not the usually
accepted one. There is understandable confusion insofar as
the central place does not refer to the ending, nor conversely
does the ending make reference to the Levites’ rebellion,
which fills the central place.
Convention 6. Rings within Rings

In the middle of the journey of the Israelites to the east,
the story is interrupted by the story of Balaam, the foreign
prophet through whose mouth God delivers prophecies of a
glorious destiny for Israel.
Convention 7. Closure, the Ending

In general terms Numbers has achieved full thematic clo-
sure. By the end, three leaders have died, the people of Israel
have followed the pillar of fire and cloud, they have fought
off the Canaanites and arrived at the banks of the Jordan, the
mission has been accomplished. Moses has fulfilled his life
task; his eyes have seen the Promised Land that they are
about to enter. At the very end of the book, in chapter 35, the
people of Israel are reckoning the borders and distributing
the land to eleven of the twelve tribes as if nothing had hap-
pened since the original numbering.
The Levites, who have figured so prominently in the ex-
position (as subordinates) and in the mid-turn (as rebels), are
not forgotten in the ending. In chapter 35 God tells Moses to
allot them cities and pastures, very necessary since they have
no claim to tribal territory. The conclusion thus makes the
necessary link with the exposition, but it omits the mid-turn.
I discuss this further in Chapter 5.
ALTERNATING BANDS
Though there are thirteen sections in all, at this stage I need to

present only the first twelve. The last one is the latch, which ought to
clip the two sequences firmly together.
At an early stage of my research on Numbers, after I had identified
the alternating sections I did not know how they should be arranged
in relation to each other. I wondered what the number 12 represented
in the editor’s thought. Perhaps it referred to the lunar calendar, a
twelve-month year, with an optional intercalary month when the days
needed to be realigned. The latter, I thought, might be represented
by the latch of the book. I looked at zodiacal circles, but I saw no light.
Then I received a letter from David Meijers, who explained that the
months of the year were associated with twelve principles of the uni-
verse, the sephirot, arranged in a circle, with two new years facing each
other, one in the first month and the other in the seventh.≥ As soon as
I organized the distinctive sections of Numbers in this fashion the
pattern was justified by the concordances that shone out of the text.
Looking at the diagram in Figure 9 and reading the sections con-
secutively down the columns, we notice that the mid-turn is at 7, an
odd number. This means that the sequence in the second column
starts with an even number, 8, which pairs with 6 on the other side.
Each of the even numbers pairs with an even number, and likewise
each odd one with an odd. I take that as a signal to read them synop-
tically. When we find that the even-numbered sections report the
laws given by God to Moses, and that the odd numbers are all narra-
tives about the desert journey, the cue to read them across is un-
mistakable. They are a set of five parallel pairs, and we find that they
are parallel in more senses than the simple pairing of odd and even
numbering. Numbers has packed the themes of the book into five
parallel rungs (Fig. 9).
Any serious reader will want to know the evidence for my claim
that the sections have been laid down in parallel. Not only I myself
but also David Goodman have to answer how we came to these firm
conclusions. He arrived at an early point to help me with the Hebrew
text on precisely this question. Table 2, showing numbered sections
arranged in their parallels, provides the main part of the evidence.
47
ALTERNATING BANDS
G 1. story
law law
3. story 11. story
9 law law 7
5. story 9. story
law law
7. story G
Fig. 9. Stories and laws alternating through the ring.
David Goodman found that the Hebrew text marks the separate
sections in the book with syntactic, verbal, and structural signals.∂
The narrative sections have a distinctive beginning. Each opens with
a time, a place, and a person or a community involved in the story.
The ‘‘when, where, and who’’ rule is not hard and fast throughout.
Sections VII and XI start by saying ‘‘who’’ but not ‘‘where’’ or ‘‘when’’
the event took place; section IX says ‘‘who’’ and ‘‘where’’ but not
‘‘when.’’ Moreover, the time-space-community indicators are not ex-
clusively used for marking the shift of style; they sometimes appear in
the middle of certain main narrative sections, marking internal divi-
sions.∑ This makes it harder for the would-be ring master. But the
legal sections (the even numbers) provide backup as they have their
own distinctive beginning just where a narrative section ends. Box 2
lists and quotes the ‘‘when,’’ ‘‘where,’’ and ‘‘who’’ statements that
open the narrative sections.
The usual formula for beginning a legal section is: ‘‘The Lord said
to Moses, Command [or ‘Tell,’ or ‘Say to’] the people of Israel . . .’’
This is used in all the legal rungs beginning at chapters 5.1, 10.1, 15.1,
48
ALTERNATING BANDS
28.1, and 33.50. The only exception is chapter 18, which varies with
‘‘The Lord said to Aaron.’’ The variation may have good reason since
it follows the great vindication of Aaron’s authority in chapter 17.
The trick is not to look for matched themes until the formal pairing
has been found. The concordances arrived at by these means are
impressive. For example, there is a purely formal correspondence
between the first two matched sections, the command to expel lepers
in section II, the first case, and to expel the Canaanites in the pairing
section, XII. An example for this exercise is the relation between
section II (chapters 5 and 6) and section XII (chapters 33–35). Chap-
ter 5 gives laws about sins that break faith with the Lord. The sinner
must confess and make restitution. If there is no one to whom restitu-
tion can be made, he pays it to the priest (5.5–10). Then follow two
cases. The first is that of a woman accused of adultery: she may have
sinned, but there is no witness; she was not caught in the act, but her
husband suspects her. The jealous husband must bring his wife to the
priest with a small cereal offering. The priest recites a conditional
curse. She will go unharmed if she is innocent, but if she is guilty her
reproductive organs will be damaged. She must say ‘‘Amen’’ to the
curse, he writes it out, washes off his writing into a bowl of water and
makes her drink it. The curse, accepted and internalized by the ac-
cused, has placed the question of guilt in God’s hands. Now her
husband can watch her body—if it rots he will know his suspicions
were well founded. If she survives, she is proved innocent and can go
on with her life.
In the same section chapter 6 seems to make another complete
break with the context. It deals with the law for the Nazirite, a man or
woman who has taken a vow to be separate to the Lord. Nazirites
must not touch the juice of grape or take any strong drink, must never
shave the head, and must never go near a dead body. These are the
conditions of their special sacred status, which they will lose if they
break any of the laws. What is a Nazirite to do if someone suddenly
touches him as he falls down dead just beside him? The rest of the
chapter details the sacrifices he must perform to annul the defilement
that he has unwittingly incurred.
49
Table 2. Numbers’ pattern of parallels, sections II–XII.
Key words have been underlined.
Laws: Laws:
II: 5.2–3, Put lepers out of the XII: 35.34, Peroration, ‘‘for I the
camp, that they may not defile the Lord dwell in the midst of the
camp in the midst of which I dwell. people of Israel.’’ 34, Distribute
Rite for wife accused of adultery by the land by lot. 35, Forty-eight
her husband. cities for Levites, six cities of
6, Nazirite’s unintended corpse refuge; law of unintended
contact. manslaughter. 33.50, Drive out the
6.22, Peroration, Aaron’s blessing: Canaanites.
‘‘and so they will put my name on
the People of Israel and I will bless
them.’’
Narrative: Narrative:
III: 7, Offering of the princes of XI: 31–33, victory against
Israel to the Tabernacle, tribe by Midianites, purify warriors,
tribe. 8, Light lamps, consecrate captives, captured treasure, and
Levites. 9, Passover, the guiding animals. The Captains’ Offering of
cloud. Setting out on the journey gold to the Tabernacle. 32,
from Sinai. Reuben and Gad settle in Gilead.
33, Summary of journeys from
Egypt to the banks of Jordan.
Laws: Laws:
IV: 10.1–11, Sound trumpets for X: 28–30, Daily offerings,
alarm, for assemblies, war, beginnings of months, Sabbaths,
gladness, beginnings of months, appointed feasts (28): sound
and appointed Feasts. trumpets, 29.1.
Narrative: Narrative:
V: 10.11, In the desert. Setting out IX: 20–27, Arrived in Kadesh,
from Sinai, the order of the host. Miriam dies. People complain of
Moses invites Hobab to guide no water, God gives Moses water
them, 11. 11.1–3, People’s first out of the rock at Meribah. Edom
complaint punished by fire, second refuses to give Israel passage.
complaint, 11.4–13, wanting meat, Death of Aaron on Mount
11.31–2, God sends quails. 11.33, Hormah. 21, Israel fights and
the Lord smote people with destroys Canaanite King at
plague. Seventy elders filled with Hormah. 21, Israel’s victory over
God’s spirit, and prophesy. Joshua Amorites, Ammonites, and king of
worried. 12, Miriam and Aaron Bashan. 22–24, Balaam, the
challenge Moses’ authority. 13, foreign prophet filled with God’s
Moses sends twelve spies to see the spirit, prophesies in God’s words,
land, their evil report. 14, People blesses Israel. 25, In Shittim
complain to Moses and refuse to people committed harlotry with
go on, Joshua and Caleb loyal. the women of Moab and bowed
Moses confronts God. down to their gods. Balaam
People repent but try to go up by blamed. Zimri and Cozbi slain by
themselves. Amelekites drive them Phinehas. God’s covenant of
back to Hormah. everlasting priesthood for sons of
Phinehas. 26, New census in the
plains of Moab. Levites in separate
count, no inheritance.
Laws: Laws:
VI: 15, Priestly perquisites, law of VIII: 18–19, Instructions to priests
drink and cereal accompaniments and Levites; priestly perquisites
of animal offerings, portions from specified offerings; tithes for
reserved for the priest. One law for the Levites.
the congregation and for the Levites’ subordination reaffirmed.
stranger; sinning through 8–21, Heave offerings and things
ignorance shall be atoned, reserved from the fire shall be the
forgiven. Man breaking the priests to eat in the most holy
Sabbath stoned. Border of blue. place.
Box 2. Defining the narrative sections.
The opening words of the narrative sections give the

‘‘who,’’ ‘‘when,’’ and ‘‘where’’ orientations.
Sections I and VII make a vertical split down the middle of
Fig. 9.
Section I, ch. 1.1–2: ‘‘The Lord spoke to Moses in the Wil-
derness of Sinai, in the tent of meeting, on the first day of the
second month, in the second year after they had come out of
the land of Egypt, saying: Take a census of all the congrega-
tion of the people of Israel . . .’’
Section VII, ch. 16.1–2: ‘‘Now Korah the son of Izhar, son
of Kohath, son of Levi, and Dathan and Abiram, the sons of
Eliah, On, the son of Peleth, sons of Reuben, took men and
they rose up before Moses.’’
Now observe the ‘‘when,’’ ‘‘where,’’ and ‘‘who’’ statements
that start sections III and XI, V, and IX, composing the two
story rungs that cut across the middle of the diagram:
First Story Rung
Section III, ch. 7.1–2: ‘‘On the day when Moses had fin-
ished setting up the tabernacle and had anointed and conse-
crated the altar with all its utensils, the leaders of Israel, the
heads of their fathers’ houses, the leaders of the tribes, who
were over those who were numbered, offered and brought
their offerings before the Lord.’’
Section XI, ch. 31: ‘‘The Lord said to Moses, Avenge the
people of Israel on the Midianites . . .’’
ALTERNATING BANDS
Second Story Rung
Section V, ch. 10.11: ‘‘In the second year, in the second

month, on the twentieth day of the month, the cloud was
taken up from over the tabernacle of the testimony, and the
people of Israel set out by stages from the wilderness of
Sinai.’’
Section IX, ch. 20.1: ‘‘And the people of Israel, the whole
congregation, came into the wilderness of Zin in the first
month, and the people stayed in Kadesh.’’
The challenge for the modern reader is to discover what the two
cases have in common: the innocent woman who is suspected by her
jealous husband, and the man who has been defiled through no fault
of his own. They are each put through a ritual that will enable them to
go on with their lives. In her case there is no evidence of sin; in his
case there was no possibility of avoiding the accidental defilement.
The connection is not explained until we look across the diagram to
read the matching section. In section XII, in the last part of chapter
35, we are given the laws on breaking the law accidentally. This makes
the match for chapter 5, the woman wrongly suspected of adultery,
and for chapter 6, the Nazirite who was defiled by accident: they are
both matched by the long disquisition on the law of murder. In this
chapter the prime issue is to know whether it is a deliberate murder or
an unintended manslaughter. If the former is the case, the kin of the
victim will take blood vengeance, but if it is manslaughter the killer
may take sanctuary in one of the six cities of refuge. The three chap-
ters, 35, 5, and 6, share the theme of the ‘‘not guilty’’ plea. The
suspected wife who may be innocent, the Nazirite who was defiled
through no fault of his own, the manslayer who killed by accident—
the three cases are taken out of the normal ritual and legal process and
given protection. The interpretation that covers all these cases is that
53
ALTERNATING BANDS
if it were not for these provisions the law might have been the cause of
an injustice.
Section III tells of the setting out on the journey from Sinai; it is
paired with Section XI, which complements it by a summary of the
journey and a list of the stopping places.
The short sections IV and X both deal with the calendar and the
blowing of trumpets. ‘‘On the day of your gladness also, and at your ap-
pointed feasts, and at the beginnings of your months, you shall blow the
trumpets’’ (10.10). By itself the word trumpets might occur anywhere,
but in fact the two Hebrew words for trumpets occur only in this rung
of Numbers. The phrase ‘‘blow the trumpets’’ is more complex. When
we come to the matching section on the other side of the central
divide, there, in chapter 29 (see Fig. 9 for sections IV and X), after
detailing the ritual calendar of appointed feasts, the key words (‘‘blow
trumpets’’ and ‘‘appointed feasts’’) are given again: ‘‘On the first day of
the seventh month you shall have a holy convocation; you shall do no
laborious work. It is a day for you to blow the trumpets’’ (29.1); ‘‘These
you shall offer to the Lord at your appointed feasts’’ (29.39).
Normally, by themselves, the distribution of short phrases, be there
ever so many of them, is not conclusive evidence of a match between
the sections in which they occur. Some later editor might carelessly
stick in the key words at random. But the combined three key phrases
(‘‘appointed feasts,’’ ‘‘blow the trumpets,’’ and ‘‘in the beginnings of
your months,’’ 10.10 and 28.11) that are repeated in the matched
sections IV and X cannot be found anywhere else in the book.∏ They
are a safe cue to the parallel.
Sections V and IX both recount episodes on the desert journey. In
both the people complain to Moses about food and water. Both de-
scribe fighting with the Canaanites: in V the people of Israel are de-
feated at Hormah, in IX they inflict crushing defeats on the Canaanite
kings, including a victory at Hormah. Both describe the working of
God’s spirit, how it fills Moses, then the seventy elders, and then the
foreign seer, Balaam.
Like other ring compositions Numbers uses various markers and
other cues to identify the separate sections. They all point to the same
54
ALTERNATING BANDS
divisions of the text. Identifying the units of structure is too important

to be left to uncertain indicators. The same caution applies to key
themes. There is often some fairly clear thematic unity that holds the
two sides of a rung together; as in the case just discussed, trumpets
and appointed feasts relate to the ritual calendar where the paired
sections, IV and X, could be read consecutively without the slightest
jar. The same sort of smooth continuity marks the move across the
divide between narrative sections.
For example, it is easy to see that a section that has two events
featuring the people of Israel angering God by wishing they were
back in Egypt is to be paired with another section telling a similar
story. Sections V and IX are both narratives, and both describe the
complaints about food and water made by the people of Israel on the
march from Sinai to the Jordan. ‘‘We remember the fish we used to
eat in Egypt freely; the cucumbers, and the melons, and the leeks, and
the onions and the garlic’’ (11.5). The two sets of complaints do not
literally repeat. The first two complaints, in chapter 11, are about lack
of meat and the summer fruit and vegetables they used to have; the
complaint in chapter 20 lists the orchard fruits. ‘‘Why have you made
us come up out of Egypt, to bring us to this evil place? It is no place
for grain, or figs, or vines, or pomegranates; and there is no water to
drink’’ (20.5–6). The parallelism between the two episodes is clear,
even though they are well separated from each other by other events.
In the first case, God’s response is harsh; he is angry, he punishes them
with fire and plague, but he also sends them manna and quails to eat.
In the second case he is benevolent; Moses draws water for them out
of a rock. But they complained again, and he sent them fiery serpents
(21.6), followed by the remedy. The great importance of these epi-
sodes is that they are structural markers, which tells us to be sure the
mid-turn lies between them.
Sections VI and VIII deal with the offerings of cereal foods and
some meats that are made to accompany major animal sacrifices, and
they are allocated to the priests. It is not difficult to see these sections
as a matching pair. Between them lies the mid-turn (deferred until my
next chapter). It is common in ring compositions for the mid-turn to
55
ALTERNATING BANDS
be flanked by two sections that are nearly the same. The parallels
before and after the mid-turn form a triad that helps the reader to
recognize the significance of the piece in the middle.
Commentators who read the book of Numbers as a linear sequence
are understandably puzzled to find repetitions. Some suggest chari-
tably that the weary editor nodded. Others take repetition as an in-
stance of the lack of organization in the book as a whole. Sometimes
there is a justifiable misgiving about the semantic fit that the key
words and other signals indicate should be there. We expect matching
sections to be related by analogy, but the parallel is not always ob-
vious. The reader who is puzzled can take it as a challenge to reflect
further and to consider the seemingly obscure similarities the editors
had in mind when they strung what we first see as two apparently
dissimilar beads on the same rope.
We have observed that the opening lines provide an unreliable
basis for identifying a new section. ‘‘On the other hand,’’ Goodman
says, ‘‘the endings are carefully worked out.’’π The clinching evidence
of the patterning lies in the finely worked perorations of each section,
legal or narrative. ‘‘In sections I, IX, and XII, the final peroration is a
summation of the narrative passage immediately preceding it, and is
often in itself a repetition or rephrasing of a formulaic passage (e.g.,
divine command, Moses’ response, conclusion/confirmation of ac-
tion), so that in the context of the whole section, it stands out clearly
as a conclusion.’’∫ Particularly interesting is Goodman’s idea that
some of the sections have an overall chiastic structure that frames and
prepares their own conclusion. ‘‘Before reaching the final passage
(9.15–23), the reader has already received an implicit signal from the
structure that the Section is drawing to a close. Or else it is signaled
by means of an embedded message, such as a word or phrase repeated
three (or more) times.’’ He gives section V’s short concluding piece
(14.40–45) as an example:
And they rose early in the morning, and went up to the hill country,
saying ‘‘See, we are here, we will go up to the place which the Lord has
promised for we have sinned.’’
56
ALTERNATING BANDS
But Moses said, ‘‘Why now are you transgressing the command of
the Lord, for that will not succeed. Do not go up lest you be struck
down before your enemies, for the Lord will not be with you.’’ [Ob-
serve the repetitive play on ‘‘up’’ and ‘‘down.’’]
But they presumed to go up to the heights of the hill country, al-
though neither the ark of the covenant of the Lord, nor Moses, de-
parted out of the camp. Then the Amelekites and the Canaanites who
dwelt in the hill country came down and defeated them and pursued
them, even to Hormah.
Goodman describes this peroration as ‘‘characterized structurally by

four pairs of couplets, two of which are arranged chiastically and are
situated in between the other two to give the sense of the passage’s
lexical keys, ‘rising’ and ‘falling.’ ’’ The first half (14.40–42) contains
the root ’lh (‘‘to go up’’) three times and begins with way-yaškimū,
‘‘they loaded up.’’ In contrast the second half has the similar threefold
semantic associations from: ‘‘and you shall fall . . . and they came
down . . . and defeated them and pursued them.’’Ω
In general the narrative sections have long, repetitive perorations
and do not rely on the last one or two lines to mark the endings. Law
endings are marked by individual words or groups of words repeated
several times.∞≠ Like paragraphing or ‘‘chapterization,’’ the position-
ing of the twelve sections conforms to a clear organizing principle of
alternation between law and narrative. If there were no other mark-
ers, this principle makes it perfectly clear when one section has ended
and another started. Once the question of how to identify the sections
is settled, we can trace their arrangement in parallels across the board,
down to the mid-turn. There we can see the gathering together of
all the threads, and then up to the ending, a chiastic pattern writ
very large.
57
five
THE CENTRAL PLACE
N U M B E R S
We have mastered the formal structure of the book of Numbers

and its outside envelope. Now we want to know what is encased inside
it. Already we have come far enough to see that ring composition is a
rhetorical form specially suited for summation and reconciliation.
When every element in the composition interacts with all the others
and nothing is extraneous or unnecessary, everything contributes to
embellish the pattern. A well-made ring, like the story of the creation
of Adam in Genesis, or like the binding of Isaac, gives the snowflake
effect; like Francis Thompson’s ‘‘filigree petal’’ it compels admiration.
The reader’s expectation that it will be exhaustively coherent, how-
ever unrealistic, is a great motive for exploring the structure in detail.
So far, we have looked carefully at the structure of Numbers. Even
without examining the mid-turn it was easy to show the book’s metic-
ulous ring construction. It is like a demonstration model. When we
come to the mid-turn, the impression of a strongly dominant form is
further enhanced, except for one unexpected gap.
I start by saying once more that a ring composition condenses the
whole burden of its message into the mid-turn. What has been seen
through straight linear reading has to be read again with a fresh eye
for the message that is in the mid-turn. Numbers would be a superb
object lesson—until the very end. There we find that the editor has
not managed to link the central meaning systematically to the ending.
In a sense this makes it even more an object lesson, since it demon-
strates the problems that arise when the rules have not been followed.
If we ignore the literary conventions to which Numbers mostly
conforms we see only a mixed bundle of laws and episodes from
the heroic story of the people of Israel traveling from Egypt under
Moses’ direction. But if we read it synoptically, as it is constructed, if
we read it along the lines instead of down, we see at least two strong
58
THE CENTRAL PLACE
political protests. One is about the Levites, the other about the Jose-
phites. Thinking in analogies and writing in parallels do not bar polit-
ical criticism. A structure of matched analogies can make satire as
biting as you could want, dramatic and difficult to ignore. The brunt
of it is lodged in the mid-turn. A mere child, used to the conventions,
could not miss it.
An Israelite child listening to a recital of Numbers would be seen
paying close attention to the opening chapters (1–4). She would note
that this is a story section of the book because it starts with a place and
a date, and one action follows another in narrative time. Knowing
implicitly that stories unfold along an internal chronology, she would
expect them to be signaled by time and place, and would also know
that Divine Laws are eternal, for all times and places. So she would be
listening knowingly to the alternating bands of story and law. The
exposition will be her main clue to the later unraveling. Knowing that
the denouement will not start until the middle she would be waiting
eagerly for the mid-turn.
As the sage child harkens to the exposition (chapters 1–4), she
would memorize the names of the dramatis personae and note the four
social divisions revealed in the story of the first numbering, the tribes,
the princes of Israel, the congregation, and the Levites. At the end of
chapter 4 she would recognize that the introductory narrative section
has come to an end because chapter 5 starts with the law of leprosy
followed by another law about breaking faith with the Lord, and more
laws.
She may have spotted the ABA chiasmus of the first two chapters,
and she will notice that the Levites occupy the middle position, the
central place in this section:
A (1.1–46) The numbering of twelve tribes by the named heads of

houses, counting all who can go out to war (omitting the sons of
Levi).
B (1.47–54) The Levites are not to be numbered; they receive
a special commission to look after the tabernacle, the altar,
and all the furnishings, and to pack and transport the whole on
59
THE CENTRAL PLACE
the journeys to come. They will camp around the tabernacle.

Unlicensed approach will arouse God’s wrath, and the intruder
will die.
A& (2.1–34) This is a return to the twelve tribes, commanding their
positions in the camp; on the east, Judah with Issachar and Zebu-
lun; on the south, Reuben, Simeon, and Gad; on the west, the
sons of Joseph, Ephraim, and Manasseh, and Benjamin; on the
north, Dan, Asher, and Napthali. The Levites are safe (or under
control) in the middle of the camp.
Well used to geometric thinking, when the child gets to hearing the
mid-turn of the whole book she will not be surprised to find that the
Levites are there, in the middle of the story and in the midst of the
strife. From hearing the first two chapters she will have concluded
correctly that this is a book about the Levites. Why should they not
be allowed to bear arms? Perhaps they cannot be trusted to be loyal.
Later events at the mid-turn will endorse this suspicion.
In chapter 3 God presents the Levites to Aaron as a gift. They are
his brethren; Aaron is a son of Levi too, but now they are all called to
do the service of the cult for Aaron and his sons. There can be no
doubt that for these other sons of Levi this is a downgrading. They
are to possess no territory, to bear no arms, and to provide physical
labor for the tabernacle, and the sons of Aaron will make the sacrifices
and act as their overseers. It is a humiliation for the Levites. This
having been made clear, the Levites are numbered, the three families
being headed by the three sons of Levi, Gershon the eldest, Kohath
the next, and Merari the youngest. The various services for the cult
assigned to each of the Levite families in chapter 4 sound like a lot of
hard physical labor. Make no mistake: it is menial work. They have
become janitors for the temple. Thus ends the first narrative section.
The young listener may doze through the laws in chapters 5 and 6
and wake up when chapter 7 indicates a return to the narrative mode
by saying: ‘‘On the day when Moses had finished . . .’’ In this chapter
the leaders or heads of families make formal offerings to the taber-
nacle. Chapter 8 prescribes what has to be done for inaugurating the
60
THE CENTRAL PLACE
Levites for their task and announces that the command has been
fulfilled. In chapter 9 the Lord gives the law for the Passover. Halfway
through chapter 9 (v. 15), the sequence reverts to the narrative mode;
the people set out under their respective military standards, following
their leaders. The discerning young reader, or listener, who suspects
that trouble is brewing for the Levites finds that there are several
troubles on the way, but the major revolt of the Levites is hidden in
silence until the turning point. It belongs in the central place.
In Numbers the central place starts at chapter 16 and includes 17
(section VII in Fig. 9 in Chapter 4). This is where the congregation
rebels against the authority of Moses and Aaron. All the disasters
foreseen in the exposition come to pass. Each of the three groups that
has been mentioned there now plays its fated role. So smooth is the
transition from the prologue to the mid-turn that the story could be
read straight on, without bothering with the intervening chapters.
In chapter 16 we hear at once about a rebellion against Moses and
Aaron. The ring leaders are Korah, the Levite of the Kohath family
(who in the exposition came in for special mention and warning, 3.27
and 4.1–17), and the sons of Reuben. They are supported by 250 of
the leaders of the congregation. The latter correspond to the cap-
tains, or princes, or heads of houses, who in the exposition did the
numbering at the command of the Lord. Now they publicly defy
Moses, accusing him of pride, self-serving, and failed leadership. It is
a takeover bid, a challenge to Moses’ authority, in clear defiance of
God’s command.
In riposte, Moses accuses the Levites. Recalling the exposition,
where they were separated from the rest of Israel and accorded privi-
leged service in the sanctuary, he asks: ‘‘Would you seek the priest-
hood also?’’ (ch. 16).
Korah assembled ‘‘all the congregation’’ against Moses and Aaron.
This brings all three groups of the dramatis personae of the exposition
into the action. The congregation of Israel figured at the beginning as
the subject of the counting, and in the count of their first born (3.40–
50). Now they are in revolt, led by the Levites and by 250 other
leaders. As events unroll, three punishments fall, one upon each of the
61
THE CENTRAL PLACE
rebel groups who had been warned in the first four chapters. Korah
himself is swallowed up in an earthquake with his men. Two hundred
of the rebel leaders (captains, heads of houses) die by fire. So the
Levites and the captains have had their due, but the congregation of
Israel is still not cowed. Next day they bitterly revile Moses and Aaron
for killing the people of the Lord (16.41). The punishment that falls
on them for this blasphemy is the plague; 14,700 died.
The wrath of the Lord has come out; Moses has been justified most
terribly, by plague, fire, and earthquake. But this is not the end. There
is another loose thread to be tied in. Aaron’s authority needs to be
publicly legitimized. Chapter 16 is only half of the midpoint of a ring
that started in chapters 1–4 with the counting of the different catego-
ries of people. In the other half chapter 17 goes back to the beginning,
with the Lord demonstrating his election of Aaron and his sons to the
priesthood.
The plague is still raging. As if intending Aaron to be justified
by a dramatic intervention, Moses sends him in to wave his censer
amongst the dying. He goes in, and the plague ends dramatically, but
the congregation persists in defying Moses. God proposes a trial in
which he will show who is in the right. All the twelve tribal leaders are
to give their ‘‘rod’’ (in function a minor ruler’s scepter) to Aaron to
keep overnight in the tabernacle. In the morning they are shocked to
find that the rod of Aaron has burst into blossom. With a wealth of
parallels echoing the beginning, we read how the divine demonstra-
tion of Aaron’s right to rule over them devastates the rebel Levites
and the others who have sided with them. It really terrifies them like
nothing before. They run to Moses and say: ‘‘Behold, we perish, we
are undone, we are all undone. Everyone who comes near to the
tabernacle of the Lord shall die!’’ (17.12).
This is a very curious wording for them to use, as there has so far
been no mention of the tabernacle in the course of the rebellion. I
suggest that the editor puts that phrase in their mouths so as to recall
the warnings against unsanctioned entry to the tabernacle in the ex-
position (chs. 3 and 4). Note there that God’s original warnings used
the words ‘‘any one who comes near’’ and ‘‘shall die’’ or ‘‘shall be put
62
THE CENTRAL PLACE
to death.’’ ‘‘You shall appoint Aaron and his sons, and if anyone else
comes near, he shall be put to death’’ (3.10). ‘‘And when Aaron and his
sons have finished covering the sanctuary and all the furnishing of the
sanctuary, the camp sets out, after that the sons of Kohath shall come
to carry these, but they must not touch the holy things, lest they die’’
(4.15). ‘‘Aaron and his sons shall appoint them each to his task, and to
his burden, but they shall not go in to look upon the holy things even
for a moment, lest they die’’ (4.17). (This was about the Kohathites, the
kinsmen of Korah, the future rebel leader.)
The richness and tightness of the weft is as impressive as in any ring
composition we are likely to read. The mid-turn of Numbers faith-
fully picks up the story of the exposition and responds to its predic-
tions. It also relates to the ending, where the Levites will be men-
tioned again. The mid-turn has not explained what it means for the
Levites not to be counted in the census with the people of Israel. This
is left to the immediately following chapter 18. The Lord said to
Aaron, ‘‘You shall have no inheritance in their land, neither shall you
have any portion among them. I am your portion and inheritance
among the people of Israel’’ (18.20).
When we get to the end of the book, after their rebellion, after the
rebels have paid the penalty for their crimes, the Lord does not forget
the responsibilities he has assumed on behalf of the Levites. He as-
signs to them generous tithes from the people, but they still need
somewhere to live. Fittingly, this will be provided in the ending, in
chapter 35.1–8. Partly because of its general convention that there
must be no loose threads, ring composition is good at grand custodial
gestures. Here, at the very end, the Levites who have caused so much
trouble are allotted forty-eight cities, of which six are to be cities of
refuge for unintended manslaughterers. Each city has pasture lands
for their flocks, so that they can live in the Promised Land, even
though they will have no tribal territory.
The Levites are one of the central themes. God has been good and
forgiving to them, but Numbers makes one message clear. No Le-
vite can aspire to the priesthood unless he is descended from Aaron.
Numbers has put a non-negotiable barrier between priesthood and
63
THE CENTRAL PLACE
the other sons of Levi. This is what the book declares, and in the fifth
century, when Numbers was getting its final revision, it was a hot
political issue. Because of the way it is elaborated through the book
and tied into the mid-turn we can assume that protecting the status of
the Aaronite priests is the principal objective of the editors of Num-
bers. This would have been for good political reasons.
When a region is shaken up by major crises, it often happens that a
renewal of spiritual energies is released. In the region of Israel the
eighth century was the scene of massive change. Particularly with the
rise of the Assyrian empire and the movement of Aramaic and As-
syrian culture into Canaan, the destruction of the northern kingdom
of Samaria and exile to Assyria, and Sennacherib’s war against Judea,
the upheavals were accompanied by the rise of new spiritual forces.
The Jerusalem temple and its cult became the object of attack by
reforming prophets. The priestly editors were inevitably aware of
the criticism directed at their established practice. They had another
cause to worry.
After the return to Jerusalem from exile, Ezra, the governor of the
province of Yehud, that is, Judah, was the Persian king’s representa-
tive. He had sometimes been criticized for being a puppet of Persia,
and for having it close to his heart to please the Persian authorities.
He believed it was better politically for Israel to be seen as a quite
distinctive culture in the region and would have dreaded the pos-
sibility of Judah being amalgamated with or absorbed into Ephraim as
one administrative unit. His claim to have restored to the people of
Judah the ‘‘Law of Moses’’ has been respected through the ages. Here
it is necessary to note that the laws he insisted on most strongly were
not recorded in Leviticus and Numbers by the priestly editors. For
one, he promoted the Levites above the priests; for another, he de-
manded that those who had married foreign wives should send them
away, and their children. He said that to marry foreign women is
against the Law of Moses. Such hostility to foreigners has no backing
from the priestly work. It is not the same law as the law the priests
were teaching.∞
This must have been critical for the loyalty of the priestly editors.
64
THE CENTRAL PLACE
They were open in their foreign relations. If they had been isola-
tionist they would never have been able to build up Jerusalem as
a rich international commercial center. They had for many genera-
tions been intermarrying with the other Levite families living abroad.
Once Ezra launched the attack on foreign wives the Jerusalem priests
would have become overnight a party in opposition. This would have
made them a political threat and explains why Ezra replaced the
priests with Levites for teaching the law. With this background we
can understand why Numbers downgraded the Levites to the level of
janitors and porters.
The priestly editors would have suffered insecurity in Ezra’s re-
gime. This accounts for two other political concerns that weighed
with them. One concerned the sons of Joseph, located in Ephraim,
now known as Samaria. In government circles they were regarded as
national enemies—fair enough, for there were wars between Samaria
and Judah. In spite of this, the priestly editors would have felt loyalty
toward the early Samarian priesthood, partly because for genera-
tions the two sets of priestly families had been intermarrying. They
worshipped the same God of Israel and learned the same Torah, but
over the intervening centuries Samaria had become a great, rich, and
powerful neighbor and, under Persian rule, a major political threat.
The other (related) cause that drew the editors’ sympathy was re-
sistance to the hegemony of Judah.
These political agenda help to explain an oddity about the book of
Numbers. The listing of the twelve tribes is repeated seven times.
The book starts with Moses being told to make a census, ‘‘Count the
children of Israel,’’ that is, ‘‘Count the descendants of Jacob.’’ Moses
counts twelve tribes (chs. 1–4). Although the sons of Levi have been
dropped, the number has been made up to twelve by splitting the
house of Joseph. By slipping in Joseph’s two sons as separate tribal
leaders Joseph now counts as two tribes. Ephraim and Manasseh
added to Benjamin bring the Josephite presence up to three out of
twelve tribes, quite a formidable constituency. The census is formally
taken again in chapter 26 when the new territories beyond the Jordan
are being allotted to each tribe, in accordance with God’s promise.
65
THE CENTRAL PLACE
Each time the count results in twelve tribes, including Benjamin and
Joseph’s two sons. On four further occasions the names of the twelve
tribes and their leaders are called out. One is the marching order, the
twelve tribes listed in chapter 2. Another is the presentation of gifts to
the tabernacle, which takes up the whole of chapter 7. The givers are
twelve princes, no more nor less. The fact that each prince on behalf
of his tribe gives exactly the same, down to the last spoon, suggests
that the count itself, the full count of twelve tribes, is the important
issue, not the inventories of identical items given twelve times. In
chapter 10 the marching order of the twelve tribes is announced. In
the listings of the twelve tribes I do not forget the presentation to
Aaron of twelve rods, one from each tribe, including Levi. This oc-
curs at the central place, chapter 17, the impressive climax that fo-
cuses on the major crisis in the narrative.
To keep listing these twelve names was a fairly clear reproach to
Judah, who claimed to represent ‘‘all Israel’’ to the Persian authori-
ties. Along with Judah, two of the Josephite tribes (Benjamin and
Samaria) were still on the political scene in the restoration period and
a source of diplomatic anxiety. (The other remaining descendants of
Jacob listed in Numbers had by this time faded out for all practical
purposes.) The choice of Joshua, a man of Manasseh, as Moses’ suc-
cessor (Numbers 27.18, 34.17) was also significant. To recall the
kinship of a hostile neighbor would not make the book of Numbers
popular reading in Jerusalem, the city of David. This combines to
suggest that a rebuke to Judah is one of the political messages of
the book. One can speculate that the priestly editors’ message was
discomfiting to Jerusalem readers who did not want to know about
the other sons of Jacob in what had been the breakaway Northern
Kingdom.
Ezra’s xenophobic foreign policy was bound to have brought him
into conflict with the priestly editors of the Pentateuch. They were on
the losing side in national politics. They disagreed with the book of
Judges on the hegemony of Judah among the tribes of Israel. The
book of Judges strongly supports the leadership of Judah: in its first
chapters it describes the conquest of Canaan and shows Judah vic-
66
THE CENTRAL PLACE
toriously leading the conquest, supported by Simeon, while the other

sons of Jacob lag behind in obeying the Lord’s command to clear out
the Canaanites from the land. Judges is particularly hard on the sons
of Joseph: it emphasizes that they failed to drive out the people of the
land ( Judges 1.22–26).≤ This is in contrast with Numbers. Whereas
Judges attributes the capture of Hebron and Hormah to Judah, Num-
bers attributes these victories to all the Israelites under the leadership
of Moses (Numbers 21.1–3). The editors of Numbers never defame
the Josephites. Their deviant political view would have put them in a
very difficult position.
They steadfastly composed their books according to their political
views, but their writing was of no avail. At the time of the redaction
the non-priestly Levites had evidently superseded the Aaronite lin-
eage of priests in their teaching functions. The Levites never ex-
ercised the priests’ sacrificial functions, but very soon we find that the
priestly role of the sons of Aaron has faded out of the history of Israel,
though they still occasionally played a diplomatic role.≥
In other words, the distinctive teachings of the priestly school were
overlooked, lost, wasted. The editors were scattered and some would
have gone into exile, voluntary or not. That is the context I suggest
for explaining the down-classing of the Levites at the beginning of
Numbers and their dire punishment for challenging Aaron’s author-
ity in the middle. In this respect, the final redaction of the book is an
anti-government protest made by the priests on their own behalf.∂
The background of submerged political conflict makes the endings of
the book of Numbers all the more interesting.
Numbers has two endings, quite distinct. One I call the ending, or
the first ending—that is, the group of concluding chapters 33.50 to
chapter 35. This is a proper ending; the great promises to Abraham
have been fulfilled and the land of Canaan is being distributed to the
tribes that were numbered at the start. On this topic the ending un-
mistakably matches the chapters of the census, 1–4 in the exposition.
The ending also has a curious point of parallel with chapters 5 and
6 in section I. This is again a very appropriate link for a ring ending to
make because it doubles the link with the start. Chapter 35.9–34 is a
67
THE CENTRAL PLACE
fine sermon on unintended sin. It justifies the creation of six cities of

refuge for the manslayer who did not intend to kill. I discussed this in
the previous chapter. We have seen that the section that is facing
chapter 1 has been well chosen to develop that very theme with two
more cases in which the law is stopped from doing harm to an in-
nocent person: the wife who is charged with committing adultery,
though there is no witness against her (5), and the Nazirite whose
body has been defiled through no fault of his (6), two analogies of
the involuntary manslayer who needs to be protected from the dead
man’s avengers. The cross-linkage made by the disquisition on inten-
tion may look somewhat contrived, but it is clear enough—all the
better for the argument that the editors were deliberately making a
ring structure (Fig. 10).
With such a regular ring, showing such conformity to the norms
for ring composition, we would expect the mid-turn to match the
exposition. Now is the time to notice that the mid-turn, though it
takes on the postures of a conclusion, makes no specific reference to
the text of the ending. Likewise, the ending makes no reference to the
mid-turn. The mid-turn, as we saw, tells of the revolt of the Levites
and their punishment. The ending, at section XII, does refer to the
Levites, but it is as if the dramatic events of the mid-turn had never
happened. This is like a hole in the weft. The ending does not forget
that the Levites are not to possess any tribal territory: this it has
learned from the exposition. Because they have no territory, forty-
eight cities are duly assigned to them, with pastures for their flocks. It
would be forcing the argument to take that to be a reference to the
mid-turn. The text says nothing to imply that the Lord’s gift of cities
to the Levites counts as a noble forgiveness for their revolt at the mid-
turn. This looks like a loose thread in this otherwise exemplary ring
composition. When we remember how punctiliously the editors of
Numbers have conformed to the overall pattern, and remembering
also the dangers of their political situation, we can suspect some gap
has been torn in the text.
We need this background when we come to the latch of Numbers.
Normally the latch makes an additional tie with the opening phrases
68
THE CENTRAL PLACE
EXPOSITION
Chs. 1–4 Census of arms-bearing men. Levites excluded
from priesthood, warned against encroaching on the
Tabernacle. Sons of Kohath in danger. Aaron to supervise
Levites. Key word clusters: ‘‘their fathers’ houses,’’ ‘‘lest
they die.’’
SECTION I ENDING Section XII

II: Ch. 5, key word cluster, ‘‘Put Ch. 33.50, key word cluster,
lepers out of the camp,’’ 5, Rite for a ‘‘Drive out the Canaanites.’’ 34,
suspected adulteress; 6, Nazirite’s distribute the land by lot; 35,
unintended corpse contact. cities for Levites; unintended
(Matching ending). manslaughter. (Matching
exposition and Section I).
MID-TURN:
Section VII, Chs. 16 and 17. Same actors as in exposition,
the Captains, the Levites, the congregation of Israel.
Revolt led by Korah, Levite of Kohath’s family; deaths of
the ring leaders, each group punished, by earthquake, fire,
and plague. Miracle of flowering rod, Aaron vindicated.
Key word cluster, ‘‘we die.’’ (Matching exposition only).
Fig. 10. Numbers’s center at the mid-turn.
and events. This time, the latch (chapter 36) does not make the con-
ventional references to the mid-turn. It repeats the case that was
newly introduced in chapter 27.1–11.
In the latch (chapter 36) Moses is consulted by the Josephites:
‘‘Then drew near the daughters of Zelophehad the son of Hepher,
son of Gilead, son of Machir, son of Manasseh, from the families of
Manasseh, the son of Joseph.’’ They ask whether the daughters of
Zelophehad, whose father died without male heirs, will lose their
lands if they marry outside their tribe (36.1–4). If there are no sons,
may the women be the heirs? Moses consulted the Lord, and received
69
THE CENTRAL PLACE
the following answer: ‘‘The tribe of the sons of Joseph is right. This is
what the Lord commands concerning the daughters of Zelophehad.
‘Let them marry whom they think best; only, they shall marry within
the family of the tribe of their father. The inheritance of the people of
Israel shall not be transferred from one tribe to another; for everyone
of the people of Israel shall cleave to the inheritance of his fathers.
And every daughter who possesses an inheritance in any tribe of the
people of Israel shall be wife to one of the family of the tribe of her
father, so that every one of the people of Israel may possess the inheri-
tance of his fathers. So no inheritance shall be transferred from one
tribe to another; for each of the tribes of the people of Israel shall
cleave to its own inheritance’ ’’ (36.5–9). In this there is no direct
connection with the exposition. Though the theme of tribal inheri-
tance is not raised at all in the first four chapters, the promised inheri-
tance for the people of Israel as a whole is a background assumption
from Genesis and Exodus, and the exclusion of the Levites from the
inheritance of land makes it explicit. The connection with the exposi-
tion is marked in this passage by making ‘‘father’s house’’ a key word
in the latch. The phrase ‘‘fathers’ houses’’ occurs three times in the
first verse, as if it had been stuck in there for a marker (36.1). The
phrase recalls chapter 1, where it occurs conspicuously seventeen
times (2, 4, 18, 20, 22 , 24, 26, 28, 30, 32, 34, 36, 38, 40, 42, 44, 45). In
the exposition the people were numbered by ‘‘their fathers’ houses,’’
and now at the end, their right to inherit land from ‘‘their fathers’’ is
vindicated by the Lord himself. The editor has added a clause in
which Moses guarantees to the Josephites their permanent status as
members of the tribes of Israel. He tacks it on the end of the book and
uses the key word cluster to connect it. To make a structural link only
by repeating the same word cluster and not to bother to find analo-
gous laws or events strikes me as rather clumsy scissors-and-pasting,
not Numbers’ best style.
The peculiarities of the Numbers latch supports Graeme Auld’s
reasons for considering the last part of Numbers to be a late addi-
tion. He sees Numbers as bridging two worlds, the earlier world of
Exodus and Leviticus, where the Levites barely figure at all, and the
70
THE CENTRAL PLACE
later world of Chronicles-Ezra-Nehemiah, where they figure promi-

nently. On stylistic grounds he shows that the ‘‘final quarter of Num-
bers depended on later elements from Joshua depicting tribal rela-
tionships, which in turn had drawn on materials in Chronicles.’’∑ This
is especially interesting because it predicates interference with the
original text on different technical grounds, and from different politi-
cal angles, from those I would have expected from government loyal-
ists in the Ezra period. Here is some late insertion that emphatically
defends the right of the Josephites to be included in the inheritance of
the children of Israel. Who would do that? I assume someone sympa-
thetic to the agenda of the priestly school.
Let us conclude with a return to the dismissive reception accorded
to the book of Numbers, the idea that it is confused and disorderly.
This seems to be a bias in the Christian readership. The early Jewish
scholars have no complaints on these grounds. They were familiar
with ring composition; writing in parallelisms was a strong Semitic
tradition. Nothing about this rhetorical style needed to be explained
to the later generations. Eventually the ring tradition lapsed, which
accounts for the baffled European scholars seeing only incoherence.
The reception accorded to it by rabbinical commentaries com-
bined detailed examination of the text with the pious veneration ac-
corded to the books of Moses. The interpretation they offered was
general, moralistic, uplifting. The book has, however, an anti-Judah
and pro-Joseph bias that would have been unacceptable to the leader-
ship at all times. I suggest that the book remained uninterpreted
because its political agenda was both impracticable and unwelcome.
71
six
MODERN, NOT-QUITE RINGS
We have now examined a particularly fine example of antique ring

composition. Numbers fits Roman Jakobson’s definition of parallel-
ism as ‘‘a system of steady correspondences.’’∞ With this example
in mind, we can better review his idea that a faculty for creating
or recognizing these correspondences lies inherent in the relation
among language, grammar, and the brain. Such correspondences
should enable us to answer our initial question: why is ring composi-
tion so widespread? If Jakobson is right, it is inevitably spread far
around the world. We should expect parallelism and rings to rise up in
any region, at any period.
Jakobson’s idea implies an aesthetic theory about the satisfaction
derived from the brain making images of itself at work and duplicat-
ing its own structure and activity. The brain works by building corre-
spondences and recognizing them. Jakobson was writing as a linguist
about this mental process of dividing and matching. His materials for
studying analogy were necessarily verbal; his examples came from
poetry, words, rhythms, and sounds. We can go on from the verbal
medium to consider creation of parallels between and across the
senses. Making divisions and seeing similarities, matching parts, like
to like—this is the essence of creativity. The internal pressure to
extend understanding by new analogies is evident all the time, for
example, when we create visual poetry, compose music for marching
or dancing, or identify similarities between bodily shape and land-
scape. We can hope that studying ring form will enable us to go
beyond words in asking more about ring composition. The next step
toward widening the question is to look for it in modern literature.
The seven rules for the construction of a ring that I presented in
Chapter 3 are not arbitrary limitations on what an author may do.
They are requirements for solving technical problems in turning the
72
M O D E R N , N O T- Q U I T E R I N G S
corner of a recital and returning to the start. The justification for the
rules is to preserve the ring shape. But why preserve it? What is the
attraction of returning to the start? This I shall have to consider at the
end of this book.
Meanwhile, it is useful to discuss what it means for a poem or
speech to be shaped in a ring. And this entails thinking about the
shapes of literature in general. Length, for example: when is the poet
satisfied that he has finished when he writes a short poem? How to
stop is one of any author’s (or painter’s, or preacher’s) problems. A
mundane and practical point of view would argue that the individual
artist is influenced by the viewers’ expectations. It sounds like a mar-
ket response, but it is more complex.
Henry James commented on the difficulty faced by the novelist
who first tries to write for the theater. The playwright’s task is largely
controlled by the theatergoers’ convenience and also that of the ac-
tors. The play must be confined in a time-space between the closing
of the offices in the evening and the last suburban trains at night.
Somehow the play must be shaped into chunks that do not put the
players into impossible positions (like being in two places at once) and
do give them time to change costumes. James intimated that the
external discipline imposed on the playwright by these factors was
salutary. In a similar fashion, the seven conventions for ring composi-
tion serve as an author’s kit for achieving balance and proportion. An
ending that has been prepared since the start, the right ending for a
poem of that sort, adds to the coherence of the whole.
No definition of the meaning of a ring shape will be true for all
civilizations.≤ There is no meaning for the idea of ring beyond that
assigned in a particular time and place. The ring may be a metaphysi-
cal model of world order—the Zodiac, for example. The book of
Numbers arranged in a twelve-part ring may stand for the revolving
seasons of the year. There is no saying whether a closed ring serves a
philosophy of closure and fixed endings, or whether the circle is seen
as one of a cyclic series that always returns to the same place. The
myth of eternal return can be taken to be comforting and stabilizing,
73
or it can be seen as a frustratingly sinister trap. Alternatively, it is

equally possible for every ending to be an opening on a new ring, a
philosophy of renewal and regeneration.≥
Jakobson’s thesis suggests a self-generating set of conventions that,
with local variations, the common structure of language causes to
emerge of its own accord. The other hypothesis is that the ring is a
peculiarly complex literary form that has been invented once and
found so satisfying that it spreads by diffusion to different regions.
Evidence does not support either thesis. The speculation about a self-
generating system gains some support from new literary construc-
tions that almost reach ring form without drawing on a prior tradi-
tion. My idea is simply that it starts with the wish to bring a text back
to its beginning before closing it down. In the effort to make a fitting
closure, elementary problems appear, to which there may only be a
fixed number of solutions. This would explain why literary construc-
tions of the same type keep reappearing all over the world.
One way to test and support the speculation would be to study
some of the almost-ring constructions in modern writing. Certain
genres lend themselves to it better than others. We could expect the
detective story to exemplify ring form because it starts with a who-
dunit question and must end by reverting to it with an answer. But to
discover a corpse at the beginning and refer to it again at the end is
not enough for the work to qualify as a ring. We need also to look out
for a structure that is held together by a strongly marked central
place, with an internal organization of parallel rungs, preferably alter-
nating in character, the two series organized inversely.
Detective fiction is generally governed by a principle that by the
time the conclusion is reached all the loose ends must be tidied up,
nothing left unexplained, everything accounted for. You would think
that this genre lends itself to ring composition because the require-
ment invites a step-by-step reconstruction of the crime at some point
after the middle. Following this rule Agatha Christie has composed
stories that almost conform to the ring form. But we have to be
careful. Even these do not quite work out as well-contrived rings.
One that superficially looks like a ring turns out, for example, to be
74
Table 3. Five Little Pigs as a parallelism.
I Carla’s enquiry, II Interviews III Five suspects’ IV Concluding

her projected with five suspects letters review and solution
marriage to John 1 1 1 John appears
2 2 2
3 3 3
4 4 4
5 5 5
only a set of parallels. Five Little Pigs comes in four sets of five chap-
ters (Table 3).∂ After the introduction the first five chapters introduce
the plot and its main characters. A young woman, Carla, whose dead
mother has been convicted sixteen years earlier of murdering her
husband, engages Poirot to find out who really did it. Unless her
mother’s name is cleared Carla feels she can never marry without
shame. The murder can only have been committed by one of five
persons who were present (and are still living). In each of the second
set of chapters Hercule Poirot interviews one of the suspects; he also
asks each to send him a written record of what they can remember.
Each of the third set of chapters consists of one of the letters written
by the suspects, presented in the same order, one chapter for each.
The last section also consists of five chapters in which Poirot un-
covers the real murderer through discrepancies in the written reports.
The whole book has a macro-parallel structure. Twenty chapters
have been meticulously matched in pairs, but there is no develop-
ment; they deal with the same event every single time. No antique
ring would be so monotonous. True that the ending joins up with the
beginning, but actually it never left it. There is no mid-turn; the
composition is not divided into two halves; everything is piled on to
the ending; the ring concept is not realized. The structure is a strong
device for controlling the plot, but it is too mechanical to count as one
of Christie’s best books, and it is not a ring.
What about another of her books, The ABC Murders?∑ The book
starts out promisingly. Poirot receives a challenging letter announcing
75
a series of murders that will defeat his ingenuity to solve. The letter is
signed ABC. Another letter tells him the first murder will happen in
Andover, then there is a murder there, and the ABC Railway guide is
found beside the corpse. He hastens there to check details. Another
letter arrives to tell him that the next murder will be in Bexhill,
and indeed there is one, and the ABC book is found beside the
corpse. The same happens again for the third murder at Churston.
The fourth occurs at Doncaster. In a lot of rushing to and fro Poirot is
accompanied by a group of supporters who are related to the victims.
The mystery is finally solved, but the systematic first half of the story
is followed erratically by reports from local people at the four dif-
ferent places and by diverse half-revealed thoughts in Poirot’s mind.
There is no systematic sequence of returns to the first series in in-
verted order, and no discernible mid-turn.
Julian Symons was generous with his endorsement on the cover: ‘‘A
masterpiece of carefully concealed artifice.’’ A really good ring com-
position hides the machinery of its construction. The structure in
this case does show rather obviously and it tends to overwhelm the
narrative.
In spite of the difficulties, detective fiction does provide interesting
comparison with ring composition on two scores. First, its structure is
very formal, like that of ring composition. Its rules can be extracted
from the stories and justified by the requirements of the genre. Sec-
ond, it has a good claim to being a genre, not just a method of con-
struction, not just a formal scaffolding for the plot. It was originally
based on ideals of deductive reasoning and scientific method. These
ideals engender rules or conventions that make them realizable.
Thomas Narcejac introduces us to the history of the genre by
going back to what the early practitioners said they were doing.∏
Austin Freeman, for instance, remarked that the writer of detective
fiction could not focus too clearly on personalities, lest the scientific
requirements be pushed into the background. The requirements of
the genre produce ‘‘a static person, with no future, no development’’;
a person’s role must be subject only to the development of the story
76
according to the pre-agreed structure. Freeman’s style is impersonal;

his aims are clarity and precision. ‘‘If a person could evolve, it would
be the end of the detective story.’’π
I once asked a successful English novelist, Margaret Drabble,
whether she had ever thought of writing detective stories. She an-
swered emphatically ‘‘No,’’ and, modestly, that she would never even
be capable of it because one would need to start at the end. While
writing her own novels she never knew what her ending was going to
be. The justice of her answer is confirmed by S. S. Van Dine’s twenty
rules for a detective story as given by Narcejac (Table 4). Van Dines
sees the reader and the writer pitted against each other in a gentle
contest, parallel to the competition between the detective and the
criminal in the story. Reading it is like a game.
The rules provide something like a golf handicap, an effort to make
a level playing field or, you could say, an attempt to protect the reader
from an unscrupulous opponent, the writer. They also protect the
game itself. When we reflect on them, we can recall authors who have
defied them, with deplorable results for the story. Unforgivable, out-
rageous, the author’s tricks are crimes committed against the naive
reader who believes that he has a fair chance. Like all cheating they
are self-defeating because they spoil the game. The rule for fair play
makes the game a very appealing analogy. It is difficult to apply such a
rule to ring composition, which in no way resembles a competitive
game, not even like drawing-room Scrabble. There is no occasion for
competition. All the restraints on free composition are devised to
resolve one problem: how to turn a ring. The decision to lead the text
back to the beginning and to do it elegantly, this is the source of the
conventions that have grown up around ring composition. For this
reason, when I scan modern works for ring form I should ensure that
the initial determination to return to the beginning is at work.
Before reading Van Dine’s twenty rules, I thought that my first list of
ten for the construction of a ring was too long. I cut down my list of in-
dispensable rules to seven. At that point the ring form still seemed to
be very restrictive. I thought it might be a problem to be addressed.
77
Table 4. Van Dine’s twenty rules.
1. Reader and the detective must have equal chances of resolving the problem.
2. The author has no right to use against the reader any tricks or strategies
other than those which the guilty party uses against the detective.
3. The true detective story must be free of any amorous intrigue. It must not
disturb the mechanism of the purely intellectual problem.
4. The guilty party must never turn out to be the detective himself or a
member of the police.
5. The guilty one may only be discovered by a series of deductions, never by
accident or spontaneous confession.
6. The detective must resolve the problem through the signs revealed in the
first chapter.
7. There is no such thing as a detective story without a corpse.
8. The detecting problem must be resolved by strictly realist means (no
clairvoyance).
9. There can only be one detective.
10. The guilty party can only be someone who plays a significant role in the
story.
11. The guilty party must never be one of the domestic staff, valet, cook, or
others. The criminal must be someone who is worth the trouble of
detection.
12. There can only be one guilty party.
13. Secret societies and mafias have no place in the detective story. It would
merge with the spy story or adventure story.
14. The crime must be rational.
15. The conclusion must be visible all the length of the story, if the reader
were only clever enough to see it. He ought to be able to guess the solution
before reaching the last chapter.
16. No long descriptive passages or subtle preoccupations with atmosphere,
they would burden the task of presenting the crime and finding the guilty
party, and slow up the action.
17. The guilty party must not be a professional criminal.
18. The crime can’t turn out to have been an accident.
19. Motive must be strictly personal.
20. Here follows a list of unpardonable tricks, ways of circumventing the
above rules.
From Thomas Narcejac, Une machine à lire: Le roman policier (Paris: De-
noël/Gonthier, 1975), 97ff. (author’s translation from the French).
How can the authors of ring compositions bear to submit to so much

regulation? But now, comparing the seven to the twenty conventions
for detective fiction I realize that the ring composer is relatively free.
The detective story is severely restricted; the rules maintain the genre
and so provide the composers with a challenge and free space for
creativity. Breaches of the rules are made all the time, and I do agree,
as a reader, that the laxity is a form of cheating that does not improve
the story.
I have no example as yet of detective fiction that conforms exactly
to the rules for ring structures. But this digression on the rules of
detective fiction convinces me that there is no reason why there
should not be one. Some detective stories are structurally nowhere
near rings, some are almost rings, and there may well be true ring
forms to be found.
I will now turn to a very different modern piece that may perhaps
be accounted as a ring variant, because the author clearly says that he
is aiming at a return to the start. Laurence Sterne’s book Tristram
Shandy is considered by many critics to be the greatest English novel.∫
At the same time, it is said to lack structure. Great and unstructured,
it seems thus worthy to be compared to the book of Numbers and the
other works named in Chapter 2 for being judged disorderly and
chaotic. In each case discussed in that chapter a hidden ring was there
to be discovered, and I wondered whether this is true also of Tristram
Shandy. I have reexamined it with the seven rules in mind, also re-
membering the implications of Jakobson’s thesis that parallelism may
crop up anywhere. Now, after the exercise, I have stopped doubting. I
believe that it almost is a well-made ring. This case becomes a ques-
tion of whether the author could have achieved anything so complex
unintentionally, or whether he knew very well how to organize a
chiastic structure for his long novel.
Commentators on Tristram Shandy vary between those saying that
its true glory is in having no structure (we have heard that before) and
others who suggest a structure of a musical or spiraling kind. ‘‘Tri-
stram Shandy, in other words, like a piece of music, is marked by a
kind of concentric involution, a structure—to change the shape of the
79
metaphor, like that of a Chinese box.’’Ω ‘‘Concentricity, the spiralling

pattern of graduated replication and varied repetition, is another such
formal pattern.’’∞≠
Trying to see how these effects are achieved will help us to reflect
on Jakobson’s meaning. Tristram Shandy earns its reputation for dis-
order by reason of endless digressions from any theme. The sense of
spiraling circles is conveyed by moving away from a theme and con-
tinually coming back to it. Long philosophical disquisitions, short
episodes and stories, personal opinions and encounters quite out-
number the chapters on the narrative themes. Sterne himself cheer-
fully admitted his many digressions. They take up most of the space,
and the story of Shandy’s life is swamped. It is like the problem with
reading Numbers: is it a book of laws interrupted by narratives? Or is
it a story of the Hebrew people interrupted by laws? It could be a
mistake to take the biographical narratives to be the main theme of
Tristram Shandy. Turn our expectations around and we could believe
that the scattered philosophy is the main work, and the narrative only
a minor set of strings meant for holding the thing loosely together. It
would then be like a Mannerist painting in which the frame has teas-
ingly become more significant than the picture it surrounds, a kind of
literary joke, a reversal of functions.
Sterne says that his digressions are progressive. He justifies them,
glories in them, and tells us that they develop the book’s intention. He
does not regard them as a weakness of the composition. ‘‘Digressions,
incontestably, are the sunshine—they are the life, the soul of reading.
Take them out of this book for instance—you might as well take the
book along with them’’ (5.22). It is up to us to decide whether this
boast is justified.
In Sterne’s day the absence of form was admired as the sign of true
inspiration. This is somewhat in the way that incoherent glossolalia is
respected as a sign of spirit possession in ecstatic cults. And in the
same way, the Persian commentators cherished what they took to be
the incoherence of the mystic Rumi, taking disorder for a sign of
inspiration. Laurence Sterne worked hard on the representation of
carefree abandon. In spite of this, I still concur with those who main-
80
tain Tristram Shandy is not disorderly. I perceive it as a complex

structure, and a deliberately created one, at that.
Tristram Shandy is much more tightly and richly organized than
the critics allow. The various themes make a well-organized pattern
of analogies (like the thirteenth-century Chinese novels described
in Chapter 1). It is unified by the intertwining of several themes,
developed according to the principle of parallelism typical of ring
structure.
The story centers on two eccentric brothers living in Yorkshire.
The elder, Walter, is a retired merchant absorbed in ancient philoso-
phy, astrology, and science. He is the father of Tristram, the narrator.
Walter’s pretentious learning and his naive superstition is one of the
targets of Tristram’s wit through the book. His brother, Tristram’s
uncle Toby, is a war veteran. Wounded in the siege of Namur in 1695
and invalided home, Toby develops a passionate interest in the subse-
quent course of the war. With the help of his batman, Corporal Trim,
he constructs miniature replicas of the battle sites on the bowling
green in his garden. It is well supplied with water that they turn into a
canal, copying the canal around the fortifications of Dunkirk. In full
army uniform they daily reenact each new campaign. This gives the
first important analogy on which the story is built: on one hand, the
war itself, its battles and sieges, and, on the other, the miming of the
war among the toy fortifications on the bowling green.
Tristram Shandy is divided into two books. The first places great
emphasis on his father’s character, on the events around the birth of
Tristram, and on the progress of the bowling green fortifications.
The second book is strung on Toby’s affair with Mrs. Wadman, a
neighboring widow. While he is working on his fortifications she is
spying on him through the hedge and maneuvering to capture his
affections, another kind of warfare. At first Toby is too engrossed with
his war games to notice; later he tries to elude her schemes. Toby
keeps abreast of the news from Flanders’s battle fields and reenacts it
on his miniature stage, until the war comes to an end with the Peace
of Utrecht in 1713. Once the order has gone out for the fortifications
in Dunkirk to be demolished, Toby, grieving greatly, orders Trim
81
to demolish his miniature fortifications. Having nothing else to do,

he starts to succumb to the widow’s wiles and falls in love with her.
The courtship is described in terms of attack, defense, siege, and
capitulation. So the narrative thread across the book concerns war in
three analogies: real war, war games, and the wars of love. They sur-
face intermittently across the whole book. They are the background
to ever-recurring episodes of false starts, beginnings that come to
naught, disappointments, and new beginnings.
Instead of a coherent argument or dominant plot, the book pro-
ceeds by offering one analogy after another. Body corresponds to city,
bodily ills and strengths correspond to a city’s fortifications, the front
of a siege stands for the front of the face, the nose the center of the
face, the groin at one time the vulnerable outworks of the city under
military siege and at another time the vulnerable male organ. When
the architectural terms are extended to both structures the words
come into collision, continually verging on indecency, and never al-
lowing one meaning to overmaster the others.
Sterne intertwines the two lexical registers all the time, the lan-
guage of military fortifications and the language for human bod-
ies. Uncle Toby hears the specialized military register in almost
any word. When Walter, speaking of grammar, says that auxiliary
verbs are important, Uncle Toby warmly agrees, praising the service
that auxiliaries provide in battle. When Walter refers to the curtain
around his wife’s bed, Toby starts off about the ‘‘curtin’’ that joins two
bastions in a rampart. A conversation about mending the bridge of
the baby Tristram’s nose is mistaken by Toby to refer to a drawbridge.
In the course of the book several narratives present elaborate parallels
between the nose, in the center of the face, and the groin in the center
of the body. In the Oxford English Dictionary ‘‘groin’’ can mean the
snout of a pig and is also a stone pier built for a harbor fortifica-
tion. This word, groin, is in frequent use because the wound Uncle
Toby received was in his own personal groin, which makes the analo-
gies tumble over each other more boisterously than ever. Nobody
but Toby and Corporal Trim knows exactly where the damage was
done, and when Toby presents himself as a suitor, the Widow Wad-
82
man, who has heard the rumors, is understandably anxious to know

whether he has been unmanned.
The glossary for military fortifications in my edition is very reveal-
ing. When Tristram’s mother is about to give birth, his father, Walter,
wants her to have a local scientific practitioner (a man) to attend her
in labor. She wants to refuse. Uncle Toby, trying to defend his sister-
in-law’s preference for the licensed female midwife, is stuck for the
right word. He wants to explain why modesty would stop her from
allowing a man midwife to approach her . . . (left as a gap); the
narrator considers that the rude word the old soldier wanted to avoid
was ‘‘her back-side.’’ But the modest fellow finds that ‘‘too bawdy.’’
He hesitates, and never finishes his sentence, though the narrator
thinks that if he had not been interrupted he would have said ‘‘her
covered way.’’ So we look up the glossary note on ‘‘covered way’’ and
find it stiff with innuendo.
A covered way is ‘‘a space of ground level with the adjoining coun-
try, on the edge of the ditch, ranging quite round the half-moons, and
other works without-side the ditch.’’∞∞ In this context it serves for an
account of the buttocks. Someone scoffed, saying that babies are born
from the front, not the back, but to mention the frontal approach
would be even more repugnant for Uncle Toby’s modesty. Uncle
Toby is a virgin; his ignorance of women and sex is another regular
butt for the merriment of readers.
The war in the Netherlands might be the analogy for the fight for
the life of baby Tristram, or the scene of delivery in childbirth might
be the analogy for the war. The sense of reality is strong; the mili-
tary puns are never so distracting for us to forget that they are talk-
ing about the woman in childbed. But neither birth nor battle, still
less the toy battlefield, is the root scene for which the others are
metaphors.
Of some kinds of books one cannot ask, ‘‘What is it about?’’ This
book is not about something external, it is about itself. Does this
mean that its main object is to celebrate the craftsmanship that it
displays? William Freedman was right to say that ‘‘Tristram Shandy is
about the way it is told. Form and content are one and both are
83
present.’’∞≤ The starts and stops, the interruptions and digressions

are equally present in two ways, in the discombobulated construction
and in its themes. The mutual mirroring between the manifold topics
and the structure of the book is the literary magic that compels the
reader’s attention.
Interruptions occur throughout. They tend to make fun of Walter,
the portentous elder brother, who is continually interrupted in the
middle of his learned soliloquies. At the beginning of volume 7 the
narrator declares that there is nothing that he abominates worse than
to be interrupted in the middle of a story. We have heard this before,
and indeed the very opening of the whole book is about an untimely
interruption of business in hand. Just as Walter is trying dutifully to
concentrate on his Saturday-night sexual routine, his wife distracts
him by asking whether he has forgotten his other weekly duty, to wind
the clock: ‘‘Good God! . . . Did ever woman, since the creation of the
world, interrupt a man with such a silly question?’’
This is the mere fabric of the novel. It will take another chapter
to unravel the pattern of parallels and concordances into which the
material is woven.
84
seven
TRISTRAM SHANDY
T E S T I N G F O R R I N G S H A P E
A pattern of alternating bands was one rhetorical device that the

biblical ring makers used to mark off units of structure. We saw this in
the alternation of law and narrative for the book of Numbers and will
see it again in the chapters on the Iliad in the alternation of nights and
days. In default of an alternating principle, Laurence Sterne relied
entirely on parallelism to mark his units of structure in Tristram
Shandy. This partly accounts for the frequent repetitions.
The next question about the structure is, do the swirling spirals of
Tristram Shandy have anything in common with the construction
of the book of Numbers? Or any connection with ring composition
in general? The period of composition stretched between 1759 and
1767. Recall that Sterne was an Oxford man, he had graduated from
Jesus College in 1737; he was in holy orders and a scholar. He surely
knew the Bible model if only by hearing fellow clerics talk about
Bishop Lowth’s famous lectures on biblical poetry delivered in Ox-
ford between 1741 and 1750.∞ He could quite conceivably have cop-
ied the chiastic model for his book.
Lowth described a chiastic pattern that marks the structure by
parallels linking the beginning, the middle, and the end, so that there
are normally three key points and strong correspondences between
them. It places the main message at the middle, effectively making the
first and last part of the composition into a frame, often very elabo-
rate. The series of chapters running from the middle back to the
beginning go in inverse order from, and in parallel with, the first
series going the other way. I will show how closely the structure of
Tristram Shandy conforms to the pattern described by Bishop Lowth.
I do not think it is quite a coincidence.
The test of what I would regard as a well-turned ring composition
is the middle, the summary that marks the turning point. Tristram
85
TRISTRAM SHANDY
Shandy has no central place. Normally, as in the Iliad, the central place
in the composition recapitulates the beginning and anticipates the
end, thus marking unequivocally the way the book is to be read and its
key analogies. To find the center the first task is to identify the units of
text that have to be paired with each other in two series, the one
descending from, the other ascending back to the beginning. As I see
it, Sterne has used two formal devices for structuring the novel: one is
to divide the work into two distinct ‘‘books,’’ and the other is to use
the constituent volumes as units of structure.
Unlike the book of Numbers or the Iliad there is no need to worry
that later scribes may have organized the divisions of Tristram Shandy.
Sterne wrote his own book. He himself would have deliberately cho-
sen the way his chapters are grouped in volumes, all numbered, nine
volumes in all, some of them very short. There seems to be no special
reason for this arrangement, except to use the volumes as dividers of
sections in a ring-like organization. If that is right, volume 5 would be
the central place in the middle of nine volumes. This would be stan-
dard practice for seventeenth-century poetry, based on Pythagorean
number theory.≤ If these volumes are the units of the ring construc-
tion, then on either side of the middle volume two strongly paired
volumes would stand on either side, in parallel. They are the sign that
there is a ring construction. They will closely correspond, each to the
other, a triangular frame for the middle, starting the return to the
beginning.
If volume 5 is the center, volumes 4 and 6 ought to correspond to
each other as the parallels framing it. I try it out for Tristram Shandy.
It doesn’t work. There is no obvious parallelism between 4 and 6. So
now I try the idea that the middle comprises not one but two volumes,
5 and 6, forming a big central place flanked by 4 and 7.
The result is positive: 4 and 7 are clearly marked as a pair. From
here we can tentatively proceed on the assumption that the central
place includes both volumes 5 and 6. We shall find that volumes 5 and
6 do perform some of the main function of a mid-turn by making the
necessary cross-references to the first and the last volumes, and more.
Evidently the book has structure and the units of organization are the
86
TRISTRAM SHANDY
Fig. 11. Half a ring: The structure of Tristram Shandy.
volumes. That is already a result. We have established that it will be

worthwhile to go further and test whether Tristram Shandy really
exhibits a macro-ring form.
I am only half convinced. The book does not fulfill all the criteria
that distinguish the antique ring forms I have been studying. If it is
judged to be a ring similar in most details to the expanded chiasmus of
the Bible, then one could suppose that Sterne consciously adapted it
from the Bible. This is tempting, but I do not think it is the right
conclusion to draw from the signs of literary structure (Fig. 11).
If it is only a nearly correct ring form, that suggests that it originated
87
TRISTRAM SHANDY
as a structure in the mind of the brilliant author, without his neces-

sarily being aware of all the details of Hebrew poetic structures or
meaning to adopt them. There is a lot of structure here, it is original,
but it does not conform to any conventional formula. That being so,
this book is an exemplar to support Roman Jakobson’s idea of parallel-
ism deep set in the brain itself.
The first of the two separate books of Tristram Shandy is mainly
concerned with Tristram’s father, Walter, and the second with his
uncle, Toby. Our first act of analysis is to draw the line down the
center.
Volume 1 is a like a classic exposition for the whole book. It sets the
scene for the narrative in each half and introduces the main characters.
Volumes 2–5 compose the first half. The end of volume 5 is written as
the ending of the first book. At the beginning of volume 6 the author
announces a new book. Volumes 6–9 constitute the second book, with
the ending in 9. The narrator, Tristram, says in so many words that he
is supposed to bring his theme back to the beginning (6.33).≥ This
entitles us to try reading the second half as an inverted ordering of the
first. It works out very well. When the volumes in the two parts are laid
out in order, the parallelism becomes visible. Reading carefully with
questions about the macro-structure in mind, we find that volume 5
and volume 6 both refer back several times to the information given in
the exposition, volume 1, so the thing is properly linked up.
Given the many years over which the whole book was composed,
the exposition was presumably written, as is often the case, after the
ending had been done. Observe how it takes account of events on
either side of the dividing middle line. Uncle Toby’s wound in the
groin receives attention in the exposition, though it is going to play a
more important part in the second book. This also suggests that the
final version was patched and rearranged for the sake of the unity of the
whole. The process recalls the editorial work of Numbers and Levi-
ticus, which assembled excerpts from very old texts. Ring composition
is an excellent form for scissors-and-paste editing.
The most heavily emphasized links connect volume 2 with vol-
ume 6, and volume 5 with volume 9. It looks like a ring pattern with
88
TRISTRAM SHANDY
Table 5. Three parallels between volumes 2 and 6.
Volume 2 Volume 6
New book New book
Bowling green fortified Bowling green fortified
Peace of Utrecht Peace of Utrecht
the exposition standing outside. This double cross-linking is such an

effective device for connecting what purported to be two separate
books that I feel it cannot have just happened unintentionally. But the
reader shall judge. The volumes 2 and 9 do not respond obviously to
each other as parallels, nor do 5 and 6. Instead, 5 crosses the line
between the two halves to reach forward with unequivocal parallels to
9, while 6 crosses the other way, reaching back to the beginning with
stories that are paralleled in 2.
To assert a parallel with confidence there need to be at least two
distinctive items found in both members of the pair, but nowhere else.
Volumes 2 and 6 have it in common to be the start of a new book.
There are three parallels between 2 and 6 (Table 5).
In volume 2.1–6 Toby and Trim thought up the idea of turning the
bowling green into a toy fortification on which they could project the
course of the war in the Netherlands (1692–1695). Volume 6.21 picks
up the thread from volume 2 almost directly, by describing the house
of Shandy Hall in Yorkshire, the garden and the bowling green, and
the first two years of miniature fortifications.
The moment of transition in Tristram Shandy is the Peace of
Utrecht, infamous in the eyes of Uncle Toby.∂ Reference to this im-
portant treaty connects the beginning to the end and the middle. The
consequent demolition of Dunkirk mentioned at the beginning (2.7)
is a crucial factor in the development of the narrative as it brings the
games on the bowling green to an end in volume 6. This is where
Uncle Toby hears of the perfidious order to demolish the fortifica-
tions in Flanders (there are delays and the actual demolition is not
89
TRISTRAM SHANDY
Table 6. The three parallels between volumes 5 and 9.
Volume 5 Volume 9
Simultaneous conversations in Simultaneous conversations in

study and kitchen parlor and kitchen
Concern for Tristram’s Concern for Toby’s castration

castration
Bridget’s hand gestures Trim’s hand gestures
reported until volume 8). We hear in volume 6 of Toby’s deep chagrin

and sense of betrayal after the Peace of Utrecht (1713).∑
In volume 2.7 Uncle Toby mentions the Peace of Utrecht in con-
nection with the shock he received in the affair with Widow Wadman.
Volume 6 deals with the Peace of Utrecht and the order to demolish
the fortifications and also with Widow Wadman’s affair with Toby
and his falling in love (Table 6). Both volumes 5 and 9 are end-
ings. Three major links between them are as follows: two simultane-
ous conversations in kitchen and front room; two persons feared to
have been castrated; and two hand gestures to demonstrate castration
without indecent language.
The first of the two conversations, in volume 5, has Walter in his
study, having heard of the death of his elder son, Bob. He is harangu-
ing Uncle Toby with a series of profound aphorisms from the an-
cients on the theme of death (5.3–6). At the same time (5.7–10),
Corporal Trim is haranguing the kitchen on the same theme. Then
we have in volume 9.20–26 an argument in the kitchen between
Bridget and Trim about Toby’s wounded groin, while Toby and the
widow are having a genteel discourse on the same thing upstairs in
the parlor.
The second parallel starts when Susannah has held the baby Tris-
tram in the open window to urinate, but the window sashes have been
destroyed, and the window comes down with a crash on the child’s
penis (5.17). Susannah fears he has been castrated and runs to tell
Corporal Trim. He then goes to describe the disaster to Uncle Toby
90
TRISTRAM SHANDY
by mime: ‘‘Trim by the help of his forefinger laid flat upon the table,
and the edge of his hand striking across it at right angles, made shift to
tell his story so that priests and virgins might have listened to it’’
(5.20).
The parallel is in volume 9.20–26. Toby declares his love and pro-
poses marriage to Widow Wadman in the parlor, and they talk about
the married state and the purpose of marriage as given in the Book of
Common Prayer. Mrs. Wadman seems to digress; she asks obliquely
whether the male member of her suitor is intact: ‘‘And whereabouts,
dear sir, did you receive this sad blow?’’ (9.26). Toby circumvents the
question by sending Trim to get the map of the citadel of Namur so
that her curiosity can be fully satisfied. Returning with the map Trim
first goes down to the kitchen to show it to Bridget. She has promised
her employer to find out from Trim the truth about ‘‘that necessary
organ,’’ and Trim tries to explain the location of the wound by plac-
ing Bridget’s hand on his own body, but she has information that
the wound was more central than the manual demonstration implied.
She contradicts him with her own manual display. Bridget is arguing
(against Trim) that Toby has indeed suffered the equivalent of castra-
tion: ‘‘ ‘Come—come—’ said Bridget, holding the palm of her left
hand parallel to the plane of the horizon, and sliding the fingers of the
other over it, in a way which could not have been done had there been
the least wart or protuberance’’ (9.28).
We learn later that Susannah, the maid in Walter’s house, has told
Bridget, and Bridget has told everyone in the village, and so everyone
knows about Uncle Toby’s mutilation. ‘‘In a word, not an old woman
in the village or for five miles around, who did not understand the
difficulties of my uncle Toby’s siege, and what were the secret articles
which had delayed the surrender’’ (9.32).
This is the end of the affair between Toby and Mrs. Wadman; their
passion has been checked. For both of them it has been a frustration, a
false start. The second half of the ring, which started out with her
wiles to entrap him, is now well closed, and even linked to volume 2.
These diagonal parallels, crossing over from volumes 2 to 6, and 5 to
9, systematically connect the two halves of Tristram Shandy. They
91
TRISTRAM SHANDY
show how much Sterne was concerned to make his work a unified
whole in spite of the break down the middle and the wild digressions
throughout.
Our next task is to demonstrate the kinds of correspondences that
make the lateral connections. Bear in mind the fulfillment of these
linking functions is one of the main reasons for the book’s reputation
for disconnectedness. And remember that we are not interested in
thematic correspondences unless verbal indicators support them in
both of the paired sections, and remember that it is word clusters that
count, not isolated words. Following these principles, I find very
strong parallels between 3 and 8, 4 and 7.
In volume 3 we hear the bad news of an accident to the miniature
bridge over the waterworks in Toby’s battleground. It is described as
‘‘crushed,’’ ‘‘splintered all to pieces’’ (3.24, p. 171). Uncle Toby is very
upset. Three chapters later in the same volume, we hear of the new-
born baby Tristram having the bridge of his nose ‘‘crushed’’ (3.27,
p. 174) by the clumsy forceps of Dr. Slop, who delivered him. So
much for the man midwife Walter had selected for his scholarly cre-
dentials. When Uncle Toby hears that Dr. Slop is in the kitchen
trying to make a new bridge, he assumes that the word alludes to his
drawbridge.
Much of volume 3 is taken up with Walter’s anguish over the disas-
ter that has overtaken his baby son. His science and ancient philoso-
phy have taught him that a beautiful, long nose is necessary for a
successful life and that a damaged nose invites disaster. He prides
himself on coming from three generations of long noses. After the
first shock he tries to learn more about noses, and collects a library on
the subject. Knowing something about his learning, we can suspect it
is a library of astrology and ancient magic. This is the first setback for
the young Tristram. Whatever else is related in this volume, it is
dominated by the two bridges, broken by carelessness and causing
deep distress. In this volume there is mention of the demolition of
Dunkirk (3.24).
There follows volume 4 with a long essay on noses and names.
Right across the ring, pointing from volume 3 to volume 8.10, we
92
TRISTRAM SHANDY
learn that not only the drawbridge but the whole of the toy fortifica-
tions on the bowling green are due to be destroyed. Although it starts
with the episode of Widow Wadman falling in love with Uncle Toby
(which is also mentioned in 3.24) , 8.10 reverts to the war in Europe.
Toby has just had news of the Peace of Utrecht. He shares the anger
of the soldiers who regarded the treaty as a dishonorable betrayal. In
his view a war should be fought until the enemy capitulates. Just as
Walter was anguished for the destruction of his son’s nose, Toby was
anguished on learning of the order to demolish the fortifications at
Dunkirk. How can his war games possibly go on?
Eventually, in this volume, Corporal Trim, under orders from
Toby, duly replicates the historical events. Bitterly grieving, he de-
molishes all the model fortifications on the bowling green. That is the
end of the stories about the toy fortifications. ‘‘Dunkirk’’ and ‘‘for-
tifications’’ are a key word cluster linking volumes 3 and 8.
With all this on his mind Toby is unaffected by the attempts made
by the widow upon his feelings. There is a great deal in volume 8
about Widow Wadman, her passion for Toby, and her wiles to attract
his attention. And plenty about his complete unawareness, his inno-
cence and ignorance of womankind. Henceforth Toby has no military
fortifications to play with; he is at a loose end and starts to think about
love (8.18, p. 465). War and love are parallel themes: military terms
are used for the widow’s maneuvers and for Toby’s early resistance
and later for his reversal of roles when he proceeds to lay siege to her
feelings. Most of volume 4 is about Walter’s obsession with names and
his dislike of the name of Tristram, and about his superstitions con-
cerning noses. The theory of names has already been described in the
exposition. His deep pain at his baby son’s disfigurement is soothed
when he thinks that by choosing the right name for the boy the
disaster can be corrected. The child shall be called Trismegistus,
the name will ‘‘bring all things to rights’’ (3.11). The name means
‘‘Thrice the Greatest,’’ a later designation of the Egyptian god Thoth,
the father and protector of all knowledge.∏ It is a typically magical
idea, smacking of his crazy learning, as superstitious as his theory
of noses.
93
TRISTRAM SHANDY
At first there seems to be no connection between volume 4 and

volume 7, the latter is such a total digression, but wait! In volume 4 we
have already been reading about sojourning in France. In volume 7
the narrator, grown-up Tristram, is very ill, so he takes a holiday on
the continent. From the start both volumes are linked by reference to
foreign parts. Volume 4 had the curious tale of Slawkenbergius in
Strasbourg, the man with an enormous nose. When we come to vol-
ume 7 we read another curious story, about two French nuns in a
carriage, abandoned by their driver. Slawkenbergius’s tale is frankly a
digression, but it does have direct relevance to the matter in hand,
noses. In volume 4 Walter’s theory of noses is repeated and Walter
plans to cancel the misfortune of the baby’s broken nose and to save
the fortunes of the family by bestowing on him a magically propitious
name. In volume 7 the story about the nuns seems to stand alone;
nothing connects it with the rest of the book at all. It seems a true
digression that has no relevance to the paired volume 4, except for
taking place in France.
But when we reread both stories, in volumes 4 and 7, with the
structure of the book in mind, we recognize them as the ‘‘progressive
digressions’’ that support the book’s main movement. In volume 4 the
two universities of Strasbourg are embroiled in a furious debate about
whether the great nose is truly a nose, or a false nose made of pa-
per. The Lutherans form the anti-nosarian group, denying that God
could or would have made such a monstrously big nose. The Popish
group are the nosarians; they insist that God could do whatever he
wanted and that he did make this huge nose, and that it is a true nose.
The blasphemous satire makes a devastating backdrop to theological
controversies that arise throughout the story of Tristram Shandy. Far
from a digression, volume 7 ridicules Walter’s theories of noses and
names expounded in volume 4. It is a joke against the French and a
joke against religion.
The further clue that the two volumes are linked is that volume 4
mentions the Abbess of Quedlinburg leading a great procession in
Strasbourg. There are no other abbesses in the whole of Tristram
94
TRISTRAM SHANDY
Shandy except in volume 7. The Abbess of Andouillets, accompanied

by a novice called Margarita, is on her way to find a cure for her
ailments (7.20). Their carriage, drawn by two mules, is driven by a
coachman who addresses them in colorful obscenities. At one point
on the journey he stops for refreshment, abandoning his passengers.
The mules go on a little way, then they stop, and nothing the pas-
sengers can do will make them move on again. Panic strikes as they
imagine themselves at the mercy of brigands and rapists. Margarita
tells the abbess that she knows two words that will force any horse,
ass, or mule up a hill, but alas, the words are sinful to pronounce, they
cannot even contemplate saying them. The abbess hits on a solution:
the thing is to divide the words in half, repeating the first syllable and
then repeating the second syllable. So they say: bou, bou, bou, 100
times, and then ger, ger, ger, 100 times. Then the same for the other
word, they say fou, fou, fou, 100 times, followed by ter, ter, ter, 100
times. They have not sinned; they never actually said the bad words.
The joke is based on the ambiguity of the French bouger, to stir,
budge, move, with suggested allusion to bougre (bugger), and the
indecency of foutre, to fuck. It must have been a successful strategy
because, though the mules did not understand the words, we are told
that the devil did, and he left the women in peace.
One might well ask, ‘‘What on earth is this story doing in this
book? It seems to have nothing to do with anything.’’ But if you
accept that the book is constructed as a ring, the answer is that the
story of the superstitious nuns is placed just here in order to match
Walter’s crazy science and the wild superstitions of volume 4. French
Catholic superstitions outdo Walter’s learned magicality and form
the link between 4 and 7. Furthermore, the story gives scope for an
eighteenth-century Protestant gibe against Catholic superstition.
I will also suggest that this episode is placed where it is because it
focuses on the idea of interrupted speech and emphasizes the master
analogy that lies at the center of the literary spiraling, thus giving
unity to the book. The dominant metaphor may be sex interrupted,
or equally well it may be interruption illustrated with anecdotes about
95
TRISTRAM SHANDY
sex. It is, as the narrator affirms, that the diversions and digressions
promote the theme, they illuminate an essay on interruption. If we
have to decide what this book is about, we could well say it is about the
disorder of life, its incessant ups and downs and sundry misadven-
tures. But as a topic that is really vacuous. There is not one theme, but
many. I claim that the disarray is only superficial. This is a book with a
crystalline structure. It is so skillfully contrived that, if it is about
anything, it is about itself. The way it is written exemplifies what is
written. Cut it at any point, the same pattern is there. Look at the
ending, for example.
It is a surprise to find a book that was written in such distracting
conditions over so many years, including sickness and good health,
presented in such a strong, subtle, and consistent structure. Written
in the racy style of Rabelais, with the weight of Erasmus’s thought,
and with the tolerance of Montaigne, its author was obviously set on
greatness.π If someone so opportunistic as Sterne is going to give
himself maximum scope to play with analogies and to achieve reple-
tion by their intertwining, he would have had to invent something
like a ring form.∫ Sterne is like Turner working away at his com-
position in the Royal Academy until the very last minute allowed for
the day of vernissage. The symmetrical structure results from his will
to perfect his attack on the vanity of learning. He had to pull together
a book that had appeared in parts over many years and had already
been set into two volumes. The structure that crosses over two end-
ings and two beginnings would be an original invention, worthy of
the author’s creative mind. I do not believe that it imitates an antique
ring.
This being said, Sterne’s decisions about the ending become the
more interesting. In effect, one ending is not enough. The first end-
ing does everything required to close the narrative in a coherent and
comprehensive sequence. The second is a latch, a little surprise that
completes the tale at a wider and deeper level.
Uncle Toby’s courtship of Widow Wadman occupies most of the
second book, volumes 6–9. It was always phrased in terms of war,
attack, defense, strategies, and sieges. It comes to its climax when
96
TRISTRAM SHANDY
Toby, supported by Corporal Trim, both dressed in full regimentals,

set out to declare love, respectively, to Mrs. Wadman and her maid,
Bridget (see Chapter 6 above). Book 1 (vol. 2) had made a great point
of Toby’s ignorance of the female sex. At the end (book 2, vol. 9), he is
nervous as they approach Mrs. Wadman’s door. ‘‘Now my uncle did
fear, and grievously, too: he knew not (as my father had reproach’d
him) so much as the right end of a woman from the wrong’’ (9.3). The
reader can see that the whole second half of Tristram Shandy has
turned upon Tristram’s uncle’s lack of education in the matter.
At the beginning of the second volume there is a reference to ‘‘the
shock Uncle Toby received in the affair of Mrs. Wadman’’ (2.7). The
expression ‘‘shock’’ is justified when we get to the distressing events of
volume 9. She tries to find out whether his wound at the siege of
Namur damaged his genital organs, and he tries to satisfy her curi-
osity about where he received the wound by sending for the map of
the battlefield. ‘‘The shock’’ is his realization that the true informa-
tion about his castration and impotence, so far from being hidden
from Mrs. Wadman, was common knowledge locally and in the vil-
lages around. Trim finally told him in three words. Toby had inter-
preted the lady’s thousand sympathetic inquiries about his wound as
signs of love; now he discovered they were reasons for rejecting his
suit (9.31). Toby will never marry Mrs. Wadman; their mutual attrac-
tion has come to naught. This is the definitive ending of the story of
Uncle Toby and the affair of Widow Wadman.
The negative conclusion rounds off the many reflections on con-
flict, hesitation, uncertainty, and frustration that have been scattered
through the whole book. It has accomplished enough of the things
that the ending of a ring is supposed to do, by way of recalling the
beginning and intervening episodes. Yet Sterne is right not to have
been satisfied. It would be a mistake to leave Uncle Toby despondent
at the end. He is the readers’ favorite character. So the story goes on,
there is a latch.
When Walter hears of Uncle Toby’s distress he is furious. He
launches on a vitriolic lecture against Widow Wadman, denouncing
womankind and lust as the source of all evil (9. 32). To give Walter the
97
TRISTRAM SHANDY
last word on sex makes a clever tie with the first chapter of the first
volume, where Walter’s interrupted sexual activity preceded the con-
ception of Tristram. While his tirade is in full spate the servant,
Obadiah, rushes in (another interruption) with news from the farm—
the cow has calved! The cow had been so long in calf that it was feared
that nothing was going to happen. She was suspected of a swollen
belly, like a ‘‘false pregnancy’’; alternatively the bull was being sus-
pected of impotence. Now there were rejoicings at Shandy Hall; both
bull and cow were vindicated, and the female sex in general.
Much as Sterne was fêted and much as this book was praised, there
had always been critics who hotly criticized him for indecency, and
not without provocation.Ω It would have been a cowardly concession
to those critics if he had ended with Walter’s diatribe against sex. Yet a
hostile critic might complain that the story about the cow proving her
fertility is not a worthy ending for a book with such high philosophic
pretensions. The latch makes an analogy between animal sex and
parturition and the deliveries of a cow and of Tristram’s mother. Is
that not indelicate? And is it not too trivial a tale for concluding so
many sage philosophical reflections?
In response to the critic I would say that the connection of the latch
with the beginning has unparalleled richness. The whole story started
with Tristram’s mother being obliged to lie in for the birth at Shandy
Hall instead of, as she had wished, in London (vol. 1). Her marriage
settlement had given her the right to a London medical specialist
when she went into labor, on the one condition that she did not
contrive a false pregnancy, a swelling of the belly, as an excuse to get
to London. Unfortunately, before she conceived Tristram she had
claimed to be with child and had persuaded her husband to take
her to London to give birth, as was her right. But the journey was
for nothing; she actually had a ‘‘false pregnancy.’’ So that is exactly
why, when later on she was really pregnant, she had to put up with the
discomforts and risks of giving birth in rustic Yorkshire, to endure
the grotesque male midwife, his clumsiness, and the breaking of the
child’s nose. The end of the whole book gives us a new version of the
opening theme.
98
TRISTRAM SHANDY
Perhaps Sterne recognized some incompatibility when he half-

dismissed the incident as a ‘‘cock and bull story.’’ ‘‘Cock’’ has phallic
reference, and ‘‘bull’’ means a ‘‘blunder, or inadvertent contradiction
of terms.’’ Christopher Ricks’s essay on ‘‘The Irish Bull’’ gives reason
to believe that Sterne was fully aware of this meaning of ‘‘bull’’ and
meant to end aptly with one more pun.∞≠ As much as the book is about
philosophy and morals, it is also about laughter, life, and love. A
bawdy latch befits the merry book.
Unraveling the structure of this book has been very worthwhile for
my central project, not a digression at all. The cross-referencing is
found to have been so careful, it suggests an overall structure that
was pieced together gradually. The more the book grew, the more it
would become unbalanced and wild, the more the author would have
sought for ways of tying the two halves together. So he has an exposi-
tion that embraces both sides in its scope. Then he links the two book
openings (2 and 6) with a long and elaborate parallelism and does the
same for the two book endings (5 and 9). After that it is easy to involve
the remaining volumes in the pattern by elaborating them as parallel-
isms. Finally, the latch is a brilliant ending; connected with volume 1
it makes what Bible scholars call an inclusio, a containing envelope for
the whole book.
With all they have in common, we must note the differences be-
tween the ring shape of Numbers and this ring of Sterne’s. The latter
has no mid-turn section that highlights a crisis in the plot, no tangle
of dilemmas is summarized here, no dangers to be evaded or prob-
lems to be solved. The turn is made more brusquely, and very de-
cisively, by announcing that one book is ended and another is begun.
It is a crude solution to the problem of working back to the beginning.
The second difference is that there are no alternating bands across the
design of the whole book. These two deviations from the ring shapes
that I have observed suggest strongly that the structure of Tristram
Shandy is the author’s personal response to the challenge of writing in
a ring.
If this is right, it explains the recurrence of ring shapes all over the
globe in terms that support Jakobson’s thesis. The brain works by
99
TRISTRAM SHANDY
making parallelisms. No other explanation is necessary. We should

expect regional variations and variations over time. It is outside the
scope of this study to start distinguishing the eastern Mediterranean
ring structures from Persian or Indian kinds. Given the initial hom-
ing impulse, the project of making the ending meet the beginning
(that is, given the wish to write in rings), we can expect that local
variations will be strongly elaborated wherever literature is highly
valued. Why authors choose to clip the end of their composition onto
the beginning, divide the ring in half, and make the two sides re-
spond to each other in parallel formation, or why they should choose
to write in rings at all, is a deeper matter. Presumably, symmetry,
balance, matched proportions, and repetition have just as much to
do with the way the brain works as the structure of language and
grammar.
Lighthearted and profound, mundane and transcendent, Tristram
Shandy hits both poles of order and confusion at once. Interruption is
not the meaning; it is a means, an instrument that keeps pointing to
how the book has been made, and keeps drawing the whole book
together in a laugh.
We have now done a lot of spadework on understanding what ring
composition is and is not. It is time to apply it to unsolved problems.
The next chapters will study Homer’s Iliad.
100
eight
TWO CENTRAL PLACES, TWO RINGS
T H E I L I A D
For those who know and love the Iliad, this and the next chapter are
going to be frustrating. In this book the concern has to be entirely
with the construction, nothing about the beauty, humor, and emo-
tional power of the story. By this narrow focus we discover that the
Iliad adheres faithfully to the seven rules for ring composition, though
admittedly this claim is controversial.
Many scholars would now agree that the Iliad is highly structured
and that the form is annular. Over time there have been many sugges-
tions about the overall structure, but no one scheme is generally
accepted. This is the paradox in its reception. Some of its most ardent
admirers used to deny that it has any overall structure at all. A text
that is like the Bible in having been transmitted from ancient times by
many voices and different scribes cannot be free of discrepancies and
anomalies. In other words, in the eyes of these scholars it would be a
waste of effort at this point in time to look for a consistent ring
structure. Some have maintained that it is a loosely connected bunch
of poems and stories, individually brilliant but separately composed
and brought together in almost haphazard fashion. It seems too me-
chanical to apply the rules of ring composition to this most gloriously
inventive, riotously varied, powerful poem. As I have found, and said
earlier, such repudiations of structure are like signposts saying: ‘‘Here
lies a hidden ring composition.’’
I do not have a lot of sympathy with the skeptics. In fact, I do not
see one ring, but I do see two. The main ring organizes the whole
temporal structure into a numerical pattern of groups of days. At the
midpoint of the main ring there is a minor ring that is organized in
parallels, night to night and day to day. It is possible that at some
moment in history one ring was superimposed on, or inserted in, the
other. I personally think that the elegance of two rings, one taking
101
off from the mid-turn of the other, the first being a frame for the
second, is just the kind of poetic contrivance that might be expected
of Homer. But that is speculation. Here I am concerned with what we
have before us now.
As to structure, there have always been skeptics, some of whom
date back to antiquity, starting with Josephus, and working forward
through the centuries to the present day. In the eighteenth century,
John Myres wrote, ‘‘Homer’s excellencies were being depreciated and
misinterpreted by the stricter French classicists mainly on the ground
of their anomalies and the difficulty of reconciling Homer’s treatment
of his subject with modern conceptions of what classical literature
should be.’’ He remarked that l’abbé d’Aubignac ‘‘as a lawyer, was
sensible of the discrepancies and anomalies of the action’’ and de-
plored the ‘‘looseness’’ and ‘‘wayward construction’’ of the poems as
they stood.∞ That judgment is already a familiar refrain in the recep-
tion of ring composition.
Several scholars have tried but failed to convince their colleagues to
accept their version of the poem’s structure. They usually lack some
means of distinguishing the organization of the poem from its theme.
Trying to recognize structure by the ordering of themes is bound to
be highly subjective. A formal literary structure is not the same as the
thematic structure. A formal structure is not based on contents; it is a
set of empty frames or containers for the contents. There should be
always some fittingness between contents and structure, like that be-
tween the rhyme and meter and the thematic content of a poem, but
searching among the contents will never of itself reveal the structure.
One simple method is to start simply with a search for repeated
formulae that make a pattern. It can start with words, rhymes, asso-
nance, alliteration, and can show also in nonverbal markers, such as
line endings, and punctuation marks, commas, full stops, or exclama-
tion and question marks. Think of the poet having to pour a newly
formed idea into bowls and jugs of different shapes, some of which are
at hand in local conventions, and some of which have to be invented.
The conventional variation in poetic structures will be found to
depend on the type of social event for which the poem is intended.
102
Dirges, military marches, funerals, and sermons need to meet distinc-

tive public expectations. The more traditionally hallowed and public
the occasion, the less freedom the poet enjoys, as for example the
strictures that the Somali poets have to observe (Chapter 2 above).
Intimate lullabies and love songs are not necessarily freer, because
their form can only take up the residual space left by the conventions
for public performance. The history of the local art forms will tell us
what the conventions are. This mode of enquiry is very congenial to
anthropologists because it locates the art form in the social events that
call for it.
Some myths are recited to celebrate boys’ initiations.≤ In another
case the myth of creation is recited at the beginning of a new political
period to affirm the relations between the different shrines of the
community.≥ Oliver Taplin follows this clue when he goes beyond
recording the phrases, words, and themes of a great myth and records
the occasions when it is recited and any prevailing circumstances that
may assist interpretation.∂ He considers the great length of the Iliad
and the intricate and overlapping cross-references that hold the huge
work together. He proposes accordingly that the Iliad is intended to
be heard or read as a whole, and delivered to an audience who are
familiar with the story, or at least with the style. He describes the
sensitive reciprocal relations between the audience and the poem. I
believe this method of using the social context to define the number
of parts in the poem to be valid.
Taplin’s next question is when and how such an audience, large
enough and distinguished enough to support the bard’s fame, will be
gathered together for long enough to hear the whole recital of the
Iliad. Presumably the parts of the poem would have been worked out
to fit into the given temporal structure. A three-day festival would be
needed to support this poem, so it would accordingly be divided into
three parts, one for each day. He proceeds to look carefully at the
openings and endings, and at cross-references between them, to iden-
tify the three parts.
Taplin’s method and argument are in principle convincing, but
when it comes to finding the precise points of the divisions between
103
parts he lacks a criterion. Size is the principle that has given him a
three-part poem; he continues to rely on size of the narrative events
to identify the three parts. For example, he suggests (to me uncon-
vincingly) that the middle of the poem identifies itself by recording
the biggest and longest battle of the siege.
There is another snag. His analysis only takes account of a small
minority of the days that are listed in the tale. Noting that ‘‘about
three quarters of the days covered in the whole poem are lumped
together in blocks in the opening and closing phases,’’ he decides to
ignore them. He seeks the structure by studying the days of action.∑
He arbitrarily concentrates his analysis of the shape of the Iliad on
the eight days of actually narrated events. I doubt that it is justifiable
to omit the long blocks of twelve or nine days that are listed. The
group of twelve days when the gods are absent, also the nine days of
plague, the twelve days after Hector died, the twelve days of Trojan
mourning—that is an awful lot to ignore. If we follow Cedric Whit-
man’s account, these very same blocks of days are essential elements
in the Iliad structure of narrative time.∏
Cedric Whitman’s appreciation of Homer’s structure has the dis-
tinction of not confounding form and content. His method is first to
ascertain the form and then to find what meanings have been pre-
sented in it. The form, as he sees it, is a pattern of days, plausible
enough. He proceeds to count all the days. Like Taplin, Whitman
regards an eight-day period of action as the centerpiece of the whole
poem. This period is designed as a set of symmetrical opposites; it
begins with a quarrel and ends with a reconciliation. The two sec-
tions (introducing and concluding) that frame it are both arranged in
blocks of days ordered in an interesting numerical pattern. The first
block is grouped into a ‘‘geometric structure’’ based on ‘‘the magic
numbers 3 and 9.’’π
In book 1, on the first day, Apollo’s priest, Chryses, appeals to
Agamemnon for the return of his daughter; the appeal is rudely re-
fused. After this day follows the group of nine days of plague sent by
Apollo in punishment for the insult to his priest; then the one day of
104
the assembly called by Achilles, when he quarrels with Agamemnon, a

day that includes the voyage of Odysseus sent to escort Chryses’
daughter home. One, plus nine, plus one, that makes eleven days. On
the twelfth day another separate block also comes to an end; the gods
return after their twelve days’ absence from Olympus. On this count
the pattern of (1–9)–(1–12) brings book 1 to an end.
This time series is the introduction for the central story of the
Trojan War, but the latter is presented over eight days, and they are
organized on a different principle. They are not an undifferentiated
block but a detailed series of specified, identifiable days alternating
with identified nights. They form a second ring. It is the mid-turn of
the first ring that I am calling the numerical ring. It is obviously so
important that it needs a separate chapter.
Book 24, the last book in the whole poem, is the inverted, numeri-
cal match to the introductory block of book 1. Book 24 starts with the
twelve-day period of Achilles’ mourning for dead Patroclus. Then
there is one day for King Priam’s visit to Achilles’ camp to ask for
Hector’s body and to convey it back to Troy. Nine days are allowed
for the Trojans to gather wood for the funeral pyre and, lastly, one for
the day of cremation and burial (24.665–67).∫ That makes (12–1)–
(9–1) for the concluding series of blocks of days. They are arranged
inversely to the order of the first group, (1–9)–(1–12). With the eight
days in the middle Whitman makes of the whole numerical scheme a
regular ABCBA structure: (1–9)–(1–12)–(8)–(12–1)–(9–1).Ω This is
a purely formal structure. It posits nothing about the contents, the
most abstract, empty pattern imaginable.
Whitman’s insight about the macro-scheme expressed in a numeri-
cal pattern is dazzling. The two blocks of days from book 1 and book
24 (which Taplin discarded from his examination of the structure)
function to complete an abstract model of the whole pattern, a numer-
ical chiastic scheme. The beginning series is matched by the end series,
and in the very middle there lies the story of the eight-day war. As the
central eight days are arranged in another ring (to be demonstrated) it
would be more accurate to show the whole structure as two rings, one
105
taking off from the middle of the other. I repeat: the outer ring is a
frame; the first half consists of blocks of numbered days, and the
second ring in the middle is the story of the war, structured on nights as
well as on separate days, after which the second half of the outer ring
resumes and concludes with a parallel list of blocks of numbered days.
As to the matching of themes, it is well known that the themes of
the introductory half-ring, book 1, make unmistakable analogies with
the closing half-ring, book 24. In the first a father (Chryses) comes to
the Greek commander to ask for the return of his child, Briseis. He is
refused. In book 24, again a father (Priam) comes to the Achilles camp
to ask for the return of his child’s body. This time the request is
granted. The beginning is conflict, the ending is reconciliation. It
strengthens the case that Whitman uncovered the original divisions.
For my discussion of the war I will use the patterns of days and nights
as editorial divisions, and only use the editors’ numbered ‘‘books’’ and
lines for reference.
The count of groups of days in the numerical ring starts with the
appearance of Apollo’s old priest, Chryses, before the whole army in
council (1.1–42). He brings a ransom and begs for his daughter to be
returned. Against the advice of the whole assembly (1.22–24), Aga-
memnon roughly refuses the request and sends the old man away with
insults. Chryses prays to Apollo to punish the Greeks. This is one day.
Then follows another homogeneous block of time in which no
nights or particular days are mentioned. Nine days of plague are sent
by Apollo in punishment for the offense to his priest, Chryses. The
text says, ‘‘On the tenth day’’ (1.54). Presumably that would be the
sum of the first day plus the nine days of plague, which is still raging.
On this tenth day Achilles calls an assembly of the troops and de-
mands an enquiry into the cause of the plague. The diviner, Calchas,
finds Agamemnon’s intransigence responsible. In the hope of getting
Apollo to end the plague, Agamemnon is persuaded to give back the
girl. Reluctantly he sends her home to her father’s island escorted
by Odysseus. (That would be the same tenth day.) Agamemnon com-
pensates himself by taking Briseis, the girl Achilles had won in war.
In anger Achilles withdraws his troops from the war against Troy.
106
The Greeks are now facing the Trojan army without their strongest
fighter (1.100–412).
On arrival at Chryse, Odysseus makes sacrifice to Apollo, who then
stops the plague (1.457); then follows feasting and singing all day long
(1.474). They are still on the tenth day. Odysseus and his crew sleep
on the island overnight (1.476) and take their boat back the next day.
‘‘When the young Dawn showed again with her rosy fingers, they put
forth again to sea toward the wide camp of the Achaeans’’ (1.477).
That will be the eleventh day.
On day 1 Thetis tells Achilles (1.23) that the gods left ‘‘last eve-
ning’’ for Ethiopia and that they will return after twelve days. In other
words, their twelve days of absence started the day before the first day
of the eleven days of the quarrel sequence. After Odysseus has re-
turned from escorting Chryses’ daughter on the eleventh day, we duly
learn: ‘‘But when the twelfth dawn after this day appeared, the gods
came back to Olympus’’ (1.493).
It has been carefully worked out. The gods were absent one day
before the insult to Chryses, absent during the nine days’ plague and
absent for the day of quarrel between Achilles and Agamemnon. Zeus
learned about it from Thetis on their first evening back (1.494–510).
Our reckoning of these two distinct introductory periods confirms
Cedric Whitman’s calculation of a system of (1–9)–(1–12) days. The
day of the gods’ return after the eleventh day is a new start: this
twelfth day is also the first day of the next series. As soon as the gods
are back we are into the second ring. Now follows the central period
of eight days of fighting. It is a new ring that serves as an oversized
mid-turn for the numerical ring, but it also has its own strongly
marked internal mid-turn (Fig. 12).
The eighth day of fighting marks the end of the mid-turn of the
outer, numerical ring. After the eighth day the numerical system takes
over, with the last two blocks of days. One is the set of twelve days for
Achilles’ mourning the death of Patroclus, in book 23. Then book 24
records one day for Priam’s visit to Achilles and the return of Hector’s
body to Troy, followed by nine days of Trojan mourning, followed by
one day for the burial. It makes a pattern of (12–1)–(9–1).
107
∞ Numerical ∞
Ring
Ω Ω
∞≤ ∞≤
∞ ∞
Nights
and
days
9
Mid-Turn
Night ∂
Fig. 12. Two mid-turns in Whitman’s two-ring model.
Whitman’s Numerical Ring

Bk. 1 The Quarrel (1–9–1 days)
Bk. 1 The Gods Abroad (12 days)
Bk. 23 Mourning for Patroclus (12 days)
Bk. 24 Mourning for Hector (1–9–1 days)
This is poetry by numbers, not at all the usual kind of ring structure.
Instead of matching words and themes over two halves of the poem,
the only matching that is done is according to the numbers of days and
the sums thereof. It is clever, abstract, entirely formal, and peculiar.
108
At the center of the introductory and valedictory pieces is the mid-

turn. It is normal practice in ring composition for the mid-turn to be
arranged as a ring. We have seen that the mid-turn of the Iliad is so
big that it dominates the whole poem prominently enough for some
scholars to feel justified in leaving the surrounding frame out of ac-
count. This central ring is not a formal continuation of the numerical
ring. It corresponds to the ring within a ring described by Otterlo, not
a digression but an expansion. It is a set of nights and days organized
by the principle of alternation. Counting them is difficult because the
division by chapters distracts the eye, and yet it has to be done as it is
also a numerical structure (Fig. 13).
In this ring the eight-day model of days and nights in the center of
the numerical ring is exemplary. It is faithful to all the conventions.
The mid-turn is the ‘‘central place’’; it conforms to Alastair Fowler’s
analysis of the role of central place in seventeenth-century poetry.∞≠
All the meaning is to be found there. We saw in Chapter 3 above that
the first basic rule is that the ending should be made to correspond
to the beginning, and both should correspond to the mid-turn. To
achieve this, the exposition or prologue must be designed to antici-
pate the mid-turn. The second convention is that the ending, when it
joins up with the beginning, must make a clear closure while also
using key words from the exposition and the first section. The third
and fourth rules require a mid-turn that divides the composition into
two. The mid-turn must be strongly marked. This is done by laying
it between two clearly marked individual sections that match each
other. They will be matched by content and by clusters of key words,
one series looking toward the middle, the other looking back in in-
verse order. Lastly, the mid-turn corresponds to the exposition, using
the same key words and repeating the themes announced there. Since
by definition the ring ending must join the exposition, both the ex-
position and mid-turn refer to the ending, thus putting the whole
piece into a tight corset.
The mid-turn is not in the middle in any quantitative sense. The
best way to recognize it is by the two supporting series flanking it on
either side and showing a conspicuous correspondence to each other.
109
Priam comes
NIGHT
to Achilles.
G 8
Gods return, DAY DAY Funeral games,

feast, laughter. 1 8 laughter.
NIGHT NIGHT
1 7
Hector’s DAY DAY The pyre for

promise. 2 6 Patroclus.
9 NIGHT NIGHT 7
2 6
Honoring DAY DAY Hector’s death

the dead. 3 6 and mutilation.
NIGHT NIGHT
3 5
Battle for the Battle for the
wall. DAY DAY ditch.
Eagle portent. 4 5 Eagle portent.
Zeus prophesies. Patroclus killed. G
Patroclus’ death. NIGHT
4
Central Place.
The night embassy to Achilles’ camp.
Fig. 13. The Iliad: A pattern of days and nights.
In Numbers we saw that the mid-turn story of the violent punishment

of the Levites’ revolt is supported on either side by matching laws
regulating offerings due to the priests and Levites. Here in the Iliad
the events of the mid-turn at night 4 are supported on either side by
the matching pair of battles. In the first battle raging through day 4
(8.1–485) the Trojans press the Greek invaders so hard that they are
110
about to board the ships and set them on fire. A turn is marked by the
very fact that the besiegers of Ilium suddenly find themselves be-
sieged and on the defensive. In the second battle (11.1–18) victory of
the Trojans seems imminent, so Achilles sends Patroclus into the fray,
bearing his arms. Patroclus is killed by Hector. This will prove to be
the real turning point, as it will bring Achilles himself into the fight-
ing. The effect is to turn the mid-turn into a triangular three-part
section, a chiasmus, ABA, another ring (Box 3).
Box 3. Dawns and nightfalls for eight days of war.
Day 1. 1.494, ‘‘But when the twelfth day after this day ap-
peared, the gods . . . came back to Olympus’’; 1.601, ‘‘Thus
thereafter the whole day long until the sun went under they
feasted in Olympus.’’
Night 1. 1.605, ‘‘when the light of the flaming sun went un-
der they went away each one to sleep . . .’’ (in Olympus).
Day 2. 2.48, ‘‘Now the goddess Dawn drew close to tall
Olympus with her message of light.’’ The fighting begins, two
truces, two single combats. This day continues to Book 7.
Night 2. 7.282–293, ‘‘Night darkens now . . . good to give
way to the night time.’’
Day 3. 7.381, ‘‘Then at dawn Idaios went down to the hollow
ships.’’ Message to Greek camp, 7.420, ‘‘Now the sun of a
new day struck.’’ 7.433, ‘‘But when the dawn was not yet, but
still the pallor of night’s edge.’’ Truce, both sides gathered in
their dead.
Night 3. 7.465, ‘‘The sun went down . . . they took their
supper’’; 7.478, ‘‘all night long Zeus was threatening evil . . .’’
Day 4. 8.1, ‘‘Dawn the yellow-robed scattered over all the
earth.’’ Great Battle; Zeus’ eagle portent, Zeus foretells
death of Patroclus, and entry of Achilles to war.
111
MODES AND GENRES

Fig. 6. Father and son confident together. Drawn by Pat Novy.
that he knew what would happen; he trusted in God’s keeping his

word, and he was right to do so, and so was Sarah.
There are also cross-references and smaller chiastic twists inside
the big ring. For example, verses 11 through 16 start and end with
God’s call, so enclosing the whole of the second section as a second
ring, or a subset of the first. Verses 7 and 13, and 4 and 13 are also
elegantly cross-linked, creating a tightly integrated whole. I hope that
these short examples show how richly worked the small rings can be.
Placing them here will help me in the next chapter to describe how a
ring composition, large or small, is constructed.
26
1. Bk. 9.1. The first reference to the beginning (Bk. 2) starts with
Agamemnon’s night speech inviting the men to go home, saying
that he sees no hope of capturing Troy (9.16–29). This is a near
replication of book 2, the first day of fighting, when Agamemnon
tests the assembled troops with the same speech (2.109–141). The
first time, Odysseus roused them to continue to wage war and they
armed again. This time it is Diomedes and Nestor who rally the
men.
2. The second return to the start refers directly to the quarrel: Aga-
memnon makes a formal apology for his injustice toward Achilles
in book 1: ‘‘I will not deny it, I was mad, I will not deny it’’ (9.115–
16). He then lists the vast wealth he promises to give to Achilles to
persuade him to relent and return.
3. Third, the leaders decide to send envoys to Achilles that very
night, to tell him of Agamemnon’s repentance and persuade him
to return. This involves reference to the initial quarrel. Phoenix,
Aias, and Odysseus set out, they are well received and feasted at
Achilles’ camp. Odysseus tells Achilles that the Argive fleet will
not be able to withstand the attack of the Trojans led by Hector (9.
230–306) and begs him to return to the fray. Achilles refuses.
Phoenix beseeches him again in a long speech.
4. Achilles’ anger is not softened. This connects with the valedictory
section, which actually refers to the anger which Achilles wreaks
on the corpse of Hector, until the gods are disgusted and send
under heavenly escort his old father, Priam, to stop him (Bk. 24).
5. Bk. 9.702. Back at the camp of Agamemnon the envoys’ report
describes the unassuaged anger of Achilles. This is the end of
chapter 9. The night embassy fulfills the function of referring to
the beginning by frequent references and connects it to the end by
continual reference to the anger of Achilles.
So far, the mid-turn does exactly what it should do. But chapter 10
is not accommodated to the conventions. On the same night there are
two spying expeditions, one team from the Argive camp, the other a
spy sent out by Hector for the Trojans. It is the end of the fourth
113
night, the end of the mid-turn. The recital here of an episode that
does not contribute to the mid-turn’s function is surprising in a poem
that so meticulously follows the ring conventions.
The two supporting days on either side of the fourth night indicate
their quite obvious parallelism. On day 4 the battle begins: ‘‘Now
Dawn, the yellow robed scattered over the earth’’ (8.1). It continues
all day. After great carnage ‘‘the darkness came too soon’’ (8.500). On
day 5, ‘‘Now Dawn rose from her bed . . .’’ (11.1). At this point a new
battle rages through seven more books, to book 18.235.
Two other parallelisms pair these two days (4 and 5). On day 4 Zeus
sends up his eagle (8.247), which the Argives take to be a good por-
tent for their armies, and Zeus also prophesies Patroclus’s death and
Achilles’ entry into the war. On day 5 the eagle portent is repeated
and the prophecies of day 4 come true. The supporting columns of
the mid-turn have exactly fulfilled the rhetorical rules for a ring com-
position. After the mid-turn the fortunes of the war change direction,
an example of mood change. It remains to look more closely at the
two sequences of chapters that run on either side of the mid-turn.
114
nine
ALTERNATING NIGHTS AND DAYS
T H E I L I A D
The fundamental question keeps recurring. What sort of mental

discipline does this literary form imply? From the list of rules one
might suppose that it exerts a heavy-handed control. To be able to
write in this exacting form might suggest a rigorous mental set in a
strongly controlling culture. Is ring composition the tip of a psycho-
logical iceberg? Is it an example of the cultural control of creativity
that the nineteenth-century anthropologists attributed to the remote
peoples they called ‘‘primitive’’? The Iliad is renowned as the most
creative literary work of our Western civilization. The great advan-
tage of being able to present it as a ring is that it gives a resounding
answer, No!
Studying the construction of the Iliad suggests that the rules of ring
composition are not like leg irons that constrain the poet’s soaring
spirit. Remember that Milton wrote Paradise Lost in exact mathemati-
cal proportions to celebrate the unity of knowledge and of creation.
Being able to use such rules for poetry is a cultural achievement in
itself; the amenities that the poet appropriates to realize the project.
Moreover, instead of suppressing poetic invention, these rules stimu-
late and support it. And incidentally we discover that they are pliant
—the thing can be patched and mended from time to time without
anyone noticing, so long as the general structure is preserved.
We have noticed already that the most important technical prob-
lem for the composition in ring form is to demarcate clearly the units
that are going to be placed in parallel with each other. For short
poems it is enough to place a few key words to indicate the clues and
responses, as in the Isaac story. Longer pieces face several technical
problems. The beginning and ending of every section needs to be
clearly marked so that there is no doubt about defining each one for
matching the equivalent set on the other side of the virtual split down
115
A LT E R N AT I N G N I G H T S A N D D AY S
the middle. The reader has to discover what marking techniques the
author used. When the marking methods are consistent through the
piece, they endow it with unity and integrity.
The book of Numbers, as we saw in Chapter 4, marks its internal
boundaries by alternating laws and narratives. Many traditions of
alternation exist in the epic form—for example, alternate recitations
in prose and poetry (called prosimetrum).∞ Male and female voices
were alternated in the singing of Latin dirges. Even in modern de-
tective stories, the narrative generally proceeds from two sides in
alternation. Chapters that describe the events from the investigator’s
standpoint alternate with chapters describing the events in the sus-
pects’ lives. Not only does alternation mark the beginning and end-
ing of equivalent sections, it unifies the piece with a common struc-
ture, creating overall balance and formal symmetry. And it relieves
monotony.
Alternating bands across the whole book indicate both where one
section ends and where to look on the other side of the ring for its
equivalent matching section. The principle of alternation is a com-
mon method of marking sections for the large-scale ring composi-
tion. It almost makes key words superfluous, though repetitions of
names, phrases, or events are also used to signal the parallel for each
member of a pair of matching sections, and of course, the pairs that
are matched verbally are also matched thematically.
No less an authority than Virgil himself revealed alternation to be
his method for marking the division of the Aeneid into twelve vol-
umes.≤ Three stories of action are separated on each side by two
interludes (Table 7).
It has one exposition, a series of five volumes down one column,
volume 7 at the mid-turn, five more volumes up the next column in
inverse order paralleling the first, so that volume 12 meets volume 2.
The rows are also divided horizontally by another principle, alter-
nation of dark and light themes. To emphasize the resemblance be-
tween Virgil’s alternating system and that of the book of Numbers, we
can present it as in Figure 14, showing dark and light volumes in a
pattern.
116
Table 7. Virgil’s Aeneid, structured by alternating volumes.
I The Trojans arrive in Carthage, Juno raises a storm

II Destruction of Troy XII Rome’s future assured
III Interlude, wandering XI Interlude, truce
IV Tragedy of love X Tragedy of war
V Interlude, games IX Interlude, siege
VI Future revealed VIII Birth of Rome
VII Trojans arrive in Italy, Juno raises war
The excellence of the mid-turn of the Iliad is not enough by itself;

we also need to take into account the series of parallelisms that cut
across the poem’s length. This is also excellently done. Once arrived
at the account of the war Homer refines and tightens the structure by
using nights for markers as well as groups of days. After the first
numerical ring (the major ring according to Otterlo), the divisions
within the eight days of war are marked by a precise sequence of
sunrises and nightfalls. The attention that the poet pays to the first
signs of dawn in the Iliad is famous, and he describes every nightfall,
not so lyrically, but just as punctiliously. Various homely words, such
as ‘‘time for supper,’’ ‘‘feasting,’’ ‘‘sleep,’’ ‘‘darkness,’’ ‘‘going back to
the ships,’’ give the signal for recognizing nighttime. The descent of
night always followed by the arrival of ‘‘rosy-fingered dawn’’ gives
alternating structural markers throughout the period of the war. So
far the Iliad is exemplary.
But at this point my eulogy for the Iliad’s compliance with ring
conventions must come to a pause. Homer has marked out the units
of structure very clearly by alternating nights and days. We are en-
titled to expect thematic correspondences to slot into the places thus
marked. This does happen insofar as the day is for fighting, the night
for feasting, assemblies, and parleys. The days are bloodthirsty to a
horrific degree; the battle lines are always shifting forward and back.
Important things also happen at night, councils of war, spying, col-
lecting the dead, succouring the wounded, and making decisions.
Each day and each night moves the story along on the metonymic
117
up. I have pointed them out in the previous chapter. There is the
concordance between the exposition in book 1 and the mid-turn in
books 8 and 9. Also note the correspondences between Achilles’ un-
quenchable anger at the mid-turn and his conciliatory behavior at the
end, books 23 and 24. Also note the parallel between the ending in
book 24 and the beginning in book 1. Fourth, there is the clear
concordance between the two battles placed at days 4 and 5 to support
the mid-turn. Fifth, it is hardly forcing it to point out that day 1 and
day 8 have in common feasting and the only mention of laughter.
This leaves us to account for weak concordances between days 2
and 7, and days 3 and 6. Their presumptive parallels are weak, barely
recognizable. Day 2 and day 7 have in common only Hector’s declara-
tion that he will allow the corpse of any warrior he slays to be taken
home, and his expectation that his own corpse will receive the same
dignity (7.66–86), which makes an inverse match to Achilles’ promise
that Hector’s body will be eaten by dogs (23.185). On day 3 the Greeks
support the proposal to honor the dead, while on day 6 Hector repeats
his oath of day 2 that he will honor the body of any warrior he slays.
Ironically, he is killed, his body mutilated and carried back to Achilles’
camp for further deep dishonor. Apart from these instances, much else
happens in those four days that does not have any thematic parallels.
Admittedly I have scraped the barrel to find these matches. Days 1
and 8, 4 and 5 have already shown strong parallels, but all the other
matches are based on the theme of corpse defilement. It might look as
if they have been put in specifically as markers, except that they are
scattered rather randomly. There is no other attempt to make the
themes of the parallel days match each other. My best interpretation
is that the structure of days and nights in the minor ring is nearly as
formal as the structure of the groups of (1–9)–12–12–(1–9) days in
the major ring, this in the sense of a purely formal structure being
empty or independent of content. The poet’s skill in this ring com-
position is to have produced two sets of four days, the last (Achilles’
hospitality in honor of dead Patroclus) being brought to converge
with and match the return of the gods on the first day. He has suc-
cessfully made and closed a ring. That may be all there is to it.
119
Table 8. Big chart of the Iliad.
1.1–492, Achilles’ quarrel and anger, 24, Achilles twelve days’ mourning,
nine days’ plague. The gods’ twelve- desecrating Hector’s corpse (24.1–
day absence. 140). Trojans’ nine-day mourning
for Hector (24.141–804).
Day 1 (1.493–611). The gods Day 8 (23.226–895). The men

return. Achilles asks Thetis to return, wake Achilles, gather up
influence Zeus against the Greeks. bones, put them in an urn for burial.
She relates story of the quarrel, Funeral games, lavish prizes.
persuades Zeus to side with Troy. Laughter on earth when Aias falls in
Hera angry. Banqueting all day. the dung (23.784).
Hephaestus jokes about his fall.
Laughter in heaven.
Night 1 (1.605–2.47). Zeus Night 7 (23.192–225). Achilles
sleepless, summons Dream summons N and W Winds
(personified) to deliver a misleading (personified) to blow on the fire,
message to Agamemnon. which then burns well. Achilles
tends it all night, sleeps in the
morning.
Day 2 (2.48–7.293). Agamemnon Day 7 (23.100–91). The Greeks

proposes to abandon the campaign, collect firewood and stack the pyre
hopeless without Achilles. Odysseus for Patroclus. Aphrodite and Apollo
rallies troops. Army units marshaled, protect the body from the sun and
battle engaged. 3.85–460, truce for from dogs. In the evening the men
single combat, Paris and Menelaus. are dismissed, friends remain, they
Paris rescued by Aphrodite (3.380). light the pyre. Achilles sacrifices
Agamemnon demands return of oxen, sheep, and twelve noble sons
Helen and looted treasure. 4.105f., of Troy. He declares: Hector shall
Pandarus breaks truce. Battle not be eaten by fire, but by wild dogs
renewed. Bks. 5, 6, 7, Argives yield (23.185).
before Hector. 7.66–86, Hector will
honor corpses of those he slays,
demands honor for his own corpse.
7.67–282, Truce, single combat,
Aias and Hector.
Table 8. Continued
Night 2 (7.293–381). Feast of the Night 6 (22). Dead Patroclus
Greeks: agree to gather and burn appears to Achilles in a dream,
the dead, build rampart, and ditch asking for burial. The Greeks take
before the ships. Assembly of the their evening meal and rest.
Trojans (345). Idaeus dispatched to
offer treasure and propose truce for
burying the dead.
Day 3 (7.381–465). Idaeus delivers Day 6 (19). Hector imagines
the messages. Offer to return the offering Helen and the treasure to
treasure accepted, suggestion of the Greeks. Achilles receives arms,
return of Helen rejected. Support reconciled with Agamemnon
for honorable treatment of the dead. (19.56). 20, 21, fights the river;
Truce, both sides gather and burn harries Trojans. 22.66, Priam
their dead. Argives build rampart expects to be killed and eaten by
and ditch. Zeus and Poseidon dogs. Battle engaged, Achilles fights
comment negatively (445–64). Hector, who repeats promise to
honor body of the slain. 255–60.
Hector is killed, Greeks mutilate
and take his body, 330–404.
Night 3 (7.464–482). Both sides Night 5 (18). Trojans take their
feasting all night long, while Zeus evening meal. Achaeans mourn
made thunder. They took the gift Patroclus all night. Thetis
of sleep. commissions Hephaestus’s arms for
Achilles. Achilles will not bury
Patroclus until he brings Hector’s
head.
Day 4 (8.1–485). First day of great Day 5 (11.1–Bk. 18). Second day of
battle. Trojans press hard on great battle. Second Eagle portent.
Argives. At midday Zeus sets his Bk. 12, Argives driven back, Hector
golden scales for Greek defeat charges the ditch. Bk. 13, Third
(8.70). Hector drives his team eagle portent. Bk. 14, Trojans cross
irresistibly to the ships. Zeus sends the ditch, Argives in rout Bk. 15.
eagle, a portent taken by Argives Patroclus joins fight, is killed by
as sign of their victory (245–52). Hector. Bk. 16, Hector about to
Zeus foretells death of Patroclus burn the ships. Bk. 17 and to defile
and return of Achilles to the fray Patroclus’s corpse, Aias drives him
(477–75). off (17.126–27).
Table 8. Continued
MID-TURN
Night 4 (8.485–Bk. 9, Bk. 10). ‘‘Now in the western ocean the shining sun
dipped.’’ Night council in the ships; the story of the quarrel rehearsed,
Agamemnon admits his fault, decision to send envoys to request Achilles to
fight with them. Night Embassy led by Odysseus as mediator, who recites
the story of the quarrel with Agamemnon. Achilles replies with the story of
the quarrel, rejects offer of gifts and return of Briseis; embassy reports back
at camp. 10, two spying raids.
Here I must recall a precious lesson from Benjamin Harshav, pro-

fessor of Hebrew poetry at Yale. I called on him for advice on the
structure of the book of Numbers. I was worried about a passage that
did not seem to fit the pattern I had discovered. He told me to cherish
this piece of non-fit, love it, exhibit it, and never, never try to hide it.
It is prime evidence of my honest search, and also evidence that I have
not forced my own pattern upon the material. A smoothly perfect
design would be suspect; a few mismatches in an ancient text that has
survived so many generations would be much more convincing to
discriminating readers.
Having laid this unforgettable advice to heart, I have suggestions
for two problems about the structure of nights and days in the Iliad.
The failure to find parallel themes filling all the formal parallels of
days and nights may be turned to good account. The Homeric bards’
performances would have been a creative interaction between the
narrative traditions and their own poetic genius. There would have
had to be room for including new developments. The strong frame-
work of days and nights provides for just such available spaces. This
would account for asymmetries on the two sides of the mid-turn.
Furthermore, it would be no trouble to add an extra day and night,
so long as another extra day and night are added at the same time
to the other side of the mid-turn and some artificial parallel inserted
into each.
There is a small problem about the number of days. I have assumed
throughout that the period of war in the Iliad lasts eight days and
122
seven nights. A well-known anomaly occurs in book 7 at day 3. The

start of the day is announced at line 381, when the Trojan emissary,
Idaios, rises at dawn to deliver a solemn message to the Greeks about
a day’s truce. When he has done this, he makes his way back to the
Trojans, line 412, who immediately start to make preparations for
retrieving and burning their dead. Then at lines 421–23, while both
sides are engaged in this sad work, ‘‘the sun of a new day struck on the
plough lands . . . rising out of the quiet water and the deep streams of
the ocean to climb the sky.’’ We get the definite impression that the
sun has properly risen while they are working away. Is this another
new day, or the same day?≥ Reasonably enough the commentators
count it as the same day—that is, the same day 3 in my reckoning of
days. But then at line 433 another new day is announced.
There has been no announcement of a night, but ‘‘when the dawn
was not yet, but still the pallor of night’s edge, a chosen body of the
Achaians formed by the pyre’’ and built on it a defensive fort. Watch-
ing this building, the god Poseidon has a conversation with Zeus.
This seems to be early morning still, but ‘‘as these two were talking
thus together, the sun went down and the work of the Achaians was
finished’’ (lines 464–65). This is presumably to be counted as the
beginning of night 3. We seem to have had two dawns following each
other with only one night. Everywhere else, in the whole book, every
time a new day is announced, the following night is announced in due
course. The system is perfectly consistent, except in this one place.
Either it is all happening on the one day, between the first dawn and
the sunset, or there were two dawns and no night in between. Perhaps
it is ambiguous because the day’s truce was not supposed to be spent
on defense works. If that is the point, the Greeks are cheating on the
truce as the Trojans did on day 2 (4.105 ff.). Or perhaps it is just a con-
fusion due to poems’ ancient and vulnerable process of transmission.
These mentions of dawn and day suggest that a night is missing
between the inauguration of day 3 and the arrival of night 3, which is
unambiguously announced by the words ‘‘And then the sun went down’’
(7.464). If there is a missing night and if we insert it, day 3 is split into
two days, and night 3 becomes night 4, and so on, with the result that
123
the eight-day period is now nine days. Adding a day would dislocate
the pairing of parallel days across the middle of the ring, so I grate-
fully choose to follow the practice of ignoring the anomaly.
We have noted that Whitman makes of the whole numerical scheme
a regular ABCBA structure with eight days of war in the middle: (1–
9)–(1–12)–(8)–(12–1)–(9–1).∂ His numerical scheme is a regular
ABCBA structure: (1–9)–(1–12)–(8)–(12–1)–(9–1).∑ This being a
purely formal structure, it posits nothing about the contents. But now
it strikes me that the (8) at the midpoint of his numerical ring is a bit of
a disappointment. A pattern of 1’s, 9’s, and 12’s might have been more
satisfactory. It may be wild to suggest that there may have been an
earlier version in which day 3 was split into two days. And another day
added on the other side of the mid-turn. It is just a speculation inspired
by the only anomaly in the count of days and in Whitman’s model.
We have now examined the two sequences of chapters, one that
runs from day 1 to night 4 and the other one that parallels it, going
from day 5 to day 8. Unexpectedly, the same analysis that demon-
strated that Tristram Shandy is almost, but not quite, in ring form and
that unraveled the alleged confusion of the book of Numbers reveals
that Homer and the editor of Numbers have used the same pattern of
composition. The main difference is that the Iliad, adhering to the
same rules, is more complex and more coherent. The Iliad’s ending,
mid-turn, and exposition match each other better. Though the pat-
tern of alternating days and nights is not as unambiguous as the
pattern of alternating narrative and law in Numbers, the poem does
conform to the principal conditions for making a ring.
With all this held in mind, if I were to ask myself honestly whether
I think the Iliad is or is not a well-formed ring, my own answer is
positive. Yes. It is an excellent example of ring structure, but over the
centuries there have been a few changes. In a work like this, so intri-
cately organized, even a small change can disrupt the system of days,
and subsequent scribes then have a hard task to set it right again. This
is what seems to have happened here around the middle of the action.
However, the crucial requirements are for parallels between the mid-
dle, the beginning, and the end. These the Iliad fully meets.
124
ten
THE ENDING
H O W T O C O M P L E T E A R I N G
Arrived at this chapter we face several questions about endings.

Our first will concern the way that actual rings come to their ending.
This includes what the ending does for the composition, how it is
done. That will be easy. More difficult is solving Jakobson’s conun-
drum described in Chapter 1, which must be attempted if my ending
is to meet my beginning.
The great philologist considered that parallelism is a faculty inher-
ent in the relation among language, grammar, and the brain. Once
the ring structure has been explained as a system of parallelisms, the
puzzle takes a new turn. Why is it so difficult to recognize? It hardly
seems plausible that the scholars balk at the circular system just be-
cause they are used to linear reading and writing. Roman Jakobson’s
theory that precisely this kind of analogic arrangement is deep wired
into the human capacity for language becomes a paradox. Surely we
should all be capable of appreciating a structure of parallelisms at
first sight?
There are several unconvincing explanations. One is simple preju-
dice against foreigners: perhaps we read archaic literature with a con-
cealed ‘‘Orientalist’’ contempt. Are we like the ancient Greeks in
regard to anyone they called ‘‘barbarian’’? Admittedly, the record
does not suggest that we do expect ‘‘Oriental’’ writings, or oral litera-
ture, to be elegant or capable of effective expression, but nor do we
seem to be specially prejudiced against them. A slightly more plau-
sible explanation might be that it is a side effect of social Darwinism
with its emphasis on moral and intellectual evolution. From this,
combined with our colonial tradition, we learned that the so-called
primitive peoples over whom we ruled were still at a childish stage of
evolution, morally and intellectually, so we were never looking for
literary sophistication.
125
THE ENDING
These bequeathed ideas may have limited our capacity for appre-
ciating other people’s artistic refinement. Once it has been discarded
as a method of literary construction, that is to say when there are no
contemporary ring compositions being produced, it is understand-
able that it should be overlooked by the readers. When it goes out of
fashion, it comes to an unlamented, unrecorded, end.
The writer who believes that a recognizable ending is necessary
will have to decide when and how to end, whether to make it a tri-
umphal justification of the original position, or a terrible warning, or
whether a modest summary will serve. For the author composing a
polished ring there is little choice; the ending is prescribed. The rules,
which I have elicited by reading endings of rings, are three:
1. The ending must evoke the beginning; it should close the ring by
touching on the topics that were opened by either the exposition,
or the section following it, or both, using some of the same words.
2. To have followed the first rule automatically evokes the mid-turn,
which has itself been designed to connect with the ending as well
as with the beginning.
3. It may make a double closure, using the option of the ‘‘latch.’’ In
this case, the first ending will finish the immediate business, con-
clude the story, or round off the laws. The second ending will set
the text as a whole in a larger context, less parochial, more human-
ist, or even metaphysical.
I pause here to ask why a poet should have the ambition to make a
polished ring. Would not something rough and ready do? In my first
chapter I described the context of recital and the function that a well-
turned poem performs by way of authenticating the status of the bard.
Professional performance always has an element of competition.∞ It is
not just a matter of payment, or just a matter of gaining fame; there is
the practical problem of the box office. A bard needs to draw an
audience; the need puts him in tacit competition with the others. If he
and his team can count on a large audience at the major feasts, they
are winning. As he climbs the ladder of success he will encounter
more and more discriminating audiences. In the triadic interaction
126
THE ENDING
between poet, audience, and poem the conventions will be contin-

ually sharpened. The pressure for quality encourages him to abide by
the rules, refine them, and make the composition more complex.
Glenn Most considers that its complexity explains the difficulties
that modern readers have in interpreting Pindar’s poetry. He holds
that ‘‘each poem of Pindar is organized systematically as an immanent
compositional unity whose form can be interpreted as appropriate for
its unique content. . . . ‘Compositional form’ depends on spatio-
temporal concreteness of the art-work . . . the arrangement of the
parts is neither capricious nor arbitrary—none of its parts can be
substituted or exchanged against some part of another text . . . with-
out the substitution being noticeable. The sequence is not random.’’≤
He concludes that the unity of the poem derives from ‘‘the formal
design whereby all the elements are integrated into a coherent total-
ity, and the formal design itself is part of Pindar’s meaning.’’ The idea
of the poet’s formal design controlling the parts and conferring unity
on the poem corresponds closely to the concept of repleteness in a
work of art. The strength of this concept is that it comes so close to
the way modern artists themselves talk about painting. Matisse, for
example, had an idea of this kind in the forefront of his mind all the
time. He was driven by the desire to realize his original vision of a
painting as closely as possible. His advice to painters was: ‘‘When
painting, first look long and well at your model or subject, and decide
on your general colour scheme. This must prevail. In painting a land-
scape you choose it for certain beauties—spots of colour, suggestions
of composition. Close your eyes and visualize the picture; then go to
work, always keeping these characteristics the important features of
your picture. And you must at once indicate all that you would have in
the complete work. All must be considered in interrelation during the
process—nothing can be added.’’≥
Matisse repeated this advice, to strive for fidelity to an original
vision, to lose nothing of it, and to add nothing, always with emphasis
on the interrelation of the elements. On the process of making a
painting he said, ‘‘I want to reach that state of condensation of sensa-
tions which makes a painting.’’
127
THE ENDING
He accepted changes that take place as the picture is being re-

worked, but the direction of this process of constant change is always
back to the beginning: ‘‘For me all is in the conception. I must there-
fore have a clear vision of the whole from the beginning.’’∂
What he wanted to achieve is nothing more and nothing less than
that original vision, and everything condensed around that point.
Roger Fry is another modern painter who looked for a similar effect,
what he called ‘‘that unity of feeling.’’∑
Their shared idea of the finished painting corresponds closely to
the philosopher’s concept of repleteness. Nelson Goodman’s episte-
mology of rightness is helpful here.∏ In Languages of Art he directs our
attention to ‘‘fittingness’’ and to the schemes of thought in which an
element may be seen as right, and particularly to connoisseurship.π
Because the latter, connoisseurship, is a key element in the situation
of bardic performance and reception, a digression is in order.
Goodman takes judgments of beauty and value out of the range of
misty universals and subjective sensibilities. His focus on ‘‘rightness
of categories’’ extends Wittgenstein’s concern with rule-following
and with rightness as central aesthetic concepts. Both men take
aesthetic judgment to be a matter of connoisseurship, knowing the
scheme of ideas, knowing the well-made garment: ‘‘What does a per-
son who knows a good suit say when trying on a suit at the tai-
lor’s? ‘That’s the right length,’ ‘That’s too short,’ ‘That’s too narrow.’
Words of approval play no role, although he will look pleased when
the coat suits him.’’∫
On these lines, the preliminary work of the art historian is to find
the right categories for placing particular works. Do the members
assigned to the category fit it? It may come up as a matter of authen-
tication (‘‘Is this a genuine Rembrandt?’’), or as a matter of correct
classification (‘‘Does this painting belong with the Mannerists?’’). Or
more profoundly, the art historians may be asking each other, ‘‘Have
we got the right categories, knowing what is being attempted? Recall
Wittgenstein’s homely example of categories for our enquiry?’’
For answering such questions, Goodman provides a list of ‘‘symp-
toms of the aesthetic’’ that include ‘‘repleteness.’’Ω He uses the term
128
THE ENDING
replete to mark out distinctive criteria for works of art, without im-
plying that they are good or bad art, but distinguishing them from
maps, charts, and diagrams.∞≠ It relates to internal cross-referencing.
Clearly, a ring composition has a facility for repleteness.
A poem or painting is replete when all the elements fit and contrib-
ute to the whole design. It is ‘‘fittingness’’ applied to the internal
relation between the features of the art object. Do they support each
other, or do they jostle, jar, and intrude? The same for literature: Do
the elements of this story fit together? Has the argument wandered
from the original idea?
The term replete has nothing to do with the content of the art
object. In this context there is no such thing as repleteness in general,
and no way of suggesting that the perfection of a work of art depends
on some general completeness of its cross-referencing. We cannot
call up universal values in order to say that the less the work is paro-
chial or narrow the more it is replete. Nor can we say that unless the
work of art is universal in its reference it cannot be said to be replete.
Repleteness makes no call for the work to be humanitarian, or to
respect the heritage of the past, or to be sensitive to the pains of
childbirth or to dangers to the environment.∞∞ The term is technical.
All that benevolent concern is dissipated when we focus on an original
project and its fulfillment.
The book of Numbers provided an example of a ring with a defec-
tive ending (Chapter 5 above). Instead of pausing to recall the middle,
the ending of the book goes straight home to the start. Neither of its
two endings refers to the midpoint; the effect is disjointed in spite of
elaborate care to make the other required cross-references.
The first and main ending makes an excellent connection with the
beginning but somehow fails to connect up with the mid-turn. The
failure is all the more noticeable in that the ending chapters are exem-
plary in ticking off, as it were, all the points made at the start when
God commanded a census of the people of Israel (chs. 1–4). At the
mid-turn the Levites had rebelled and been punished. When they
turn up for mention in the first ending (ch. 35), instead of rebuke they
receive forty-eight cities to live in, as if the chapter followed straight
129
THE ENDING
on from the exposition. There they have been told they cannot have a
tribal territory, so some arrangement has to be made for them to have
somewhere to live. Nothing is said about their revolt against Moses.
Perhaps they have been forgiven, so perhaps no explanation is neces-
sary, but there does seem to be a gap.
The second ending is the latch (ch. 36), where we would look in
vain for some mention of the Levites. It follows on the first ending,
but makes a quick switch of subject. The whole style of this book is
almost everywhere rigidly consistent with ring practice. It uses, with
textbook correctness, alternation and chiastic forms for indicating
structural units. The exposition has placed the Levites in a servile
position in the Temple, the mid-turn passionately records the Le-
vites’ perfidy. After such meticulous allegiance to the conventions, we
expect the endings to follow the time-worn principles. We are en-
titled to expect some criticism of the Levites in the ending, but there
is nothing. Clearly the editors were fully capable, and probably will-
ing, so I assume the plan laid down in the exposition was interrupted.
The text we have before us in chapter 35 has lost some and gained
some additional material that forms the latch.∞≤
Is it due to inadvertence? Why would the careful editors grow
tired? Anyway, there is no other sign that they were tired. In the
census of twelve tribes the exposition has given three places for the
Josephites they are wishing to protect: one for Benjamin and one for
each of the two sons of Joseph (three heirs, instead of the one heir
each given to the other patriarchs in these lists). The Josephites’ claim
to be counted with the children of Israel has no other precedent in the
book. We are not ever told that anyone has challenged it. We are not
told why they need high-level support from the mouth of Moses
himself. This latch is discomfiting. If it does confer extra cohesion, it
applies at an unarticulated level of political loyalties. So correct in
every other way, at the end the book of Numbers turns out to be a
flawed ring.
We have seen the perfect exemplar, the Iliad, whose middle and
ending both depend directly on the beginning. If we want modern
examples written without an expressly recognized tradition, we can
130
THE ENDING
easily find authors who deliberately intend to take their work along
the homeward route. Laurence Sterne, just after the middle of Tris-
tram Shandy, lets fall that he knows he is supposed to go back to the
beginning: ‘‘all which being considered, and that you see ’tis morally
impracticable for me to wind this round to where I set out’’ (6.33).
His carefully structured book justifies my enterprise by being mod-
ern, being greatly appreciated, being famous for lack of structure, and
yet by his admitting that he expects to go back to his beginning, and
doing so by a series of inversely ordered parallelisms.
Ring composition belonged to the high style, an elegant form of
writing for reciting at important events. As such it called for a proper
ending. For other elegant literary forms, even more constrained by
stylistic demands, such as pattern poems, endings were achieved in
different ways.∞≥ For example, a text could be projected on a physical
model. A pattern poem cuts the lines of a poem shorter and longer so
as to depict on the page a figure of the subject of the work: the lines on
the page present an outline, the poem and the image converge. This
convention gives no scope for varied endings. When the lines of
poetry have completed the visual shape on the page, the poet must
stop or the picture will be spoiled. There is no need to herald an
ending, no decision to be taken; the ending was implicated in the
initial design.
The fashion in Europe for figure or pattern poems, starting from
the Hellenistic tradition, was strong in the Middle Ages and even
more popular in the late Renaissance. ‘‘Virtually every seventeenth-
century poet attempted one or two. The design on the printed page is
a pattern of the theme. Poets seem to have taken a special pleasure in
the visual aesthetics of print.’’∞∂ Sometimes it seems to have been a
kind of charming, clever decoration, more like a game or a greeting
card. Alastair Fowler also shows that the pattern poem could be more
than a gimmick; the poet found contemplative depth in reflecting on
the analogy between the visual and verbal presentation (Fig. 15).∞∑
In modeling the structure of a book upon the structure of a physical
object, the book of Leviticus goes several steps further. This book is a
projection of the tabernacle. God dictated the proportions of the
131

Fig. 7. The scroll. Drawn by Pat Novy.
I have abstracted the following seven rules or conventions from

long ring compositions. They are not rules in the sense of there being
something hard and fast about them. Breach carries no penalties, but
insofar as they are commonly observed they are like rules. They are
responses to the technical problems of coming back gracefully to the
start. Other technical problems arise out of the solution adopted for
circumventing the first.
35
Fig. 15. The Goblet, from Bernhard Praetorius, ‘‘Alcimendontis Poculum’’,

Press mark: HAB 49 Poet (3). Courtesy Herzog August Bibliothek
Wolfenbüttel: 49 Poet. (8).
THE ENDING
desert tabernacle to Moses in the book of Exodus (ch. 25). The build-
ing consists of three compartments separated by two screens: the first,
very large, the entrance and the court where the worshippers make
sacrifice; the next, smaller, the sanctuary where only the priests may
enter. It contains the table for the showbread, the altar of incense, and
the menorah, the seven-branched candelabra. Lastly, the smallest, the
Holy of Holies, contains the Mercy seat and the Ark of the Covenant,
a figure of a cherubim on each side. Nobody can enter it except the
high priest.
The book is likewise organized in three sections of diminishing
size. It consists of laws, separated by two narratives, which I take to
correspond to the two screens. The sections of the book preserve the
relative proportions of the sections of the tabernacle. The first large
section of the book corresponds to the large court of sacrifice, and the
book’s contents in this section actually state the laws for sacrifice. The
second section of the book is smaller; it ordains the liturgical work of
the priests through the year and prescribes rules for their marriages
and households. In this respect it corresponds faithfully to the holy
place reserved for priests, and it describes what has to be done with
the incense, oil, and bread whose furnishings are in that compart-
ment. The third part of the book is very small indeed, like the Holy of
Holies, only three chapters long: it is about the covenant that is
supposed to be kept there. So the book has been carefully projected
upon the architecture of the tabernacle and on the proper activities of
the place.
When the book comes to the pages that correspond to the end of
the building it is modeled upon, it has automatically come to an
end. To go on would spoil the design.∞∏ The analogy between the
abstract structure of the written contents and the solid object on
whose shape it has been projected gives the book a strange trans-
parency. The reader looks through the words, or past them, and,
visualizing the object, can intuit the depths of the analogy. At first
Leviticus looked like a dry list of laws, but now, seeing it in three
dimensions, it exemplifies the House of God. That does change the
way it is read. And moreover, the tabernacle where God dwells among
133
THE ENDING
his people exemplifies Mount Sinai, where God originally met his
people and gave his laws to Moses. Tabernacle, holy book, and holy
mountain, presented so compactly, yet so vast in reference, mirroring
each other in two and three dimensions, they stand for everything
that is covered by God’s law.
The pattern poem is a way for a text to say more than it says. This
solves a theological difficulty. Frank Kermode, on the idea of the
classic, reflects on ‘‘the difficulties we encounter when we ask what
happens when modern minds engage ancient texts. From day to day
we must cope with the paradox that the classic changes, yet retains its
identity. It would not be read, and so would not be a classic, if we
could not in some way believe it to be capable of saying more than its
author meant; even, if necessary, that to say more than he meant was
what he meant to do.’’∞π
The poem is undeniably a way of saying more than the words say,
and even of saying more than the author could possibly have meant.
In the case of Leviticus the hidden analogy has expanded the meaning
to encompass the Lord’s ordering of his infinite universe.
Seeing the Leviticus text as a projection of the tabernacle is a reve-
lation of the same order as produced by reading a ring according to its
structure. The impact of a composition would obviously be much
enriched by having a meta-structure. If the verbal structure is being
projected on to something else outside itself, it is making another
analogy at a meta-poetic level. And this projection provides a further
kind of ending or completion.
The previous pages have shown that the task of coming back to the
starting point (the defining feature of ring composition) accounts for
its complexities and its frailty. Asking why the author burdens the task
of writing with the problems of return, we can suggest that writers
and readers share a home-seeking urge. The effect of the ending on
the overall coherence of a ring composition is crucial. If the ending
does not perform all its functions, neatly connecting up the various
themes that have gone before and picking out the key points like a
street lamp, readers are excused for losing their way.
There is nothing final about homecoming. Any recognized kind of
134
THE ENDING
ending is loaded with intimations of other endings. After a rest, the

home-comer will set out again, and return again. Nonetheless, with
the sight of the place and the familiar smells and sounds, everything
announces ‘‘home.’’ No matter how many more expeditions there will
be, each successive return to the start is an ending. A series of minor
endings is characteristic of large works, like chapters in a book. By
analogy, there is no reason to expect that each literary ending is final;
an ending says only that some part of the work in hand is complete.
More volumes may follow, like the four of the Pentateuch that follow
Genesis. Their sequencing may have been anticipated, as in an alpha-
betic ordering.
Endings are not all happy; there are the griefs of failure, defeat, and
death. The analogies press upon each other. No wonder that many
agents who are free to choose, like writers and painters, will try to
postpone ending. Whether they succeed depends on how much au-
tonomy they enjoy. Social convention has to put out a heavy hand to
oblige artists and players to stop.
The idea of reciprocity that governs all kinds of transactions under-
pins an idea of ending. An unrequited debt is waiting to be settled;
requital makes an ending. Justice and fairness, punishment and re-
venge, and the great series of tit-for-tat exchanges, they all depend on
construction of substitutability, they expect correct requital, neither
too much nor too little, and at a stipulated time. At micro-levels the
measuring of equivalence for matching gift and counter-gift may be
free, but for important transactions the community eventually comes
to help its members by imposing agreed techniques for calculating
requital.
Time limits for settlement may be significant, if coordination is
important. Gate money and box-office institutions require time lim-
its. Some games need both a whistle to signal the beginning and a
fixed time for play. Some firmly fix the ending into the structure of the
game, as in some card games: when all the cards have been drawn, or
when the players have all had their turn, the game must stop.
Without some external pressure, most writers have difficulty in
knowing when to stop revising. Painters and sculptors are reluctant to
135
THE ENDING
lay down whatever they are using, knife, brush, or chisel. Sometimes
the inspiration for the next work may push the current one into the
background of concern, and so it ends at the point at which the artist
is bored. Competition between artists can make it harder for one to
relinquish the work if there is still a chance to do better than the
rivals. Musical composers, painters, writers may only stop when they
feel like it if they have the sole responsibility for structuring their own
piece. As soon as their work involves collaboration, however, as a
composer with an orchestra, a journalist with a newspaper, a painter
with gallery officials, the need for coordination makes endings a mat-
ter of negotiation and convention. It means that the kind of pressure
on ending according to conventions results from the kind of society
that is giving out the rulings.
Social motives (like competition) may tempt an artist to postpone
the ending for a long time. J. M. W. Turner used to send his pictures
in to the Academy in a very rough state, expecting to finish them in
the three days allowed for ‘‘varnishing.’’ He was not the only painter
to go on finishing his picture while it was hanging on the wall of the
exhibition, but he made a big performance of it. Lawrence Gowing
wrote of Turner: ‘‘One of the motives of the performances on the
varnishing days was certainly to demonstrate the force that the in-
trinsic color of painting possessed in his hands, overwhelming every
other artist. They were the final manifestations of power to out-rival
everyone, past and present. His contemporaries retaliated; soon the
galleries were full of painters tuning up their pictures in competition
and the rooms reverberated with color.’’∞∫
Different literary kinds require different kinds of endings. Ob-
viously, any old conclusion cannot be foisted on any composition. It
has to have been anticipated to some extent by the process of con-
struction. The result of a mathematical accounting process can be
indubitably correct; there is no room for interpretation. For a work of
fiction interpretation is free. Its ending may have been prepared by
building up some sort of crescendo, or by some steady decline herald-
ing dissolution. Or, if neither of these is appropriate, the acceptable
ending may review what has gone before. Fashion always exerts a
136
THE ENDING
constraining grip; there is always less autonomy and scarcer option

for eccentric endings than we imagine.
These thoughts suggest further that ‘‘homing’’ is another of our
fundamental mental resources, like making analogies and parallel-
isms. It is worth reflecting more on the gratification of going home,
and the balance between the joy of exploring and the joy of returning.
The traveler sets out with a destination in mind, reaches it, turns
around, and travels back the same route to the beginning. Even very
small animals, or insects, can do it. A mouse sallies forth from its
retreat to get food, gets some, then turns around, comes back the
same way, and arrives. Through our animal origins we are gifted with
navigational skills, which we recognize when we watch a homing
pigeon or track wasps to their nest.∞Ω Some deep-laid faculty has nur-
tured this process in human beings. We know what homing means;
we do it all the time, we can recognize a return to the beginning when
we see it, and we can transpose it to a literary form. It would be
satisfying for this argument if we were able to say that,
. . . the end of all our exploring

Will be to arrive where we started
And know the place for the first time.
Through the unknown, remembered gate
When the last of earth left to discover
Is that which was the beginning. (T. S. Eliot, Little Gidding, V)≤≠
No wonder the ending is so important for a long ring composition:

it can link together everything in the text, creating coherence for the
whole. When the ending of a composition wraps up everything that is
there, the whole is endowed with a special power. The same for a
painting or a carving, or music—its dense coherence inspires feelings
of respect, even of awe. It is something to be valued for itself, like
poetry, regardless of what narrative or instrumental functions it may
perform. You might use the picture for a blanket, the carving to stop a
leak, the poem to advertise a brand drink.≤∞ Its possible usefulness is
irrelevant to its quality as a work of art. The thing is too impressive to
be cherished for generations because of any incidental ability it may
137
THE ENDING
have to refer to something outside itself. It is not made for a practical

purpose, like a map. It is not necessarily made for a limited occasion,
such as a graveside elegy, a marriage blessing, a speech for a school
prize-giving. It is made to be itself, a complete thing in its own right.
This explains why the ring has a formal ending, but it does not say
why it has died out in so many places. Nor does it solve Jakobson’s
conundrum. Another chapter on endings will be necessary to do jus-
tice to both thoughts.
138
eleven
THE LATCH
J A K O B S O N ’ S C O N U N D R U M
After all this, Jakobson’s paradox is unchanged. We believe what he

says, that writing in parallels comes to everyone naturally—but we do
not understand why we are slow to recognize it. Recall that I wish to
understand both why ring composition fell out of fashion in the east
Mediterranean hinterland in the fourth and fifth centuries, and also
why now we have trouble recognizing rings.
I suspect that I have not been explicit enough about this latter
difficulty. Let me take advantage of the latch to give at the very end
one more case of a typical misreading of a ring. Erich Auerbach’s
interpretation of Bishop Gregory’s Histories of the Franks illustrates
how easy it is to mistake the genre and the motives of the writer if the
central loading is not recognized.∞ Saint Gregory was the Bishop of
Tours in the sixth century, and he was highly esteemed in his time.
Auerbach saw a big contrast between the bishop’s degenerate, deca-
dent form of writing and the classical Latin of the previous period
when Roman splendor was at its height. He recognized Gregory as a
famous, influential writer in his period; he praised his style as vivid,
psychologically superb, dramatic, and emotional. But he deplored its
being repetitive and over-detailed. He especially drew attention to
the syntactical control by which the Roman prose of the Golden Age
made connections that ‘‘reach the height of subtlety, exactness and
diversity—an observation which applies not only to conjunctions and
other devices of subordination, but also to the use of tenses, word
order, antithesis and numerous other rhetorical devices, which are
likewise made to serve the same end of exact, subtle, yet pliable and
richly shaded disposition. . . . Gregory’s language, on the other hand,
is but imperfectly equipped to organize facts; as soon as a complex of
events ceases to be very simple, he is no longer able to present it as a
coherent whole. His language organizes badly, or not at all.’’≤
139
THE LATCH
The point that concerns me here is the charge that Gregory is just
not able to present his story coherently, implying intellectual weak-
ness. Auerbach compares Gregory’s style in which, for lack of a con-
trolling syntax, words and emotions come gushing out, with classical
Latin, where ‘‘the whole is strictly controlled by the order imposed by
impersonal syntax, the sense is infinitely richer and better articu-
lated.’’ Though he praises Gregory’s style for its immediacy and sen-
sory directness, he finds that the text gives an ‘‘impression of dis-
order,’’ it is ‘‘confused and imprecise,’’ there is ‘‘a general lack of
orderly arrangement in the grammatical structure.’’ He concludes
‘‘that Gregory is not capable of arranging the occurrences them-
selves in an orderly fashion.’’≥ Auerbach pours scorn on the boorish,
country-bumpkin style. He may praise it for being concrete, sensory,
spontaneous, but he dismisses it as utterly disorganized and weak
in syntax.
This ought to present Auerbach with a puzzle, since he fully recog-
nizes that the good breeding and education of the bishop implied that
he would have been capable of writing in the classical style. He also
notes that Gregory himself had said emphatically that he knew his
style seemed boorish, but that he had deliberately adopted it so that
his Histories of the Franks would be more accessible to his congrega-
tion. The story Auerbach has selected to exemplify Gregory’s writing
turns out to be a ring composition, which imposes its own syntactical
rules and frame of reference. Gregory believed that this style would
be more widely understood than the rigid grammatical formulae of
the classical Latin writers. Choosing to write like this was for him
what Martin Heinzelmann calls ‘‘pastoral strategy.’’∂
The story is about a blood feud between two men. By the end of the
story the words and actions of the beginning are repeated in inverse
order. In the first scene, a servant is struck with a sword and falls down
dead. Sicharius, the central character, tries to avenge him; fighting
ensues, the would-be avenger (Sicharius) escapes. In the last scene it is
Sicharius who is struck with a sword and falls down dead, and the
servant who escapes. Auerbach gibes at the word-for-word repetition.
He does not notice that the first statement, about being struck with a
140
THE LATCH
sword and falling down dead, and its inverted form at the end, make
the inclusio for the whole story.
As it is a tit-for-tat tale of revenge told in clear parallelisms we
should look for the middle to see what it is all about. The central place
is not difficult to find. The story comprises three episodes, each about
a crime and the court’s verdict on it. The middle crime with the court
that tried it is at the central place between the other two; all three are
presented as parallelisms. The three episodes trace the development
of the feud, with violations of the peace on an escalating scale, the
crimes more horrendous each time. In each episode the status of the
court also escalates, so first, the people’s court, second, the judges’
and the bishop’s court, and last, in highest degree of judicial rank, the
city court.
The first two courts require wergild to be paid, but compensation is
refused each time by the offended party, who intends to wreak his own
revenge. The third trial results in a reconciliation of the two enemy
avengers. But after all that carnage, justice was not seen to have been
done, the two avengers have both gotten off lightly. The last big
scandal is that both refused to accept compensation; both persisted in
murderous vengeance till the end. But the story goes on. The two
enemies make friends, then they quarrel, and one, suddenly enraged
by a tasteless joke, does kill his friend in a final revenge. This is the
last scene, which repeats the first in reverse: Sicharius is killed and the
servant escapes.
Auerbach was misled by the title of the book; the stories were not
meant for a history for historians. The bishop’s court is central. The
whole object and meaning of the story is there. Bishop Gregory’s long
sermon on Christian peace and his fervent denunciation of feud and
vengeance are in the middle. We must conclude that the text is a
message for the faithful about the teaching of the Church. That is
why Gregory has used the simple language of everyday and dispensed
with the strict and rigid grammatical structures of the high style—
precisely as he had said. The little puzzle about why someone so
well educated as Gregory should compose in such a simple manner
disappears.
141
THE LATCH
Auerbach’s reproaches against inelegant Latin and ‘‘grammatical

monsters’’ do not concern us so much as his claim that Gregory’s text
is unintelligible and his thinking muddled. He says that the history
has left out what was necessary and included much unnecessary detail;
the reader is left in the dark. Gregory lacks historical judgment: he
has chosen to write about minor episodes, so trivial that no classical
historian would have bothered to mention them at all. Auerbach evi-
dently considers not only the syntax to be weak but the author to be
weak in the head.
Qualified scholars have turned the edge of his censure against the
critic, in defense of Gregory.∑ Auerbach’s scoffing remarks about
Gregory’s inability to see his work as a whole and his sneers about
incoherence, disorder, and mental confusion are misguided. You can-
not mistake the bishop’s sermon on Christian teaching for the mean-
derings of a feeble-minded historian manqué, not when you know
how the framework has been constructed and you know what is in the
middle.
At this point, we should try to take stock of ourselves and our own
cultural environment. Gregory’s histories lost out in comparison with
classical Latin writings, which belonged to such a different civiliza-
tion. It is possible that now that ring compositions have started to
reemerge from a long period of neglect they also find themselves in
the wrong kind of civilization.
Part of the answer to why ring composition has died out will be that
nowadays its very repleteness has a downside. To some it may seem
too highly contrived. Take away one word, change the pronunciation
of another, the fittingness is compromised. It becomes difficult to
know what was meant. It calls for a high degree of connoisseurship in
the composers and readers, but connoisseurs easily disappear. With-
out their steady social support the balanced cross-references and deli-
cate symmetries will be a liability. The more replete it is, the more
precarious the constituent correspondences will be in their reliance
on each other. According to discussions of postmodernism, we are
now living in a culture that resists boundaries and shuns formal end-
ings. In literature, painting, and music, there is a preference for open-
142
THE LATCH
ness, part of a more general bias against formality and structure. If

that is true, it may account for some of the current difficulties in
reading these firmly bounded, well-structured writings.
If we are indeed in a period that prefers open-ended solutions,
then we ourselves might risk going a step further in open-endedness.
We could try questioning the popular idea that postmodernism is a
unique state, experienced only by the generations that have reacted
against the certainties and confidence of the Renaissance. If we are in
such a period we can also be open to the thought that we are not
unique, and that the explanation for the disappearance of ring com-
position in various regions was the result each time of recurring states
of mind very similar to our own. We can try the idea that postmodern
uncertainty and skepticism are examples of a cultural undercurrent
that surfaces when the pillars of a local modernism are shaken and the
old system is coming to grief.
When an antique piece of furniture, ornament, or style of painting
that has once been valued, then has been disvalued, and then resusci-
tated in general esteem, Rubbish Theory applies.∏ The expiry or sur-
vival of objects does not depend on any natural robustness. Old lace
and old ceramics in themselves are very perishable. Like everything
else once valued, they could eventually fall out of fashion and be
chucked on the midden. A change of fashion may change their status
into delicate antiques, but it only works to save a few of the relegated
objects. A vast majority are destroyed as rubbish. Things can endure
only if they are so highly valued that resources are dedicated to their
preservation. The natural perishability of these ancient relics ac-
counts for why only a few examples remain.
A literary genre that survives through many centuries, even millen-
nia, is not a specially tough or resilient medium; it survives because it
has been seen to be worth protecting. Its survival is like the survival of
old lace, wrapped in protective tissues and cherished by its owners (or
museum curators). The initial question was why the merits of ring
composition are still disregarded even now that they are resurfacing
after aeons of oblivion. Why is no one noticing the elegant corre-
spondences and witty puns, or enjoying the titillating analogies that
143
THE LATCH
resonate across the structure? The same Rubbish Theory states that
once an object has become classified as rubbish it becomes invisible.
No one notices it, no one even sees it. Only after many decades a few
examples may be fished out of the darkness and slime, cleaned up, and
placed in glass cases. Then they are formally transformed into dur-
able objects and acknowledged as genuine antiques. Then anyone will
at once perceive their value, and the archaeologists will start digging
for more of them.
It follows that survival depends on contemporary esteem. Any
genre is at risk from social change. Not only genres—de Saussure, the
great philologist, was deeply pessimistic about discourses, or words,
or even letters, retaining stable meanings over any short period, or
even between performances. I mean pessimistic in the sense of ex-
treme doubt, simple skepticism, not in any sense of sadness, rather in
the excitement of an important discovery about truth.π Indeed, it is
an important topic. It is hard for anyone raised in a modernist cul-
ture ever to imagine that the value of truth could be diminished. By
‘‘modernist’’ I mean one whose unchallenged categories make strong
boundaries and which supports the categories with functioning in-
stitutions. Yet this diminishing is what many Bible scholars dread
from the questioning of historical truths. David Damrosch puts his
finger on the mood: ‘‘In this way, a desire for living truth may bring
the canonical critic rather close to the hermeneutic relativism of post-
structuralist and deconstructionist literary theorists, who apply a
Nietzschean philosophy of language to argue for the radical indeter-
minacy of all meaning.’’∫
A culture of uncertainty implies radical indeterminacy. Its prophet
and spokesman was Jorge Luis Borges. Evelyn Fishburn quotes him:
‘‘If I am rich in anything, it is in perplexities, not in certainties’’ was
Borges’s upbeat way of expressing what Jean-François Lyotard would
later call the postmodernist loss of faith in grand narratives. An es-
sential skepticism permeates his work, seeking constantly to under-
mine our belief in all systems of knowledge and global explanations,
and indeed questioning the very possibility for such systems to exist.
Rather than in their truth, their value would lie, for Borges, in their
144
THE LATCH
ability to astonish, that is to say, as ‘‘branches of the literature of

fantasy.’’Ω
It is difficult to write about the end of the fashion for composing in
rings without recognizing a present fashion against closure in our
own culture. In a brilliant chapter entitled ‘‘Closure and Anti-Closure
in Modern Poetry,’’ Barbara Herrnstein Smith says that our contem-
porary poets seem ‘‘to reflect a general preference for, and deliberate
cultivation of the expressive qualities of weak closure: even when the
poem is finally closed, it is not usually slammed shut—the lock may
be secure, but the ‘click’ has been muffled.’’ Alongside of this trend,
she notes that the traditional forms of poetic ending are still used and
respected. But in ‘‘all contemporary art anti-closure is a recognizable
impulse,’’ and it ‘‘reflects changing presumptions concerning the na-
ture of art itself.’’∞≠
In a culture of mistrust, language itself becomes suspect and de-
ceiving. In the same book Smith says: ‘‘What is particularly signifi-
cant for poetry, as opposed here to art and music, is the suspicion of
language. . . . Language is the badge of our suspect reason and hu-
manity. It is the lethal trap sprung for truth; it is the reliquary of the
mortmain of the past; it categorizes and codifies, obliterating the
complexities, subtleties and ambiguities of experience.’’∞∞ She goes on
to say: ‘‘Postmodernism is not simply concerned with dismantling the
Enlightenment’s notion of an ascertainable, objective truth, or with
distinguishing between truths and falsehood. . . . But with proposing
the notion of a decentered reality in which a multiplicity of truths
collide in an unhierarchical existence.’’∞≤
Total mistrust and a change of reality—this is powerful stuff, way
beyond my sights. Yet it illumines my central question. We, living
in the postmodernist age, are ‘‘perhaps the heir to too many revolu-
tions . . . we know too much and are sceptical of all that we know, feel
and say. All traditions are equally viable because all are equally sus-
pect. Where conviction is seen as self-delusion and as all last words
are lies, the only resolution may be in the affirmation of irresolution,
and conclusiveness may be soon seen as not only less honest but less
stable than inconclusiveness.’’∞≥ Too much change and too much chal-
145
THE LATCH
lenge to belief, these could be the times in which a polished ring

composition would not be the right medium for expressing what is
uppermost in readers’ minds.
If a culture is indeed heavily against boundaries, rules, and closures
as such, the ring shape would seem too formal, artificial, mechanical.
It will not be popular when the preference is for natural spontaneity.
Instead of being taken for granted as a normal thing, the very idea of
closure is felt to be ambiguous. Is it not better to remain open? Smith
describes a postmodern tendency toward anti-closure in all the arts
and humanities.∞∂ It is true. Our own is an example of a culture averse
to artistic and poetic closure. Modern painting should not have a
finished look, modern music avoids closure, postmodern literature
exemplifies the trend by keeping suspense alive, leaving all possibili-
ties open.
We tend to assume that postmodernism is something that has only
happened to ourselves. It cannot be invoked to explain changes of bias
in ancient civilizations. What we learn about postmodern aversion to
sharp endings would seem irrelevant to the venerable cultures that
first wrote and then allowed ring compositions to lapse. Some readers
will protest that postmodernism is a unique manifestation of our
Western history. But without having been through the Enlighten-
ment, plausible symptoms of the same syndrome can be found in
other periods and places.
Civilizations may rest on different principles of organization. Re-
sorting to cultural theory, I will need to distinguish competitive in-
dividualism from hierarchy.∞∑ The first culture is opportunistic; it
pushes for change and has to find ways of dealing with problems of
identity and uncertainty without blocking innovation. It is attuned to
change; its view of past and future is short term. The other culture is
traditionalist: it tries to put a brake on competition, it reveres and
memorializes the past, and it has a long-term perspective.∞∏
Even ancient societies go through traditionalist and opportunistic
phases. For an example of change from one to another, I cite Ka-
trina McLeod’s fine essay ‘‘The Political Culture of Warring States
China.’’∞π From the eighth to the seventh centuries bce the Chi-
146
THE LATCH
nese Chou dynasty was defeated by the Western barbarians. Formerly

the Chou kings were ‘‘redistributive chiefs.’’ They had been at the
center of a feudal-style network of giving and receiving; in return for
land, families, horses, chariots, and precious metals given to sub-
chiefs, they took tribute and military service. After they were de-
feated, the new kind of regime was based on individual negotiation;
political leadership and military organization developed competi-
tively. The various states achieved varying degrees of centralization.
They competed with each other, made covenants to regulate their
affairs, continually broke their alliances, until a few large states even-
tually emerged. Large numbers of skilled, literate, younger members
of the competing lineages who found themselves without any ter-
ritorial allegiance or institutional constraints, became expert in new
technologies of weaponry, strategy, metallurgy, and fortification and
had much to offer to the new class of peripatetic generals and man-
agers of the estates. By the fifth to the third centuries, the period of
the Warring States, Chinese culture had become thoroughly individ-
ualistic, entrepreneurial, and competitive. Strikingly relevant to the
theme of closure, certainty had disappeared, and so had cognitive
coherence and confidence in language.
The perception of cognitive dissonance in the philosophical texts
was underlined by what Arthur Waley has called the ‘‘language crisis’’
of the period. ‘‘Warring States philosophers wrote about their per-
ception that words (literally ‘names’) and objects (literally the ‘solid’
or ‘really existing’) no longer matched or corresponded.’’∞∫
The ancient Chinese concern for cognitive coherence and lan-
guage suggestively echoes the contemporary concerns with language
and uncertainty. We ourselves with wars and changing technology
have been shifting into a more opportunistic mood, and in certain
aesthetic fields we have a bias against hard-and-fast lines, boundaries,
and endings. Such political shifts, entailing similar effects, could hap-
pen anywhere at any time. After the debilitating wars with Persia,
could a Greek precursor of this modern aesthetic mood have caused
ring composition to fall out of fashion? If so, it may explain some of
our scholarly tendency to rubbish it.
147
THE LATCH
Rules for endings and beginnings are only symptoms. The whole
structure of the ring is rule bound, this without constraining the
creative energies of the poet. After a major revolution or after a pro-
longed war, the survivors long for new forms of expression and signal
their own vitality by rejecting the old. I will take license to answer
Jakobson’s conundrum by reference to social and cultural causes. The
same factors that cause ring composition to lose repute in the first
place could have been at work with us over the past fifty years to hide
it from our vision.
148
NOTES
Chapter 1. Ancient Rings Worldwide
1. Glenn W. Most, The Measures of Praise: Structure and Function in Pindar’s

Second Pythian and Seventh Nemean Odes (Gottingen: Vandenhoeck and
Ruprecht, 1985), 11.
2. Jack R. Lundbom, ‘‘The Inclusio and Other Framing Devices in Deuter-
onomy 1–28,’’ Vetus Testamentum 46 (1966): 296–315.
3. Jacob Milgrom, Numbers: The Traditional Hebrew Text with the New JPS
Translation (Philadelphia: Jewish Publication Society, 1990).
4. Robert Lowth, De Sacra Poesi Hebraeorum. The Oxford Lectures on Po-
etry (1753; trans. Boston, 1829), 157ff.
5. Milgrom, Numbers, xxii.
6. Roman Jakobson and Krystyna Pomorska, Dialogues, trans. Christian
Hubert (Cambridge: Cambridge University Press, 1983), 102.
7. James Fox, ‘‘Roman Jakobson and the Comparative Study of Parallel-
ism,’’ in Roman Jakobson: Echoes of His Scholarship, ed. C. H. van Schoone-
veld and D. Armstrong (Lisse: De Ritter Press, 1977).
8. Jakobson and Pomorska, Dialogues, 103.
9. Quoted in ibid., 102.
10. Quoted in ibid.
11. Martin Schwartz, ‘‘The Gathas and Other Old Avestan Poetry,’’ in La
Langue Poétique Indo-européenne, ed. G.-J. Pinault and D. Petit (Paris,
2006), 1.
12. H.-P. Schmidt, ‘‘Die Komposition von Yasna 49,’’ in Prati Danum, F. J.
Kuiper (London: Mouton, 1968), 170–92. Martin Schwartz, ‘‘Gathic Com-
positional History, Y29 and Bovine Symbolism,’’ in Paitimana: Essays in
Iranian, Indo-European, and Indian Studies in Honor of Hanns-Peter Schmidt,
vol. 2, ed. Siamak Adhami (Costa Mesa: Mazda, 2006), 194–249.
13. Schwartz, ‘‘Gathic Compositional History,’’ 196.
14. Léon Vandermeersch, ‘‘Les origines divinatoires de la tradition chinoise
du paraléllisme littéraire,’’ Extrème-Orient-Extrème-Occident 11 (1989):
11–32.
15. Sarah Allan, The Shape of the Turtle: Myth, Art, and Cosmos in Early China
(Albany: SUNY Press, 1981).
149
NOTES TO PAGES 8 – 14
16. I am grateful to the Czech sinologist Milena Dolezelova for opening to

me this aspect of the history of parallelism.
17. Vandermeersh, ‘‘Les origines divinatoires.’’
18. Ibid.
19. Hua L. Wu, ‘‘The Concept of Parallelism: Jin Shengtan’s Critical Dis-
course on The Water Margins,’’ in Poetics East and West, ed. Milena
Dolezelova-Velingerova. Monograph Series of Toronto Semiotics 4 (To-
ronto, 1988–89), 170.
20. Milena Dolezelova-Velingerova, ‘‘Seventeenth-Century Chinese Theory
of Narrative: A Reconstruction of Its System and Concepts,’’ in Poetics
East and West, 149–55.
21. Dolezelova-Velingerova, ‘‘Seventeenth-Century Chinese Theory of
Narrative,’’ quote on p. 152.
22. Robert Alter and Frank Kermode, The Literary Guide to the Bible (Cam-
bridge, Mass.: Harvard University Press, 1987).
23. Robert Alter, ‘‘The Characteristics of Ancient Hebrew Poetry,’’ in Alter
and Kermode, Literary Guide to the Bible, 611–24; Joel Rosenberg, ‘‘Jere-
miah and Ezekiel,’’ in Alter and Kermode, Literary Guide to the Bible, 184–
206.
24. Ibid., 190.
25. Simon Weightman, ‘‘Structure and the Mathnawi.’’ Presented at the
Rumi Conference, January 25–26, 2002, School of Oriental and African
Studies (SOAS), and Islamic Centre of England. All quotes from Weight-
man from these papers.
26. Seyed Ghahreman Savi-Homani, ‘‘ ‘Love the Whole and Not the Part’:
Investigation of the Rhetorical Structure of Book One of the Mathnawı̄ of
Jalāl al-din Rumi’’ (D.Phil. thesis, University of London, 2003).
27. I am grateful to Eberhard Fincke for this information.
28. Albert Lord, The Singer of Tales (Cambridge, Mass.: Harvard University
Press, 1960). For G. Nagy on creative recall, see his Poetry as Performance
(Cambridge: Cambridge University Press, 1996), 16.
29. Jack Goody, The Power of the Written Tradition (Washington, D.C.:
Smithsonian Institution Press, 2000), 21–22.
30. Aldous Huxley, Vulgarity in Literature: Digressions from a Theme (London:
Chatto and Windus, 1930), 15.
31. Jean Starobinski, Les Mots sous les Mots: Les anagrammes de Ferdinand Saus-
sure (Paris: Gallimard, 1971).
32. Douglas R. Hofstadter, Le Ton Beau de Marot: In Praise of the Music of
Language (New York: Basic Books, 1997), 305.
150
33. Umberto Cassuto, From Noah to Abraham ( Jerusalem: Magnes Press,

1997).
34. I thank Robert Murray for allowing me to publish this analysis of Genesis
2–3 from his unpublished lecture notes.
Chapter 2. Modes and Genres
1. Alastair Fowler, Kinds of Literature: An Introduction to the Theory of Genres

and Modes (Cambridge, Mass: Harvard University Press, 1982), 18.
2. From Leonard B. Meyer’s Music, the Arts, and Ideas: Patterns and Predic-
tions in Twentieth-Century Culture (Chicago: University of Chicago Press,
1967), as quoted in Frank Kermode, The Genesis of Secrecy: On the Interpre-
tation of Narrative (Cambridge, Mass.: Harvard University Press, 1979).
3. Ibid., 162.
4. W. A. van Otterlo, De Ringcompositie als opbouwprincipe in de Epische
Gedichten van Homerus (Amsterdam: Verhandling der Koninklijke Neder-
landsche Academie van Wetenschappen, 1948). Although the book is
written in Dutch, Otterlo has appended a brief resumé in French.
5. G. B. Gray, The Forms of Hebrew Poetry (London and New York: Hodder
and Stoughton, 1915), 37. (The Iliad is actually written in dactylic hex-
ameters, so combining ring form and meter.)
6. This structure would correspond to that of the book of Numbers (which I
shall show in Chapter 4) where the external ring narrates chronologically
the epic journey of the people of Israel from Egypt to the Jordan and is sys-
tematically interrupted by catalogues of laws given by God to Moses and
Aaron. (I am not sure, however, that the interruptions are in ring form.)
7. Jonathan Magonet, Bible Lives (London: SCM Press, 1991), 27.
8. I thank Sybil Stokes for going to look at the monument to check these
details.
9. E. Auerbach, Mimesis: The Representation of Reality in Western Literature
(Princeton, N.J.: Princeton University Press, 1953), 11.
10. Shalom Spiegel, The Last Trial: On the Legends and Lore of the Command to
Abraham to Offer Isaac as a Sacrifice: The Akedah (New York: Behrman
House, 1979), 110–11.
11. Sebastian Brock, ‘‘From Ephrem to Romanos,’’ in Greek Literature in the
Byzantine Period: Greek Literature, vol. 9, ed. Gregory Nagy (New York:
Routledge, 2001), 1–151.
12. Isaac AQEDA, version from Romanos, listed in Le Muséon 99 (1986):
66–67.
151
13. Auerbach, Mimesis, ch. 1, ‘‘Odysseus’ Scar.’’

14. This is my own interpretation of the famous passage that has received
much attention. I have examined only eighteen verses of chapter 22.
Rabbi Jonathan Magonet has published an analysis of the whole Abraham
story, starting from God’s call to Abraham in Genesis 12 and concluding
at the end of chapter 22. Predictably, the mid-turn of this larger ring
comes at a different place. Magonet, ‘‘Abraham,’’ ch. 2 in Bible Lives
(London: SCM Press, 1992), 23–33.
15. Vladimir Propp, Theory and History of Folklore, trans. A. Y. Martin and R.
P. Martin, ed. A. Lieberman (Manchester: Manchester University Press,
1954).
16. Maurice Bloch, ed., Political Language and Oratory in Traditional Society
(New York: Academic Press, 1975).
17. Emile Durkheim, The Elementary Forms of the Religious Life: A Study in
Religious Sociology, trans. Joseph Ward Swain (London: G. Allen and Un-
win, 1915).
18. Joseph Henrich and Francis Gill-White, ‘‘The Evolution of Prestige,’’
Evolution and Human Behavior 22 (2001): 165–96.
19. B. W. Andrzejewski and I. M. Lewis, Somali Poetry: An Introduction (Ox-
ford: Clarendon Press, 1964), 42.
Chapter 3. How to Construct and Recognize a Ring
1. J. W. Welch and D. B. McKinlay, eds., Chiasmus Bibliography (Provo,

Utah: Research Press, 1999).
2. Ibid., 168–69.
3. Roland Meynet, L’Analyse rhétorique: Une nouvelle méthode pour comprendre
la Bible (Paris: Les Editions du Cerf, 1989): ‘‘Il faut sentir à quel point ce
parti contrecarre les habitudes. Il fait plus: il parait contrarier l’impres-
sion immédiate que le texte biblique produit sur le lecteur de l’occident,
qui le trouve souvent malcomposé, pas composé du tout, heurte,’’ 9.
4. Nils W. Lund, Chiasmus in the New Testament (Chapel Hill: University of
North Carolina Press, 1942).
5. Umberto Cassuto, From Noah to Abraham ( Jerusalem: Magnes Press,
1964).
6. Jonathan Magonet, Form and Meaning: Studies in Literary Techniques in the
Book of Jonah, 2nd ed. (Sheffield: Almond Press, 1983).
7. Gary Rendsburg, The Redaction of Genesis (Winona Lake, Ind.: Eisen-
brauns, 1986).
152
8. John Bligh, Galatians: A Discussion of St. Paul’s Epistle (London: St. Paul
Publications, 1969), 37–40.
9. Leonard Muellner, The Anger of Achilles: Menis in Greek Epic (Ithaca, N.Y.:
Cornell University Press, 1996).
10. Gregory Nagy, Poetry as Performance: Homer and Beyond (Cambridge:
Cambridge University Press, 1996), 16.
11. Alastair Fowler, Kinds of Literature: An Introduction to the Theory of Genres
and Modes (Cambridge, Mass.: Harvard University Press, 1982), 29.
12. Douglas Hofstadter, Le Ton Beau de Marot: In Praise of the Music of Lan-
guage (New York: Basic Books, 1997). My own experience as a member of
the Highgate Limerick Society goes to support this opinion.
13. Ibid., 272.
14. Ibid.
Chapter 4. Alternating Bands: Numbers

Second Pythian and Seventh Nemean Odes (Gottingen: Vandenhoeck and
Ruprecht, 1985).
2. Mary Douglas, In the Wilderness: The Doctrine of Defilement in the Book of
Numbers (Sheffield: Sheffield Academic Press, 1993).
3. See David Meijers, ‘‘The Structural Analysis of the Jewish Calendar and
Its Political Implications,’’ Anthropos 82 (1987): 604–10.
4. David Goodman, ‘‘Note on Identifying Beginnings and Endings,’’ in
Douglas, In the Wilderness, 123–26.
5. Ibid.
6. David Goodman’s note on two words for ‘‘trumpets,’’ ibid., 141–43.
7. Quoted in Douglas, In the Wilderness, 123.
8. Goodman, ‘‘Note on Identifying Beginnings and Endings,’’ 124.
9. Ibid., 125.
10. The list of endings for law sections is given in ibid., 126.
Chapter 5. The Central Place: Numbers
1. Mary Douglas, ‘‘Responding to Ezra: The Priests and the Foreign

Wives,’’ Biblical Interpretation 10, no. 1 (2002): 1–23.
2. Graeme Auld, ed., ‘‘Judges 1–.25, in The Conquest under the Leadership
of the House of Judah,’’ in Understanding Poets and Prophets: Essays in
Honour of G. A. Anderson (Sheffield: Sheffield Academic Press, 1993);
153
M. Weinfeld, ‘‘The Period of the Conquest and the Judges as Seen by the
Earlier and the Later Sources,’’ Vetus Testamentum 17 (1967): 93– 111.
3. J. Blenkinsopp, ‘‘The Judaean Priesthood during the Neo-Babylonian
and Achaemenid Periods: A Hypothetical Reconstruction,’’ Catholic Bibli-
cal Quarterly 60, no. 1 (1998): 25–43. I thank Gary Rendsburgh for alert-
ing me to references to the sons of Aaron in important political roles in
Joshua 22.13, 30–32; 24.33; Judges 20.28.
4. Mary Douglas, Jacob’s Tears: The Priestly Project of Reconciliation (Oxford:
Oxford University Press, 2004).
5. Graeme Auld, ‘‘After Exodus and Before Numbers,’’ in The Book of Levi-
ticus: Composition and Reception, ed. Rolf Rendtorf and Robert Kugler (Lei-
den: Brill, 2003), 41–54.
Chapter 6. Modern, Not-Quite Rings
1. Roman Jakobson: ‘‘There is a system of steady correspondences in com-

position and order of elements on many different levels: syntactic con-
structions, grammatical forms and grammatical categories, lexical syn-
onyms and total lexical identities, and finally combinations of sounds and
prosodic schemes.’’ Roman Jakobson and Krystyna Pomorska, Dialogues,
trans. Christian Hubert (Cambridge: Cambridge University Press, 1983),
102–3.
2. Rodney Needham discusses the idea of ‘‘natural universals’’ and warns
that they cannot be expected to have the same interpretation across cul-
tures. Needham, Circumstantial Deliveries (Berkeley: University of Cali-
fornia Press, 1981), 59.
3. Evelyn Fishburn touches on these possibilities in a chapter in ‘‘Traces of
the Thousand and One Nights in Borges,’’ Variaciones Borges 17 (2004):
143–57.
4. Agatha Christie, Five Little Pigs (New York: Dodd, Mead, 1942).
5. Agatha Christie, The ABC Murders (reprint, New York: Dodd, Mead,
1977; orig. William Collins, 1936).
6. Thomas Narcejac, Une Machine à lire: Le roman policier (Paris: Denoël/
Gonthier, 1975).
7. Ibid., 87.
8. The text used in this and the following chapter is from Laurence Sterne,
The Life and Opinions of Tristram Shandy, Gentleman, ed. Melvyn New and
Joan New, with an introductory essay by Christopher Ricks (New York:
Penguin, 2003).
154
9. William Freedman, Laurence Sterne and the Origin of the Musical Novel
(Athens: University of Georgia Press, 1978), 146–47.
10. Ibid., 29.
11. Sterne, Tristram Shandy, Appendix, Glossary of Terms of Fortification,
668.
12. Freedman, Laurence Sterne, 29.
Chapter 7. Tristram Shandy: Testing for Ring Shape
Some parts of this chapter were presented at the Christian Literary Studies
Group conference, Corpus Christi College, Oxford, November 2000, and
published as ‘‘Shandean Mirrors, Puzzles of the Master Analogy,’’ in The Glass
14 (Winter 2001/Spring 2002): 14–29.
1. Robert Lowth, De Sacra Poesi Hebraeorum, based on thirty-four lectures in
Latin delivered at Oxford between 1741 and 1750.
2. Alastair Fowler, Triumphal Forms: Structural Patterns in Elizabethan Poetry
(Cambridge: Cambridge University Press, 1970).
3. Parenthetical references to Tristram Shandy refer to volume and chapter
number.
4. The main Treaty of Utrecht, involving all the European powers, was
finally signed in 1713. It achieved a complete reorganization of Europe,
the dissolution of the Austrian empire, and the freeing of trade.
5. Uncle Toby received his wound and was invalided home in July of 1695,
so some time has elapsed before the events of this book take place.
6. A collection of Greek and Latin religious and philosophical writings was
ascribed to Hermes Trismegistus (first to third centuries a.d.) A fusion of
Stoic, Neo-Pythagorean, and Eastern religious elements, with much cos-
mological and astronomical teaching. (See The Oxford Dictionary of the
Christian Church [Oxford: Oxford University Press, 2005], s.v. ‘‘Hermetic
Books.’’)
7. According to Campbell Ross, Sterne read and admired all three. Eras-
mus’s protests against critics who objected to his rude attacks on theolo-
gians in L’Eloge de la folie are very similar to Sterne’s own apologia in
Tristram Shandy.
8. A point reiterated in Ian Campbell Ross’s biography, Laurence Sterne: A
Life (Oxford: Oxford University Press, 2001).
9. Ibid.
10. Christopher Ricks, ‘‘The Irish Bull,’’ chapter 4 in his Beckett’s Dying Words
(Oxford: Clarendon Press, 1990), 153–56.
155
Chapter 8. Two Central Places, Two Rings: The Iliad
The text used in these chapters is Richard Lattimore’s translation, The Iliad of
Homer (Chicago: Chicago University Press, 1951).
1. John L. Myres, Homer and His Critics, ed. Dorothea Grey (London: Rout-
ledge & Paul, 1958), 46–47.
2. Jack Goody, The Myth of the Bagre (Oxford: Clarendon Press, 1972),
186–87.
3. E. D. Lewis, ‘‘A Quest for the Source: The Ontogenesis of a Creation
Myth of the Ata Tana Ai,’’ chapter 9 in To Speak in Pairs, Essays on the
Ritual Languages of Eastern Indonesia, ed. James J. Fox (Cambridge: Cam-
bridge University Press, 1988).
4. Oliver Taplin, Homeric Soundings: The Shaping of the Iliad (Oxford: Clar-
endon Press, 1992).
5. Ibid., 15.
6. Cedric Whitman, Homer and the Heroic Tradition (Cambridge, Mass.:
Harvard University Press, 1958), 257.
7. Ibid., 256.
8. Parenthetical references are to book and line number; thus, in this case,
book 24, lines 665–67.
9. Whitman, Homer and the Heroic Tradition, 257.
10. Alastair Fowler, Triumphal Forms: Structural Patterns in Elizabethan Poetry
(Cambridge: Cambridge University Press, 1970).
Chapter 9. Alternating Nights and Days: The Iliad
1. Arthur Hatto introduced this term to me, with many lively examples
and the reference to the volume he edited called Traditions of Heroic and
Epic Poetry, vol. 2 (London: Modern Humanities Research Association,
1989).
2. G. E. Duckworth, Structural Patterns and Proportions in Virgil’s Aeneid: A
Study in Mathematical Composition (Ann Arbor: University of Michigan
Press, 1962).
3. Malcolm Willcock, A Companion to the Iliad: Based on the Translation
by Richmond Lattimore (Chicago: Chicago University Press, 1976), 82,
relating to 7.421: ‘‘This is still the same morning, Idaios’ visit to the
Greek camp has taken place very early. Line 433, this must be the dawn of
a new day.’’
156
4. Cedric Whitman, Homer and the Heroic Tradition (Cambridge, Mass.:

Harvard University Press, 1958), 257.
5. Ibid.
Chapter 10. The Ending: How to Complete a Ring
1. I acknowledge stimulus on the competitive aspects of a bard’s recital from

Douglas Frame.
Second Pythian and Seventh Nemean Odes (Göttingen: Vandenhoeck and
Ruprecht, 1985), 42.
3. ‘‘Sarah Stein’s Notes, 1908,’’ ch. 4 in Matisse on Art, ed. Jack Flam (New
York: Phaidon, 1973).
4. Henri Matisse, ‘‘Notes d’un peintre,’’ La Grande Revue 52, no. 24 (De-
cember 25, 1908). I am grateful to Christopher Green for illuminating
this for me and for giving me these references to Matisse.
5. Christopher Green, ed., Art Made Modern: Roger Fry’s Vision of Art (Lon-
don: Courtauld Gallery, Courtauld Institute of Art, in association with
Merrell Holberton, 1999).
6. Mary Douglas, ‘‘Rightness of Categories,’’ in How Classification Works:
Nelson Goodman among the Social Sciences, ed. Mary Douglas and David
Hull (Edinburgh: Edinburgh University Press, 1992), 239–69.
7. Nelson Goodman, Languages of Art: An Approach to a Theory of Symbols,
2nd ed. (Indianapolis: Hackett, 1976).
8. Wittgenstein, Lectures and Conversations on Aesthetics, Psychology, and Re-
ligious Belief, Cyril Barrett, ed. (Oxford: Blackwell, 1966), 5.
9. Quotes from Goodman, Languages of Art, 252. ‘‘A symbol is the more
replete according as proportionately more of its features are functioning
symbolically. For instance, in a linear diagram, only differences in posi-
tion between points on the line, with respect to the coordinates, are
significant; the thickness and color of the line do not matter. But in a line
drawing—a linear depiction—all these and other features are significant.
The drawing is more replete than the diagram,’’ Nelson Goodman and
Catherine Z. Elgin, Reconceptions in Philosophy and Other Arts and Sciences
(Indianapolis: Hackett, 1988), 123.
10. This is part of his distinction between analog and digital systems. Analog
systems are ‘‘dense’’ in the sense that the units of which they are made are
not, like alphabets or thermometers, finely differentiated, whereas the
157
units in digital systems have to be strictly differentiated. There is no sense

in which they can be replete because they are constructed for purposes
that focus only on certain limited properties.
11. This is an oblique reference to T. S. Eliot’s theory of the classic, as dis-
cussed in Frank Kermode, The Classic: Literary Images of Permanence and
Change (Cambridge, Mass.: Harvard University Press, 1983).
12. Graeme Auld, ‘‘After Exodus and Before Numbers,’’ in The Book of Levi-
ticus: Composition and Reception, ed. Rolf Rendtorf and Robert Kugler (Lei-
den: Brill, 2003), 41–54.
13. Margaret Church, ‘‘The First English Pattern Poems,’’ PMLA 61 (1946):
636–50.
14. Alastair Fowler, ‘‘Cut without Hands: Herbert’s Christian Altar,’’ ch 2 in
Presenting Poetry: Composition, Publication, Reception ed. Howard Erskine
Hill and Richard A. McCabe (Cambridge: Cambridge University Press,
1995), 41–51.
15. Jeremy Adler, ‘‘Technopaigneia, Carmina Figurata and Bilder-Reime:
Seventeenth-Century Poetry in Historical Perspective.’’ Comparative
Criticism 4 (1982): 107–47.
16. This literary construction is analyzed in chapters 10 and 11 of Mary
Douglas, Leviticus as Literature (Oxford: Oxford University Press, 1999).
17. Kermode, The Classic, 80.
18. Lawrence Gowing, Turner: Imagination and Reality (New York: Museum
of Modern Art, 1966), 44–45.
19. I thank William Benzone for this information.
20. T. S. Eliot, ‘‘Little Gidding,’’ The Four Quartets, 1942, Collected Poems,
1909–1962 (London: Faber and Faber, 1963), 222.
21. Goodman, Languages of Art.
Chapter 11. The Latch: Jakobson’s Conundrum
1. See ‘‘Jorge Luis Borges’’ in E. Auerbach, Mimesis: The Representation of

Reality in Western Literature (Princeton, N.J.: Princeton University Press,
1953), 99, 74.
2. Ibid., 90.
3. Ibid., 82–83.
4. Martin Heinzelmann, Gregory of Tours: History and Society in the Sixth
Century (Cambridge: Cambridge University Press, 2001), 100.
5. Peter Godman, Poets and Emperors: Frankish Politics and Carolingian Poetry
(Oxford: Clarendon Press, 1987), 8–9, and Heinzelmann, Gregory of Tours.
158
6. Michael Thompson, Rubbish Theory: The Creation and Destruction of Value

(Oxford: Oxford University Press, 1979).
7. Jean Starobinski, Les Mots sous les mots: Les anagrammes de Ferdinand Saus-
sure (Paris: Gallimard, 1971). Saussure collected an analyzed line from
legends, between 1905 and 1909, a long exercise on the nature of identity.
8. David Damrosch, The Narrative Covenant: Transformations of Genre in the
Growth of Biblical Literature (San Francisco: Harper and Row, 1987), 12.
9. E. Fishburn, ‘‘Jorge Luis Borges,’’ ch. 10 in Postmodernism: The Key Fig-
ures, ed. Hans Bertens and Joseph Natoli (Malden, Mass.: Blackwell,
2002), 99, 74.
10. B. Herrnstein Smith, Poetic Closure: A Study of How Poems End (Chicago:
University of Chicago Press, 1968), 234–60, 237.
11. Ibid., 241.
12. Fishburn, ‘‘Jorge Luis Borges,’’ 57.
13. Smith, Poetic Closure, 240–41.
14. Ibid.
15. Aaron Wildavsky, Michael Thompson, and Richard Ellis, Cultural Theory
(Boulder, Colo.: Westview Press, 1990).
16. Cultural theory: This contrast is analyzed in a big literature that at the
latest review counted more than seven hundred titles.
17. Katrina C. D. McLeod, ‘‘The Political Culture of Warring States China,’’
in Essays in the Sociology of Perception, ed. Mary Douglas (Boston: Rout-
ledge, 1982), 140–57.
18. The Analects of Confucius, trans. Arthur Waley (London: Allen & Unwin,
1964), 21–22.
159
INDEX
Aaron, 45–46, 49, 60–63, 66, 69, Art, 127–29, 135–38, 142–43,
151n6; sons of, 62, 63, 67, 154n3 145, 146
The ABC Murders (Christie), 75–76 Assyria, 64
Abraham, 152n14; and Isaac, 18– Auerbach, Erich, 24, 139–42
26, 34, 37, 58, 115 Auld, Graeme, xi, 70–71
Adam and Eve, 14–16, 58
Adler, Jeremy, xii Balaam, 37, 46, 54
Adultery, 49, 53–54, 68 Behl, Aditya, xii
Aeneid (Virgil), 116, 117, 118 Bible: chiastic structure of, xi, 2, 3,
Alter, Robert, 10 33, 87; creation story and Gar-
Alternation principle: in Aeneid, den of Eden in, 14–16, 18, 58;
116, 117, 118; in book of Num- dating of writings of, xi; and her-
bers, 37, 43–57, 59, 85, 116, meneutic relativism, 144; Isaac
118; in detective fiction, 116; in and Abraham in, 18–26, 34, 37,
epic form, 116; in Iliad, 37, 85, 58, 115; misunderstanding of lit-
105, 117–24; in large-scale ring erary structure of, 10–11; New
compositions, 116; in Latin Testament of, 33; parallelism in,
dirges, 116; in ring composition 2, 3–4, 10, 12, 33; poetry of, x,
generally, 37, 116 3–4, 10; ring composition in, x–
Analog versus digital systems, xi, xiii, xiv, 19–26, 152n14. See
157–58n10 also Genesis, Book of; Moses;
Analogies: in book of Leviticus, Numbers, Book of; and other
133–34; and parallelism, 14, 56; books of the Bible
ring composition as double Bligh, John, 33
sequence of analogies, 34; and Borges, Jorge Luis, 144–45
satire, 59; in Tristram Shandy, Borneo, 5
81–82, 96; as wild and willful, 13 Brain and parallelism, 4–6, 40, 72,
Ancient ring compositions, xi, 1– 74, 79, 88, 99–100
16, 28, 38. See also Bible; Iliad Brock, Sebastian, 19–20
(Homer); Numbers, Book of; Burmese, 5
and other books of the Bible
Antithetic parallelism, 3–4 Canaan, 64, 66, 67
Apter, David, ix–x Cassuto, Umberto, 14, 33
161
INDEX
Central loading: in book of Gen- De Saussure, Ferdinand, 13, 144,

esis, 37; in book of Numbers, 159n7
37, 45–46, 58–71; in Iliad, Detective fiction, 74–79, 116
37, 86; in ring composition Deuteronomy, Book of, xiii, 3
generally, 37. See also Mid- Diagram versus drawing, 157n9
turn Digital versus analog systems,
Chiasmus and chiastic structure: in 157–58n10
Bible generally, xi, 2, 3, 33, 87; Divination, Chinese, 8, 9
in book of Numbers, 56–57, 59; Dolezelova, Milena, xii
definition of, 31; in Iliad, 111; Doniger, Wendy, xii
and intentional fallacy, 32; Drabble, Margaret, 77
Lowth on, 85; Milgrom on, 4; in Drawing versus diagram, 157n9
New Testament, 33; panel form Durkheim, Emile, 28
for, 20, 22; and parallel struc-
ture, 6; and pedimental, xiii; Eden story, 14–16, 18, 58
spontaneous instances of, 32– Egypt, 13, 40
33; Welch’s criteria for, 32, 33; Eliot, T. S., 137, 158n11
and Zoroastrian literature, 7. See Ending/closure: of book of Levi-
also Ring composition ticus, 133–34; of book of Num-
China: Chou dynasty in, 146–47; bers, 41, 56–58, 63, 67–71, 129–
divination in, 8, 9; literature of, 30; of collaborations, 136; com-
5, 8–10, 81; Warring States pletion versus, xiii; of different
period of, 147 literary kinds, 136–37; of games,
Christie, Agatha, 74–76 135; and homing, 134–35, 137;
Chronicles, Book of, 41, 71 of Iliad, 41, 130; and reciprocity,
Circles. See Ring composition 135; of ring composition gener-
Closure. See Ending/closure ally, 23–24, 37–38, 40, 43, 126,
Creation stories: in Bible, 14–16, 134–35; rules for, 126; of sec-
58; Hawaiian creation chants, 5; tions in book of Numbers, 47,
recitation of, at beginning of 56–57; and social motives, 136;
political period, 103; ring com- of Tristram Shandy, 96–99;
position of, 27 writers’ and artists’ reluctance
Cues for sections. See Indicators concerning, 135–36, 145
for sections Ephraim, 64, 65
Cultural theory, 146–47 Erasmus, 96, 155n7
Eve and Adam, 14–16, 58
Damrosch, David, 144 Exodus, Book of, xiii, 44, 70, 133
D’Aubignac, L’Abbé, 102 Exposition. See
David, 4 Prologue/exposition
162
INDEX
Ezekiel, Book of, 10 Gowing, Lawrence, 136

Ezra, 64–66, 71 Gray, G. B., 17
Great Britain, 29
Falen, James, 42 Greek classics. See Homer; Iliad
Figure or pattern poetry, x, xii, (Homer)
131, 132, 134 Gregory, Saint, 139–42
Finger rhymes, 12 Griffiths, Alan, xii
Finnish oral poetry, 4 Grody, Jack, 12–13
Fishburn, Evelyn, 144
Five Little Pigs (Christie), 74–75 Harshav, Benjamin, 122
Fowler, Alastair, 41, 131 Hartman, Geoffrey, ix
Fox, James, 4–5 Hatto, Arthur, 156n1
Freedman, William, 83–84 Hawaiian creation chants, 5
Freeman, Austin, 76–77 Heinzelmann, Martin, 140
Fry, Roger, 128 Hermes Trismegistus, 93, 155n6
Herodotus, xii
Galatians, Letter to, 33 Hierarchy, 146–47
Games, 135 Histories of the Franks (Saint Greg-
Garden of Eden, 14–16, 18, 58 ory), 139–42
Gathas of Zoroaster, 6–7 Hofstadter, Douglas, 14, 41–42
Geller, Mark, xii Homer: Muellner’s study of rela-
Genesis, Book of: Abraham and tionship of episodes in, 39;
Isaac in, 18–26, 34, 37, 58, 115, Otterlo’s study of, 4, 17, 117.
152n14; central loading in, 37; See also Iliad (Homer); Odyssey
chiastic macro-structure of (Homer)
entire book, 33; Eden story in, Homing, 134–35, 137
14–16, 18, 58; and pedimental Hopkins, Gerard Manley, 5
composition, xiii, xiv; and prom- Hornblower, Simon, xii
ised inheritance for Israelites,
70; ring composition in gener- Iliad (Homer): Achilles in, 37,
ally, 33 105–7, 111, 113, 114, 119–22;
Genres of literature: definition of, Agamemnon in, 104–7, 111,
17; detective fiction, 74–79; and 113, 120, 122; Apollo in, 104,
ring composition, 74; rules of, 106, 107, 120; audience for, 103;
41; survival of, 143–44 battles in, 110–14; Odysseus in,
Glossalalia, 80 105–7, 111; scholars on struc-
‘‘The Goblet,’’ 132 ture of, 101–6; Trojan War in,
Goodman, David, 47, 56–57 37; Whitman’s macro-scheme
Goodman, Nelson, 128–29, 157n9 of, 104–6, 124
163
INDEX
Iliad (Homer) (continued) Jonah, Book of, 33

—ring composition elements: Josephites, 59, 66–67, 69–71, 130
alternating days and nights Josephus, 102
(dawns and nightfalls), 13, 37, Joshua, as successor for Moses, 66
85, 105, 112–13, 117–24, Joshua, Book of, xiii, 41, 71, 154n3
156n3; central loading, 37, 86; Judah, 64–67, 71
chiasmus, 111; ending, 41, 130; Judea, 64
general discussion, 6, 18, 41, Judges, Book of, 66–67, 154n3
101–14, 124; mid-turns, 105,
107, 108, 109–14, 117, 119, 122; Kent State killings, 19
numerical ring, 105–9, 124; Kermode, Frank, 17, 134, 158n11
panel form of, 110; parallelism, Key words and phrases: and alter-
114, 119–22; problems in struc- nation principle, 116; in book of
ture of nights and days, 122–24; Numbers, 50–51, 54, 69, 70;
two rings in, 101–2, 105–6 and parallelism, 116; in ring
Inclusio, 2, 24, 39, 99, 141 composition generally, 34, 36,
Indian Sufi romance, 38 37, 115
Indicators for sections: of book of Kierkegaard, Søren, 21
Numbers, 37, 45, 54–57, 116; of
ring composition generally, 36– Language, suspicion of, 145
37, 115–16 Latch: of book of Numbers, 47,
Individualism, 146–47 68–71, 130; of ring composition
Indonesia, 5 generally, 43; of Tristram
Intentional fallacy, 32 Shandy, 87, 96, 97–99
Isaac and Abraham, 18–26, 34, 37, Latin writings, 116, 139–42
58, 115 Levites, 44–46, 59–65, 67–71,
Isaiah, Book of, 4 110, 129–30
Leviticus, Book of, x–xi, xii, 64, 70,
Jacob, 3 88, 131, 133–34
Jakobson, Roman: on parallelism Literacy: and memorization, 12–
as deep set in brain, 4–6, 40, 72, 13; and parallelism, 12; as proof
74, 79, 88, 99–100; solution to of worth, 28
conundrum of, 125, 138, 139, Literary genres. See Genres of
148 literature
James, Henry, 73 Literary Guide to the Bible, 10
Jeremiah, Book of, 10, 11 Literary versus thematic structure,
Jerusalem, 64–66 102
Jin Shentang, 9–10 Little Gidding (Eliot), 137
Job, Book of, 10 Lloyd, Geoffrey, xii
164
INDEX
Lord, Albert, 12 ites’ complaints, 40, 54; and

Lowth, Robert, xii, 3–4, 85 Israelites’ victories, 67; and
Lund, Nils W., 33 Josephites, 69–70, 130; Joshua
Lyotard, Jean-François, 144 as successor of, 66; and land for
Israelites, 2, 41, 46; Levites’
Madagascar, 5 revolt against, 45, 61–62, 129–
Madhumalati (Manjhan), 38, 39 30; and Promised Land, 46; and
Magonet, Jonathan, 19, 33 tabernacle, 52, 133
Manjhan, 38, 39 Most, Glenn, 1, 43, 127
Mark, Gospel of, xii Muellner, Leonard, xii, 38–39
Mathnawi (Rumi), 11–12 Murray, Robert, 15
Matisse, Henri, 127–28 Music, 136, 137, 142–43, 145
McLeod, Katrina, 146–47 Myres, John, xii–xiii, 102
Meijers, David, 47 Mythology, x, 27, 103
Memory, 12–13
Metaphors, 42 Nabakov, Vladimir, 42
Meyer, Leonard, 17, 158n11 Nagy, Gregory, 41
Meynet, Roland, 33 Narcejac, Thomas, 76, 77, 78
Mid-turn: of Aeneid, 118; of book Natural universals, 154n2
of Numbers, 45, 47, 55–56, 58– Nazarites, 49, 53, 68
71, 129; of Iliad, 105, 107, 108, Needham, Rodney, 154n2
109–14, 117, 119, 122; of ring Nehemiah, Book of, 71
composition generally, 1–2, 24, New Guinea, 5
34, 36, 37, 40, 43, 55–56; and New Testament, xii, 33. See also
Tristram Shandy, 85–86, 99. See Bible; and books of the Bible
also Central loading Nietzsche, Friedrich, 144
Milgrom, Jacob, xiii, xiv, 2, 4 Nixon, Richard, 19
Milton, John, 115 Noth, Martin, 11
Miracles, xii Novels: Chinese novels, 8–10, 81;
Misreading of ring composition, detective fiction, 74–79. See also
x, 3, 8–11, 43, 56, 58, 71, 139– specific authors and titles
42 Novy, Pat, xiii
Modernist culture, 144 Numbers, Book of: author’s lecture
Montaigne, Michel de, 96 on, xi; Balaam in, 37, 46, 54;
Moses: authority of, 45; and census census of twelve tribes in, 59,
taking, 65–66; divine commands 65–66, 67, 130; compared with
to and responses of, 48, 52, 56, Tristram Shandy, 99; involuntary
151n6; God’s giving of laws to, manslayer in, 53–54, 68; Israel-
on Mount Sinai, 134; and Israel- ites’ complaints and God’s
165
INDEX
Numbers, Book of (continued) Odysseus, 18, 105–7, 111

responses in, 40; Josephites in, Odyssey (Homer), 18, 24
59, 66–67, 69–71, 130; laws and Otterlo, W., 17, 18, 27, 37, 117
narrative in, 47–51, 57, 80, 116,
151n6; Levites in, 44–46, 59– Painting, 127–29, 135–38, 142–
64, 67, 68, 69, 110, 129–30; 43, 146
misreading of, 11, 43, 56, 58, 71; Panel form: of binding of Isaac, 20,
Nazarites in, 49, 53, 68; plague 22; of book of Numbers, 47, 48;
as punishment in, 62; political of Iliad, 110
agenda in editing of, 64–67, 71, Papuan languages, 5
88; significance of twelve sec- Paradise Lost (Milton), 115
tions in, 47, 73; woman accused Parallelism: and analogies, 14;
of adultery in, 49, 53–54, 68 antithetic parallelism, 3–4; in
—ring composition elements: Bible generally, 2, 3–4, 10, 12,
alternation principle, 37, 43–57, 33; in book of Numbers, 3, 45,
59, 85, 116, 118, 151n6; central 46, 47–51, 53–56; Chinese
loading, 37, 45–46, 58–71; forms of, 8–10; compared with
chiastic structure of sections, ring composition, 6; definition
56–57, 59; conventions of, 43– of, 5, 72; in Histories of the Franks,
47; defining narrative sections, 141; Hopkins on poetry and, 5;
52–57; endings, 41, 56–58, 63, in Iliad, 114, 119–22; Jakobson
67–71, 129–30; exposition/pro- on, as deep set in brain, 4–6, 40,
logue, 41, 44, 59, 67–68; general 72, 74, 79, 88, 99–100; and key
discussion, ix, 6, 40–41, 43–57, words, 116; and literacy, 12; and
58, 99; indicators marking indi- poetry, 5; in ring composition
vidual sections of, 37, 45, 54–57, generally, 34, 36, 40, 43, 55–56;
116; key words and phrases, 50– synonymous parallelism, 3, 4; in
51, 54, 69, 70; latch, 47, 68–71, Tristram Shandy, 81–83, 85–95,
130; as macro-composition of 99, 131; worldwide examples of,
large ring, 16; mid-turn, 45, 47, 4–5, 99–100. See also Ring
55–56, 58–71; opening lines of composition
sections, 48–49, 52, 56; panel Pattern or figure poetry, x, xii, 131,
form of, 47, 48; parallel series, 3, 132, 134
45, 47–51, 53–56; passage not Paul, 33
fitting pattern, 122; and pedi- Pedimental writing, xii–xiii, xiv.
mental composition of Hex- See also Ring composition
ateuch, xiii, xiv; phrase inver- Pentateuch, xiii, 32, 43, 66, 135.
sion, 2; split into two halves, See also Bible; and books of the
45, 48 Bible
166
INDEX
Persia, 64, 65, 66, 80, 147 Ricks, Christopher, 99

Picture poems. See Figure or pat- Rig Veda, 6
tern poetry Ring composition: advantages of
Pindar, 1, 43, 127 complexity of, 27–30; alterna-
Pittock, Joan, xii tion principle in, 37, 116;
Plays, 73 ancient ring compositions, xi, 1–
Poe, Edgar Allan, 5 16, 28, 38; central loading in,
Poetry: of Bible, x, 3–4, 10; Chi- 37; compared with parallelism,
nese poetry, 5, 8, 9; closure and 6; compared with syntax, 14;
anti-closure in, 145; figure or construction and recognition of,
pattern poetry, x, xii, 131, 132, 31–42; definition and structure
134; Finnish oral poetry, 4; of, x, 1–2, 31–34, 43; disap-
Hofstadter on poetic form pearance of, 17–18, 139, 143–
versus poetic metaphor, 41–42; 44, 146, 147–48; double-ring
length of, 73; and parallelism, 5; form of, 24–26; as double
ring composition of, 126–27; sequence of analogies, 34; end-
rules for, 115; seventeenth- ing/closure in, 23–24, 37–38,
century poetry, 86; of Somali, 40, 43, 126, 134–35; external
29, 103; structures of, 102–3; enclosing ring versus rings
translation of, 42; Vietnamese within rings, 18, 39; functions
poetry, 5. See also specific poets of, 13–14; indicators to mark
Postmodernism, 142–46 individual sections of, 36–37,
Priamel, 4 115–16; key words in, 34, 36,
Prologue/exposition: in book of 37, 115; latch in, 43; as late
Numbers, 41, 44, 59, 67–68; in archaic style, 17–18; link
ring composition generally, 36, between start and finish in, 1,
43; in Tristram Shandy, 87, 99 31, 33–34, 134–35; macro-
Propp, Vladimir, 27 compositional form of, 16, 33;
Prosimetrum, 116 meaning in middle of, x, 7, 31–
Proverbs, Book of, 3 32; and meanings of ring shape,
Psalms, Book of, 3 73–74; mid-turn of, 1–2, 24, 34,
Pushkin, Aleksandr, 42 36, 37, 40, 43, 55–56; minimum
Pythagorean number theory, 86 criterion for, 1–2;
minor/smaller rings in, 37, 38,
Rabelais, François, 96 43; misreading of, x, 3, 8–11, 43,
Radday, Yehuda, xi 56, 58, 71, 139–42; mood
Rendsburg, Gary, 33 change in, 39; origin of, 12–13;
Repetitions. See Key words Otterlo’s classification of ring
Repleteness, 10, 128–29, 142–43 structures, 18, 37, 117; parallel
167
INDEX
Ring composition (continued) Split into two halves: in book of

series in, 34, 36, 40, 43; of Numbers, 45, 48; in ring com-
poetry, 126–27; pro- position generally, 23, 24, 36,
logue/exposition in, 36, 43; rea- 43; in Tristram Shandy, 81–
sons for difficulty in recogniz- 82, 86
ing, x, 125–26, 139; regional Sterne, Laurence: education of, 85;
variations in, 6, 38, 39, 100; and influences on, 155n7; Lowth’s
repleteness, 10, 128–29, 142– influence on, 85. See also Tris-
43; rings within rings in, 18, 37, tram Shandy (Sterne)
39, 46; rules/conventions of, xi, Sufi romance, 38
1–2, 7, 35–38, 43–47, 72–73, Symbol, Goodman on, 157n9
77, 79, 115; and split into two Symons, Julian, 76
halves, 23, 24, 36, 43; uses of, Syntagmatic axis, 38–39
27–28, 41; worldwide occur- Syntax, 13, 14, 140, 142
rences of generally, x, 4–5, 6, 12,
74, 99–100. See also Chiasmus Tabernacle structure and Levi-
and chiastic structure; Numbers, ticus, 131, 133–34
book of; Pedimental writing; Taplin, Oliver, 103
and specific authors and titles Thai, 5
Rosenberg, Joel, 10, 11 Thematic versus literary structure,
Ross, Campbell, 155n7 102
Roth, Wolfgang, xii Thompson, Francis, 58
Rubbish Theory, 143–44 Thucydides, xii
Rumi, 11–12, 80 Timor, 5
Tortoise, 8, 9
Samaria, 64, 65 Translation, 41–42
Samuel, First Book of, 4, 10 Tristram Shandy (Sterne): Abbess
Sarah, 21, 22, 26 of Andouillets in, 94–95; castra-
Saul, 4 tion in, 82–83, 90–91, 97; child-
Savi-Homani, Seyed Ghahreman, birth in, 83, 92, 98; commenta-
11–12 tors on structure of, 79–81;
Schmidt, Hanns-Peter, 6–7 compared with book of Num-
Schwartz, Martin, 6–7 bers, 99; composition of, over
Segal, George, 19 many years, 85, 88, 96; crit-
Sennacherib, 64 icisms of, 98; death in, 90;
Sephirot, 47 digressions and interruptions in,
Smith, Barbara Herrnstein, 145–46 80, 84, 92, 94–96; French nuns
Social Darwinism, 125 in, 94; language of military for-
Somali poetry, 29, 103 tifications and language for
168
INDEX
human bodies in, 82–83; umes as units of structure,

Lowth’s influence on, 85; narra- 86–87
tor of, 81, 88; on noses and Turner, J. M. W., 96, 136
names, 82, 92–94; Peace of Turtle, 8, 9
Utrecht in, 87, 89–90, 93,
155n4; sex interrupted as meta- Van Dine, S. S., 77, 78
phor in, 84, 95–96, 98; Slawken- Vietnam War protests, 19
bergius in, 94; themes of, 81, 96; Vietnamese poetry, 5
Toby in, 81–83, 88–93, 96–97, Virgil, 116, 117, 118
155n5; toy fortifications in, 81–
82, 89, 92, 93; Tristram as baby Waley, Arthur, 147
in, 90, 92–94, 98; Walter in, 81, The Water Margins, 8
82, 84, 88, 92–95, 97–98; war Weightman, Simon, xii, 11
in, 81–82, 89–90, 92–93; Welch, J. W., 31, 32, 33
Widow Wadman’s love for Toby Wellhausen, Julius, 43
in, 81–83, 90, 91, 93, 96–97 Whitman, Cedric, 104–6, 107,
—ring composition elements: 108, 124
analogies, 81–82, 96; division Willcock, Malcolm, xii
into two distinct ‘‘books,’’ 81– Witte, Dianne, x
82, 86; ending of, 96–99; form Wittgenstein, Ludwig, 128
and content, 83–84, 96; general Wu, Hua I., 8
discussion of, 79, 85–100, 124,
131; lack of central place, 85– Yin and Yang, 8
86, 99; latch, 87, 96, 97–99; Yugo-Slav peasants, 12
linking and cross-referencing,
91–99; parallelism, 81–83, 85– Zelophehad, daughters of, 69–70
95, 99, 131; prologue/exposi- Zodiac, 47, 73
tion, 87, 99; repetitions, 85; vol- Zoroastrian literature, 6–7
169

(The Terry Lectures Series) Mary Douglas - Thinking in Circles - An Essay On Ring Composition-Yale University Press (2007)

Uploaded by

Copyright:

Available Formats

(The Terry Lectures Series) Mary Douglas - Thinking in Circles - An Essay On Ring Composition-Yale University Press (2007)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(The Terry Lectures Series) Mary Douglas - Thinking in Circles - An Essay On Ring Composition-Yale University Press (2007)

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What structures does the author discuss?

What structures does the author discuss?

the terry lectures

Becoming: Basic Considerations for a Psychology of Personality

YALE UNIVERSITY PRESS ≤ NEW HAVEN AND LONDON

Printed in the United States of America.

Library of Congress Cataloging-in-Publication Data

A catalogue record for this book is available from the

The paper in this book meets the guidelines for perma-

The deed of gift declares that ‘‘the object of this foundation is

1 Ancient Rings Worldwide 1

2 Modes and Genres 17

3 How to Construct and Recognize a Ring 31

4 Alternating Bands: Numbers 43

5 The Central Place: Numbers 58

6 Modern, Not-Quite Rings 72

7 Tristram Shandy: Testing for Ring Shape 85

8 Two Central Places, Two Rings: The Iliad 101

9 Alternating Nights and Days: The Iliad 115

10 The Ending: How to Complete a Ring 125

11 The Latch: Jakobson’s Conundrum 139

1. Ancient Rings Worldwide

6. Modern, Not-Quite Rings

To publish at last an essay on ring composition is a great personal

hearing Wolfgang Roth’s lecture on the pattern of the miracles in

F' Exod. 35–40

E' Lev. 1–Num. 10:10

D' Num. 10:11–36:13

splitting of the Jordan

“Put off your shoes...holy”

B' Josh. 5–12 pesah

A' Josh. 13–24

A new interest in ring composition has lately arisen. This antique

preparatory to working back to the beginning, and the whole series

He expostulates, but when they promise to stay ﬁghting with their

ated and understood by scholars when it includes only a few sen-

When Israel went out of Egypt

A If only we had died

Antithetic parallelism: an image illustrated by its contrary, in the same

Synthetic or constructive parallelism: the parallel lies only in the con-

Example from Deuteronomy 32.2:

Or antithetic and synthetic parallelism can be combined.

Example from Isaiah 9:

Bible scholars do not use Lowth’s terminology nowadays, but they

further evidence of it in major collections of texts from the Aus-

Ring composition is parallelism with an important difference. It is

Table 1. Zoroastrian hymns or gathas from the thirteenth

1—11 1—11 1—12

Schmidt’s seminal observations of elements of ring composition in

We can go even further back in time to discover ancient Chinese

[To view this image, refer to

Fig. 2. Ancient Chinese divination on the tortoise shell.

‘‘Repetition,’’ ‘‘clumsiness,’’ ‘‘episodic’’! The complaint of incoher-

endowing it with rhythmic dynamism and dramatic tension. The

the rules of a well-conceived and well-executed ring composition as I

spread it changes the whole basis for social life by redistributing

cross-referencing. A citation from Douglas Hofstadter brings home