THE QUIXOTIC NOVEL
FROM THE POINT OF VIEW OF THE NARRATIVE
JOHN J. CONNOR
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
Dr. Alistair Duckworth inspired this study months
before it was begun and directed it, with patience and
understanding, through the many months it took to write
it. I will always be grateful to him for his guidance
Dr. Gregory Ulmer read the manuscript in an in-
complete version and at the most crucial moment in the
enterprise, when I thought I was at a dead end, showed
me how to proceed. The better parts of Chapters Four
and Five, in particular, I owe to him.
I am also grateful to Dr. William Childers, Dr. John
Perlette, and Dr. Fernando Ibarra, the other members of
my supervisory committee. They were always helpful and
encouraging and, although given little time, they read my
manuscript with care.
Finally, I owe much more than can be expressed here
to Jesse, my wife. She has endured some trying times and
lent me strength. Of course, I could not have done it
TABLE OF CONTENTS
ACKNOWLEDGMENTS...... ........................ ....... ii
ABSTRACT ...... .......................... ........ iv
CHAPTER ONE INTENTIONS, PROBLEMS, AND
METHODS ........ ............ .... ..... . ........ .
NOTES ...... .................................... 20
CHAPTER TWO DON QUTXOTE IN LITERARY HISTORY:
A SURVEY OF INTERPRETATIONS ........... ........ 23
NOTES .... .................. ...... ....... ..... .. 71
CHAPTER THREE SANCHO PANZA FROM THE POINT
OF VIEW OF THE NARRATIVE ........................ 76
NOTES ...... ................. ...... .. ...... .... 126
CHAPTER FOUR THE CHARACTER OF DON QUIXOTE
AND TIE DISCOURSE OF DON QUIXOTE ................ 128
NOTES ................ .................. ......... 196
CHAPTER FIVE MOBILE FRAGMENTS: ELEMENTS OF
THE NARRATIVE SYSTEM OF THE QUIXOTIC NOVEL
IN MOBY DICK, MADAME BOVARY, MIDDLEMARCI,
AND THE GREAT CATSBY................. .. ..... ..... 198
NOTES ........... .... .... . ...................... 250
BIBLIOGRAPHY ............ ..... .... ................... 251
BIOGRAPHICAL SKETCH..................... .. ........... 255
Abstract of Dissertation
Presented to the Graduate Council
of the University of Florida
in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy
THE QUIXOTIC NOVEL
FROM THE POINT OF VIEW OF THE NARRATIVE
John J. Connor
Chairman: Alistair M. Duckworth
Major Department: English
This study is an examination of the narrative system
of the Quixotic Novel. A method of analysis synthesizing
Vladimir Propp's methods in Morphology Of The Folk Tale
and Roland Barthes's methods in S/Z is proposed, described,
and used to investigate narrative structures in Don Quixote
and four other Quixotic Novels, Moby Dick, Madame Bovary,
Middlemarch, and The Great Gatsby.
The narratives of the later Quixotic Novels seem to
be linked to each other and to Don Quixote by their connec-
tion to an abstract substratum, a narrative system which is
too complex to be fully described but which can be identi-
fied. Certain narrative elements found in the narrative
system of Don Quixote reappear in each of the later novels.
These "mobile fragments" take different configurations
in each of the novels, but by studying these configurations
we can compare the novels and we can examine how narrative
structure engenders meaning in the five novels.
INTENTIONS, PROBLEMS, AND METHODS
This study began as an attempt to describe the narrative structure
of the Quixotic Novel. I was attracted to the subject for two reasons:
first, because I had noticed the term "Quixotic Novel" was used fairly
often and rather loosely to characterize a cluster of interesting novels,
but the term had, as far as I could tell, no commonly accepted, author-
ized definition; and second, because certain critical texts had inter-
ested me in the "structural" analysis of narrative. I hoped I would be
able to fine some simple formula to describe the narrative system in Don
Quixote and use that formula to analyze other members of the genre
of the Quixotic Novel.
It seemed to me then, and it does still, that the genre of the
Quixotic Novel is not characterized so much by metaphysical, philosophi-
cal, or psychological content (as it is often argued) as much as it is
characterized by a certain narrative structure. I have found, unfortu-
nately, that there seems no simple way of describing that structure.
Instead, it seems best to describe the narrative structure of the
Quixotic Novel as a "complex of narrative patterns." Another way to
describe it is as "an elastic structure": although no single narrative
formula generates all the narrative in any single Quixotic Novel, the
narratives of Moby Dick, Madame Bovary, Middlemarch, The Great Gatsby,
and, I think, other Quixotic Novels show tendencies to repeat the
narrative patterns and forms in Don Quixote.
Chapters Three and Four of this study are an investigation of the
patterns and fonrs in the narrative system of Don Quixote. Chapter
Three is an argument for the demystificationn" of the character of
Sancho Panza: an examination of Sancho's functions in two kinds of
narrative sequence suggests that his character is not an independent,
homogeneous entity; it is, instead, a complex of functions generated
by the necessities of the narrative discourse. Chapter Four is a
qualification and extension of the arguments of Chapter Three: I
argue there that by examining the relationship between Don Quixote's
character and the discourse of the novel we can make a preliminary
sketch of the "model" that generates Don Quixote. Chapter Five is a
brief investigation of the other novels named above: I argue there
that the analysis in Chapters Three and Four helps us to identify what
Roland Barthes would call "the mobile fragments" in the Quixotic Novel-
the units "whose differential situation engenders" meanings in Moby
Dick, Madame Bovary, Middlemarch, and The Great Gatsby.
It seems to me that we can recognize Quixotic Novels intuitively:
reading Moby Dick or Madame Bovary we sense that Melville and Flaubert
are repeating and modifying patterns we have read in Don Quixote and
in other Quixotic Novels. Of course, neither Melville nor Flaubert nor
any novelist has set out to write a novel according to the formulas
and diagrams used in this study to describe the patterns in Don Quixote.
Nevertheless, Cervantes's narrative system is repeated and modified in
Moby Dick, Madame Bovary, and other novels.
The question of how this system is passed from Cervantes down to
other authors is an interesting one. Some structuralists would argue
that it is not Cervantes who determines the form of the Quixotic Novel;
it is a language system that "speaks" Cervantes. Although my own
opinion is close to that and in Chapter Four I suggest a theory about
the mechanics of Cervantes's influence on other authors, I intend to
leave the final resolution of this question to better qualified scholars.
My intention here is more limited: to describe as best I can elements
in Don Quixote which reappear in other, later novels. This study is,
then, only a first step toward a fuller understanding of the genre of
the Quixotic Novel. I hope it is a useful beginning. I hope my reader
will find here a new way of looking at Don Quixote and the cluster of
There are methodological problems with a study of this kind. First
of all, there is the problem of Cervantine criticism: it is immense
(Manuel Duran believes no scholar could read even half of it in a life-
tim3 and it contains some of the best work by some of the most learned
and brilliant readers of Western literature. Ivan Turgenev, Miguel de
Unamuno, Ortega y Gasset, Jorge Luis Borges, and W. H. Auden have all
written about Don Quixote, as have Salvador de Madariaga, Erich Auerbach,
Leo Spitzer, Gerald Brenan, E. C. Riley, Harry Levin, Rene Girard,
Dorothy Van Ghent, Claudio Guillen, Robert M. Adams, and Robert Alter.
Many of the studies of Don Quixote, in fact, deserve full-length studies
themselves. In Chapter Two I have examined a few of them in brief
detail. My intention in that chapter is not to summarize all that has
been written about Don Quixote, but rather to give an example of the
kind of approach generally adopted by Cervantine critics. Generally,
critics interested in examining Cervantes's influence on literary
history do not focus on what Roland Barthes would call the "implicit
system of units and rules" in the narrative of Don Quixote.
Barthes's S/Z (1970, trans. 1974), his study of Balzac's
"Sarrasine," is one of the texts that interested me in investigating
the narrative system of Don Quixote. The other text is Propp's
Morphology Of The Folk Tale (1928, trans. 1958, 1968). Using these
two texts as sources creates a second methodological problem: the
body of structuralist criticism, although not yet as immense as the
body of Cervantine criticism, has become so large that it is now
difficult to define exactly what structuralist criticism is. It is
for this reason that I have avoided using the tenns structuralistt"
and "structuralism" to describe the kind of criticism I am doing in
the analysis of the narrative of Don Quixote. I have instead adopted
a phrase from Propp which characterizes his kind of analysis and, I
think, the kind of analysis I am doing here. Propp looks at his tales
"from the point of view of the narrative." He means that the text
must be seen as a self-generating entity, a system operating by laws
which are not apparent to the average reader. Propp compares his ap-
proach to that of the granmarian cataloguing the parts of speech of a
language: in precise, objective descriptions the analyst finds the
"abstract substrata" that determine the form of narrative. Whether
or not this approach is structuralistt" (Robert Scholes calls Propp,
"the first pope of the orthodox sect [of structuralism]" Propp's
Morphology is a valuable work of criticism and the methods in it are
worth imitating. I have the same feeling about S/Z. Roland Barthes
has gone on to other things and other methods since his "structural
analysis" of "Sarrasine" in 1970 and so has continued to change his
readers' ideas of what "structural analysis" is, but for me his most
valuable work appears in S/Z. I think a synthesis of Propp's methods
in his Morphology and Barthes's methods in S/Z is possible. This
synthesis is described in the second half of this chapter and forms the
starting point for my approach to the Quixotic Novel "from the point
of view of the narrative."
Two other methodological problems should be mentioned. The use
of symbolic notation is a problem. Perhaps the most unappealing as-
pect of Propp's Morphology and of much recent narratology is the use
of mathematical and pseudo-mathematical terminology and formulas to
describe literary elements. Propp ends his book with a six-page
appendix listing more than two hundred symbols for narrative elements
(W* stands for "wedding and accession to the throne"; H stands for "a
game of cards"). In a recent essay analyzing two passages of narrative
William 0. Hendricks concludes a line of thought by stating, "The terms
S1 and S2 are contraries whereas S1 and S1 are contradictories. A
relation of presupposition exists between S] and S32, i.e. SL implies S2
(likewise, S2 and S ) but not vice versa." This kind of mathematical
description bothered me, not because it cannot be used effectively--
Propp, Hendricks, and, recently, John Holloway have used symbolic nota-
tion very effectively--but because it seems, at first glance, disingenu-
ous. I intended to avoid using such notation in this study. I found,
however, that there seemed no easier way to describe the narrative
patterns and repetitions described in Chapters Three and Four than by
the use of symbolic notations and diagrams. I ask my reader to bear
with me. I have tried to keep the symbols and diagrams to a minimum.
The last methodological problem I will mention is the problem of
the list of Quixotic Novels. Just as there is no authorized defini-
tion of the term "Quixotic Novel" there is no authorized list of all
the members of the set. In an essay entitled "The Quixotic Principle"
Harry Levin lists more than two dozen books which seem to work by the
same principle (which he avoids defining): Joseph Andrews is like
Don Quixote because it is about the problem of literary influence;
Fathers And Sons and War And Peace are like Don Quixote because they
have characters (Bazarov and Count Pierre) who are "quixotic zealots";
Pickwick Papers, "with its paunchy middle-class knight and its boot-
shining cockney squire, is decidedly Cervantine"; A Farewell To Arms
is like Don Quixote because it explodes "the rhetoric of abstraction";
Joyce is like Cervantes in that he is essentially a parodist; the list
goes on. Leo Spitzer in an essay entitled "On the Significance Of
Don Quixote" names Sterne, Goethe, Melville, Flaubert, Balzac, Mau-
passant, Thackeray, Tolstoy, Proust, Mann, and Faullner as imitators
of Cervantes. Originally, I hoped to be able to verify (or falsify)
the various lists of Quixotic Novels by describing the simple formula
that structured the narrative of Don Quixote. As soon as I realized
that the structure of the narrative system in Don Quixote was a com-
plex of patterns rather than a simple pattern, however, I knew I would
not have the time to consider many of the novels various critics have
claimed are "Quixotic." I have had to limit my investigation to four
which are named often as members of the genre. Toby Dick, Madame Bovary,
Middlemarch, and The Great Gatsby will certainly be included in the com-
plete history of the Quixotic Novel, if it is ever written.
2. Propp And Barthes: A Synthesis
No structuralist has yet produced a model of narrative form con-
vincing enough to be universally accepted. Instead, it sometimes
seems there are almost as many theoretical structures as there are
structuralist theorists: Levi-Straus's mythemes, Claude Bremond's
triads, Claude Greimas's actants and modes, Tzvetan Todorov's minimal
schema, and many others. (There are at least two excellent surveys
of these various theories: in Robert Scholes's Structuralism In Lit-
erature and in Jonathan Culler's Structuralist Poetics .) It seems
the only theory with which all of the above-named critics agree is that
there is a system in narrative. All seem to agree with Barthes who
has declared, "No one can produce a narrative without referring himself
to an implicit system of units and rules."
Even the most unsophisticated reader has, I think, an intuitive
understanding of narrative which is close to (although not quite the
same as) Barthes's: narrative is essentially a special form of lan-
guage; it operates by laws that do not apply to ordinary discourse.
Even the most unsophisticated reader reads narrative in a different
spirit and with different expectations than he reads other pieces of
printed material. Opening Don Quixote we are expecting Cervantes to
perform a certain task, to give us a novel, to give us a unified--and
so, by implication, a systematic--work.
The difficulty narratologists have encountered is that it is much
harder to describe this system of units and rules than it is to sense
that it is there. We know it is there, but we haven't yet satisfactor-
ily been able to put it into a meta-language; we haven't yet managed
to find an absolutely satisfactory way to articulate the system.
I find Propp's Morpholojg Of The Folk Tale and Barthes's S/Z es-
pecially admirable because each combines an intuitively appealing
theory (in Barthes's case it is actually a system of theories) with a
close practical application. Propp examines one hundred Russian fairy
tales; Barthes examines Balzac's short story "Sarrasine." Both critical
texts are complex and difficult, but both are based on fairly simple
principles. Furthermore, they seem to be, at least in a few ways,
closely related, and a synthesis of the two methods is possible.
Propp believes, ard presents quite convincing evidence to prove,
that there are a limited number of elements that make up all fairy
tales (or, as he carefully puts it, all the fairy tales he examines).
He argues that there are really just two kinds of elements that make
up all tales: "spheres of action" (the roles characters play) and
"functions" (the acts characters perform). There are, in the one
hundred tales he reads, only seven different spheres and only thirty-
one different functions. The number of possible permutations of the
spheres and functions is, of course, large, but the number of basic
elements is small. "This explains," Propp writes, "the two-fold quality
of a tale: its amazing multiformity, picturesqueness, and color, and,
on the other hand, its no less striking uniformity, its repetition."
Barthes is also interested in repetition in narratives. In S/Z
he argues that no writer invents a story or even the elements in a
story; the writer, instead, invokes the "already-written." The
already written is the source of the five codes Barthes finds operating
Ln "Sarrasine." Balzac, according to Barthes, does not create some new
tale out of his imagination; he draws upon, and so repeats, certain
formulas and elements already existing in the language.
It is in this way that Propp and Barthes are close. (Barthes has
read Propp, by the way.) Both maintain that taletellers do not invent
tales in the way it is usually thought that they do. What taletellers
do is "interweave" (the word appears in both books) certain, recogniz-
able, familiar, generic elements. By this argument taletelling is not
an act of creation; it is an act of combination, of perrutation.
The two theories are close in another way. Both Propp and Barthes
understand chronology to be composed of a series of smaller, chronologi-
cal units. Propp calls these units "moves"; Barthes calls them elements
in the "actional code." Neither is as precise about these elements as
he might have been, however.
Propp's distinction between what he means by a move and what he
means by atale is perhaps the least precise passage in what is, on the
whole, a careful and very precise analysis. He recognizes that the
definition he has presented for the tale can also be used to define
tales-within-tales and even smaller chronological sequences, but he is
not quite able to describe the differences separating each of these
Iorphologically, a folk tale may be termed
any development out of a villainy, or a lack
through intermediary functions, to marriage or
to other functions used in the capacity of the
denoument....This type of development is termed
by us a move. Each new act of villainy, each
new lack creates a new move. One folk tale
may have several moves; and one must first of
all determine, when analyzing a text, the
number of moves it consists of....One move may
irriediately follow another; yet they may also
interweave; a development which has begun pauses,
and a new move is inserted. Singling out a move
is not always an easy matter, but it is always
possible with complete exactness. However, if
we conditionally define the folk tale as a move,
this still does not mean the quantity of folk
tales corresponds exactly to the quantity of
moves. Particular devices of parallelism,
repetitions, etc., lead to the point where one
folk ta c ray be comprised of (sic) several
Propp (or his translator) is not, clear on this point: a tale cannot
be both "any development out of a villainy or a lack" and also consist
of several moves which are themselves developments from villainies
and lacks. Nevertheless, Propp's suggestion about the kinetic force
of narrative has an appeal and has been adopted by several other
analysts: narrative moves, by this argument, from a deficiency (a
villainy or a lack) toward the liquidation of that deficiency. Tzvetan
Todorov has explained the principle in other words:
All narrative is a movement between two
equilibriums which are similar but not identical.
At the start of the narrative there is always a
stable situation....Subsequently, something occurs
which introduces a disequilibrium....At the end 19
of the story....the equilibrium is re-established.
Barthes calls the smallest actional elements in the
chronology of "Sarrasine" the "proairetisms." In his sys-
tem the actional code is one of the two "irreversible" codes
(the other is the hermeneutic code), that is, it can be read
in only one direction and it thus creates (with the hermeneu-
tic code) the chronological structure of the narrative.
These sequences work because the reader recognizes them.
The actional code structures the narrative because the
reader knows what to expect once he realizes that the narra-
tive contains a stroll, a murder, a rendezvous, or some other
familiar sequence. Barthes explains:
Actions can fall into various sequences which
should be indicated merely by listing them, since
the proairetic sequence is never more than the
result of the artifice of reading: whoever reads
the text amasses certain data under some generic
titles for actions (stroll, murder, rendezvous),
and this title embodies the sequence; the sequence
exists when and because it can be given a name,
it unfolds as this process of naming takes place,
as a title is sought or confirmed...20
The sequence exists, then, and the narrative has chronological
structure because the reader is familiar with the generic form
the sequence embodies: "the sequence exists when and because
it can be given a name..." "Naming" a sequence means recog-
nizing its genre; if we can recognize the genre, the sequence
must exist somewhere in the already written (and, of course,
in the already read). If the sequence did not previously exist,
if the author managed to create some absolutely original se-
quence, we could not name it, we would not know it: there can
be no generic title for something that has never been known to
Propp's move can be compared to Barthes's proairetism
by observing how Propp and Barthes describe the kinetic force
of these sequences. For Propp it is a movement from the
invocation of a deficiency to the liquidation of the deficiency,
from an imbalance to the restoration of balance; for Barthes
the sequence "unfolds" as the reader names it. The two de-
scriptions seem different unless we examine one important
passage in S/Z where Barthes draws his theory close to
Propp's. It is in his analysis of the first clause of the
first sentence of "Sarrasine." The story begins, "I was
deep in one of those daydreams..." Barthes maintains that
the antithesis implied here (between dreaming and waking)
is an element in the symbolic code, but then in one of
those strange and ingenious transitions that characterize
his work, Barthes goes on to describe the actional code,
the code of the proairetisms: the sequential nature of the
antithesis is important--dreaming must be following by
The state of absorption formulated here
(I was deep in...) already implies...
some event which will bring it to an end
(when I was roused by a conversation...)
Such sequences imply a logic in human
behavior. In Aristotelian terms, in which
oraxis is linked to proairesis, or the
ability rationally to determine the result
of an action, we shall name this code of
actions and behavior proairetic...21
In that transition is hidden the link between Barthes's
analysis of chronology and Propp's: both of them find at
the heart of chronological structure the sequential connec-
tion which is as old as reason, the sequential connection
between cause and effect.
A fairy tale, according to Propp, contains chronological
movement because a cause is invoked (a dragon steals a king's
daughter, thus performing a villainy and creating a defi-
ciency) and an effect is promised (something will complete
the sequence: the king will save, or lose, his daughter).
"Sarrasjne," according to Barthes, contains chronological
movement because the Droairetism is a "logico-temporal"
element (daydreams must always lead to awakenings in a
sequence which can be entitled "To be deep in"). The usual
paradigm of proairetisms, Barthes argues,
is something like begin/end, continue/stop.
....T he nature of the phenomenon established
by the notation is capped by a conclusion
and consequently seems to be the subject to
some logic (as long as temporality appears:
the classic narrative is basically subject
to the logico-temporal order). Writing
"the end" (a phrase which is precisely both
temporal and logical) thus posits everything
that has been written as having been a tension
which "naturally" requires resolution, a
consequence, an end...22
A hybrid form of the two explanations would look like
this: the fundamental tissue of chronology is woven out
of recognizable cause and effect sequences. A story has
chronology because it contains the sequential action of
cause and effect; there may be several sequences and they
may interlock (one may pause while another moves toward
its conclusion), but each cause must have an effect; each
time a deficiency is invoked, a cause announced, the narra-
tive must eventually move toward the liquidation of that
deficiency, toward the effect which will bring the conclu-
sion to the sequence. The kinetic energy of narrative
chronology is created by disequilibrium: an imbalance moves
the reader into and through the narrative, the reader reads
on to have his sense of equilibrium restored.
There is another clear connection between the Morphology
and S/Z: each involves the demystification of character.
Both Propp and Barthes are intent upon examining characters
as narrative elements rather than as mimetic constructions.
Propp proposes that character be examined "from the point
of view of the narrative." He analyzes character in the
hundred tales by tabulating the functions performed by the
dramatic personae. "The functions of a folktale's dramatic
personae must be considered its basic components," he argues.
Propp looks at character to see how the character's acts
energize the narrative.
Barthes's demystification of character seems more
radical than Propp's because "Sarrasine" is a serious work
of literature, but the approach is not entirely new. In
Anglo-American letters the demystification of character goes
back at least to 1933 when L. C. Knights published his well-
known, iconoclastic essay "How Many Children Had Lady Macbeth?"
Knights argues there that the psychologically-based and-biased
criticism of A. C. Bradley and others has led to an exaggerated
and inaccurate emphasis on "character," a word Knights is care-
ful to surround in quotation marks. The habit of regarding
Shakespeare's persons as real, he argues, "is responsible for
most of the vagaries that serve as Shakespeare criticism."
Knights believes the analyst must keep in mind that Shakespeare's
plays are essentially dramatic poems: character is just one
element in "the whole dramatic pattern of each play."
Knights could not have anticipated the special intensity
with which the author of S/Z demystifies character, however.
At one point in his analysis Barthes examines a passage in
which the central character chooses to keep his rendezvous
with La Zambinella despite the warning of a stranger who has
whispered, "Be on your guard, Frenchman. This is a matter
of life and death. Cardinal Cicognara is her protector and
doesn't trifle." Barthes's commentary is a masterpiece of
his special methods. It is entitled "The Story's Interest"
and reads in part:
....Sarrasine is not free to reject the Italian's
warning; if he were to heed it and refrain from
pursuing his adventure, there would be no story.
In other words, Sarrasine is forced by the dis-
course to keep his rendezvous with La Zambinella:
the character's freedom is dominated by the dis-
course's instinct for Dreservation....This game
is an economic one: it is to the story's interest
that Sarrasine ignore the stranger's dissuasion;
he must at all costs keep the duenna's rendezvous.
In other words, he keeps it to ensure the very
survival of the anecdote.... 27
This is, in my opinion, one of the most valuable passages
in S/Z. The rule of narrative described here--that the
discourse has an "instinct for preservation"--is a rule I
tried to keep in mind while examining Don Quixote and the
other novels. I will return to it later. Sancho Panza and
Don Quixoi.t atie successful characters, I will argue, not be-
cause they represent types or the extremes of man's understand-
ing of his place in the world, but because they serve so well
the discourse's "instinct for preservation." Sancho and
Quixote are successful characters because the functions they
perform as characters generate narrative.
There is at least one important difference between
Propp's and Barthes's methods: Barthes is also intent upon
demystifying the cultural significance of the literary work.
For Barthes (and other structuralists) the work is a kind
of bricolage, a collection of already-written fragments.
For Barthes theme is not, as it is ordinarily considered,
the connection between story and meaning or form and content;
theme in S/Z is just another formal element (to be found in
the semic and cultural codes). In what is perhaps the best
known passage in all his work Barthes contrasts in a meta-
phorical comparison the traditional understanding of form and
content with his own understanding. "Until now," he writes
in an essay entitled "Style And Its Image,"
we have looked at the text as a species of fruit
with a kernel (an apricot, for example), the
flesh being the form and the pit being the content;
it would be better to see it as an onion, a con-
struction of layers (or levels or systems) whose
body contains, finally, no heart, no kernel, no
secret, no irreducible principle, nothing except
the infinity of its own envelopes--which envelop 28
nothing other than the unity of its own surfaces.
Propp would not have accepted this argument in his
Morphology. Propp intended his analysis of the fairy tales
to be only the first step to a full understanding of them.
He saw his work as the necessary systematic beginning to
later more "interesting" speculation (about the cultural
value of the tales, about the apparent similarity of tales
in many cultures, etc.) He writes at the end of his first
We shall not refuse to take upon ourselves
the rude, analytical, somewhat laborious work
[of compiling the morphology of one hundred
stories which is complicated by the fact that
it is undertaken from the standpoint of abstract,
formal questions. Such crude, "uninteresting"
work of this kind is a way to generalize "inter-
Alan Dundes, one of Propp's American advocates, re-emphasizes
this point in his introduction to the most recent English
translation of the Morphology:
Propp's systematic approach has unfortunately
dealt with the structure of the text alone...In
this sense, pure formalistic structural analysis
is probably every bit as sterile as motif-hunting
and word-counting [two methods Propp criticizesI
...Clearly, structural analysis is not an end
in itself! Rather it is a beginning, not an end.
It is a powerful technique of descriptive ethno-
graphy inasmuch as it lays bare the essential
form of the folklorisitic text. But the form
must ultimately be related to the culture or
cultures in which it is found.30
Neither Propp nor Dundes could adopt Barthes's onion metaphor.
For them the text still has a center. Perhaps their way of
looking at the text can be compared to the way we look at
an artichoke (if we must choose a species of fruit): the
text has layers that must be peeled away, but when the
analyst has peeled them away, he finds there not nothing,
he finds the heart.
This difference in approaches is a manifestation, I
think, of a difference in critical attitudes which Robert
Scholes has described as the difference between "high
structuralism" and "low structuralism." Barthes is a "high
structuralist"; Propp is a "low structuralist." "Barthes,"
along with Levi-Straus, Michel Foucault, Jacques
Lacan, and Jacques Derrlda is a star performer,
an individual who must be approached as a system
in himself, and understood for the sake of his own
mental processes. Whatever contributions of these
men are absorbed by the general culture, their
texts will not suffer the same absorption but will
remain--like philosophical texts, which they are,
and literary texts, which some of them aspire to
be--as unique objects to which later thinkers must
return in order to grasp the ideas and methods
that have developed therein...
There is also a "low structuralism," practiced
by men...whose aspirations are more humble...The
low structuralist writes to be immediately useful,
to be ultimately superseded.31
Propp's attitude, the attitude of the "low structuralist,"
is the attitude I have adopted here. Propp considers himself
one worker in a community of workers; the points he makes
about the literature he examines are intended to help others
in their own investigations of that literature. Barthes be-
lieves that in analyzing the text he, in a sense--a very
complex sense--creates the meaning of the work. Propp believes
that in analyzing the text he can help other workers in the
community of workers to go on to further analyze the text.
The study that follows will be a success if it helps some
reader go on to further analyze the genre of the Quixotic
1. I have consulted Current Literary Terms, ed. A. F.
Scott (London: Macmillan, 195), A Dictionary Of European
Literature, ed. Laurie Magnus (London: Routledge, 1926),
A Dictionary of Literary, Dramatic, and Cinematic Terms,
ed. Sylvan Burnett, Morton Berman, and W'iliam Burtcn (Boston:
Little, Brown, & Company, 1971), A Dictionary Of Literary Terms,
ed. Harry Shaw (New York: McGraw-Hill, 1972), A Dictionary
Of Modern Critical Terms, ed. Roger Fowler (London: Routledge
and Kegan Paul, 1973), A Dictionary Of World Literary Terms
ed. Joseph T. Shipley (Boston: The Writer, 1970), Ensayo de
un Diccionario de la Literatura, ed. Federico Carolos Sainez
de Robles (Madrid: Aguilar, 1954), and The Oxford Companione
To English Literature, ed. Sir Paul Harvey (3rd ed.; Oxford:
2. In an essay entitled "The Structuralist Activity,
Critical Essays, trans. Richard Howard (Evanston, Ill.: North-
western University Press, 1972), Barthes observes, "The struc-
turalist activity involves two typical operations: dissection
and articulation. To dissect...is to find in the object
certain mobile fragments whose differential situation engenders
a certain meaning; the fragment has no meaning in itself, but
it is nonetheless such that the slightest variation wrought
in its configuration produces a change in the whole..." p. 216.
3. Manuel Duran, Cervantes (New York: Twayne, 1974),
4. Roland Barthes, "An Introduction To The Structural
Analysis Of Narrative," trans. Lionel Duisit, New Literary
History (Winter, 1975: VI, 2), p. 238.
5. Vladimir Propp, Morphology Of The Folktale, trans.
Laurence Scott (2nd ed., Austin, Texas: University of Texas
Press, 1968), pp. 89-90.
6. Robert Scholes, Structuralism In Literature, New
Haven: Yale University Press, 1974), p. 59.
7. Propp, pp. 149-155.
8. William 0. Hendricks, "The Structural Study Of
Narration: Sample Analyses," Poetics //3 (The Hague: Mouton,
1972), p. 109.
9. Harry Levin, "The Quixotic Principle," The Interpre-
tation Of Narrative: Theory And Practice, ed. Morton W. Bloom-
field (Cambridge, Mass.: Harvard University Press, 1970), pp. 45-66.
10. Leo Spitzer, "On The Significance Of Don Quixote,"
Cervantes: A Collection Of Critical Essays, ed. Lowry Nelson,
Jr. (Englewood Cliffs, N. J.: Prentice-Hall, 1967), pp. 82-97.
11. Scholes, cf. note #6.
12. Jonathan Culler, Structuralist Poetics: Structuralism,
Linguistics, And The Study Of Literature (Ithaca, N. Y.: Cornell
University Press, 1975).
13. Barthes, "An Introduction To The Structural Analysis
Of Narrative," p. 238.
14. Propp, pp.23-24
15. Ibid., p. 21.
16. Barthes, S/Z, trans. Richard Miller (New York: Hill
And Wang, 1974), p. 21.
17. Barthes, "An Introduction To The Structural Analysis
Of Narrative," p. 238.
18. Propp, p. 92.
19. Tzvetan Todorov, The Fantastic: A Structural Approach
To A Literary Genre, trans. Richard Howard (Cleveland: The
Press of Case Western Reserve University, 1973), p. 163.
20. Barthes, S/Z, p. 19.
21. Ibid., p. 18.
22. Ibid., pp. 51-52.
23. Propp, p. 89-90.
24. L. C. Knights, "How Many Children Had Lady Macbeth?,"
Explorations (New York: George W. Stewart, 1947), p. 30.
26. Barthes, S/Z, p. 134.
27. Ibid., p. 135
28. Barthes, "Style And Its Image," Literary Style: A
Symposium, ed. Seymour Chatman (London: Oxford University
Press, 1971) p. 10.
29. Propp, p. 16.
30. Alan Dundes "Introduction" in Propp, pp. xil-xili.
31. Scholes, pp. 157-158.
DON QUIXOTE IN LITERARY HISTORY:
A SURVEY OF INTERPRETATIONS
There are, broadly speaking, two kinds of theories
explaining Don Quixote's place in literary history. First,
there are those theories which develop from the assumption
that narrative has evolved, and is evolving, in time and
argue that Don Quixote can be placed somewhere on a con-
tinuum describing that evolution. Critics who use this
diachronic approach usually try to note how Cervantes
adopted and improved the narrative techniques of the authors
before him or how authors since his time have adopted and
improved Cervantes's techniques. Secondly, there are those
theories which develop from a more or less ahistorical point
of view. Critics who use this synchronic approach usually
argue that Don Quixote is a timeless model of narrative suc-
cess which authors have attempted, and will continue to
attempt, to imitate. Critics who believe in the first approach
tend to look upon Cervantes's achievement as one step in the
development of narrative; critics who believe in the second
approach tend to look upon Cervantes's achievement as an un-
surpassed ideal or a continuously viable model.
In this chapter I will attempt to detail a few of the
examples of each critical approach. In the process I hope to
synopsize a few of the most interesting analyses of Cervantes's
place in literary history, to give some idea of the variety of
interpretations of Don Quixote, and to note some limitations
of some of these interpretations.
The two categories of critical approaches are, I should
say again, broad. It goes without saying that many of the
critics themselves would object to being placed in one group
or the other; of course, none of them is under any obligation
to stay within any boundaries I mark. I have placed them in
the two groups below according to how I understand the general
direction of their arguments. Although Auerbach observes that
Don Quixote is a unique narrative whose "noncritical and non-
problematic" presentation of everyday reality Auerbach "cannot
imagine" ever being attempted by any other author, Auerbach
belongs in the first group. The idea that narrative is evolv-
ing, that there is a "trend toward reality" in fiction, is
an all-important assumption of Auerbach's Mimesis. The case
of Salvador de Madariaga is an opposite example. Although he
takes the time in Don Quixote: An Introductory Essay In
Psychology to note the state of narrative in Spain in the
seventeenth century, Madariaga seems to me more interested in
a ahistorical reading of the novel.
Despite the lack of precise boundaries, there are some
real differences between the two approaches. The diachronic
theories are much the harder to synopsize. It seems that
every critic who believes literature is evolving has his own
special history of that evolution to present. There among
such readers one finds little agreement about Cervantes's
novel. Each critic, looking for ways to justify his own
version of literary history, reads Don Quixote differently.
In Section One below I have attempted to synopsize three of
these histories--Erich Kahler's in The Inward Turn Of Narra-
tive (1957-59, trans. 1973), Michel Foucault's in Madness
And Civilization (1961, trans. 1965), and Auerbach's in
Mimesis (1946, trans. 1953). The rather detailed description
of the three analyses is necessary to point to the differences
among them and to show how variable understandings of Don
Quixote can be. The three theories make a neat triskelion.
Kahler believes Cervantes is the first to write "the modern
novel, the first to use "ascending symbolism"; he argues
that Cervantes maps out new narrative territory in Don
Quixote. Foucault, by contrast, argues that Cervantes was
one of the last practitioners (with Shakespeare) of an old
form of narrative, one that was out of date by the seventeenth
century; he focuses on Cervantes's use of madness, arguing
that the madness in Don Quixote is in "an ultimate state...
beyond appeal" and that this use of madness is unlike the
modern narrative's in which madness is a "knot" rather than
a "denoument." Auerbach's position contradicts each of
the other two. Where for Kahler Don Quixote represents a
new species and for Foucault a dying species, for Auerbach
Don Quixote is a fascinating, but ultimately unimportant
freakish species. One outside the main line of evolution:
So universal and multilayered, so noncritical
and nonproblematic a gaiety in the portrayal
of everyday reality has not been attempted
again in European letters.5
Section Two below describes some of the theories of
the second kind. The critics who adopt ahistorical readings
of the novel are in close agreement about the nature of
Cervantes's achievement. Although there is some difference
about the precise details, most of these readers (including
Leo Spitzer, Harry Levin, Salvador de Madariaga, and Lionel
Trilling) agree that what Cervantes manages in Don Quixote
is a masterful unity of disunities. Most of these readers
find that the novel is a structure generated by a duality.
One calls this essence the duality of "disenchantment" and
"illusion," another the duality of "words" and "deeds,"
a third calls it "the duality of heroes," but all are
describing, I think, the same essence: Don Quixote has at
its heart a pairing of opposite forces.
Section Three below is given to an examination of
Rene Girard's Deceit, Desire, And The Novel, a history of
the Quixotic Novel, which is an original and provocative
blend of the synchronic and diachronic approaches.
1. Three Diachronic Analyses
For Erich Kahler it is how Don Quixote means rather
than what it means that makes it a seminal work. In his
The Inward Turn Of Narrative Kahler calls Don Quixote "in
a sense...the first modern novel" because, "For the
first time the individual case is more than an individual
case; the story of an individual bears a superindividual
significance." Kahler does not present a close reading
of the novel, but he does present an elaborate and inter-
esting analysis of the history of narrative and show where
Don Quixote exists as a turning point and seminal influence.
Kahler traces in this study of Western narrative from
the Iliad to Tristram Shandy a development of human con-
sciousness. He argues that, despite the assumptions of
some thinkers, reality has not always been the same for all
men throughout history. Man's understanding of reality is
constantly evolving because of the constantly changing
relationship between his mind, his inner space, and "objec-
tive reality," his outer space. Art does not just allow
us to see reality differently (as Auerbach argues in Mimesis)
it changes reality. Throughout history, art, like other
interactions between mind and reality, has enabled man to
stretch his consciousness so that there has been a "pro-
gressive internalization of events" as man's understanding
of inner and outer spaces grow. In the twentieth century
the enormous expansion of physics and psychology are examples
of this evolution. But in The Inward Turn Of Narrative
Kahler is interested in demonstrating how narrative has
been both a tool and a measure of the evolution of con-
Kahler finds three dimensions in "the early stages in
the process of the internalization of narrative." The
first dimension is first explored in tales like Boccaccio's
falcon novella (V, 9) in the Decameron (1348-1353) and
Madame de La Fayette's La Princesse de Cleves (1678),
psychological expressiveness is no longer a
matter of scattered passages; it pervades
and embraces the entire narrative. l narration
has become a psychological process.
These stories are infused with psychology; in fact, Kahler
argues in a brief commentary on the falcon novella, the
story takes place almost entirely on a psychological plane:
the story develops and turns as the reader understands the
mental state of the impoverished knight who sacrifices his
last possession to the woman he loves and the feelings of
the woman who loses her son and takes the knight as a lover.
(It goes without saying that much of Don Quixote takes place
entirely on a psychological plane: "The Cave Of Montesinos"
would be the best example.)
The second dimension early narrative maps, according to
Kahler's analysis, is "the breadth of the realm of the
senses." Kahler finds this dimension first appears in the
narratives of Rabelais, Grimmelshausen, and Cervantes. In
their satires "the independent world of nature was established
outside the human sphere" and "Nature and society can no longer
be taken for granted and regarded as a unity." In these stories
the milieu becomes an entity and so becomes separated from man
and consciousness. Man becomes conscious--these narratives make
him conscious--of the world out there; a distance and dichotomy
is created: the narrative recognizes the existence of inner
and outer spaces. In Don Quixote this dichotomy is an obvious
one; indeed, the dichotomy between Don Quixote's inner space
(the chivalric ideals and his madness) and his outer space (the
real world) is, of course, the crux of the narrative. Kahler
The open confrontation between a romantic imagination
and Sancho Panza's rough-and-ready reality of peasant
girls, barbers, and muleteers produces that double
irony which is directed as much against the deluded
knight as a ainst the victorious but inferior world
The irony in the novel is an effect of the separation of
Don Quixote's inner space from his objective reality. In separat-
ing these two Cervantes is able to satirize them both: he can
mock the knight for seeing mills where there are only giants and
he can mock the knight's world for failing to live up to his
expectations of it (as in the many scenes where Sancho, for
example, demonstrates that his appetites are more important
than his ideals).
But it is in the exploration and exploitation of the
third dimension of early narrative that Cervantes leads the
way with his masterpiece. Don Quixote participates with
other narratives in the mapping of "the depth of the psyche"
and "the breadth of the realm of the senses"; it is, how-
ever, the first to explore the third dimension--"the height
of ascending symbolism."
This last part of Kahler's theory of the development of
"early narrative" is the most intriguing for the student of
Cervantine criticism. Kahler believes Don Quixote began an
entirely new kind of story, narrative he calls "complete
fiction." "Complete fiction" is characterized by the use
of modern symbolism, "ascending symbolism."
This new symbolism is ascending because it
proceeds not from a supernatural, extrahuman,
or prehuman event whose reality is assumed,
but because it rises from below, from a purely
human natural world, from individual characters
and events which Crom the outset possess only a
representative, not an actual, reality. In fact,
they have been invented by the artist for the
sake of this representation.17
According to Kahler, Cervantes's narrative world is quali-
tatively different from Dante's, for example, because it
(Don Quixote) is a total creation of a man's imagination.
The character of Don Quixote, the mad knight, is an essence
from which meaning develops upward, not an image for which
meaning has been assigned from above. Dante uses symbolism
in The Divine Comedy but it is the old symbolism (older than
allegory even), "descending symbolism." In "descending
symbolism" the meaning of a symbol comes from a reality
greater than the symbol. In Don Quixote and other works of
"complete fiction" the symbol and its meaning are more
intricately intertwined. Kahler observes:
Dante's symbolic procedure consists in fitting
the variegated patterns of life on this earth
to the preexistent divine cosmos, in deriving
earthly life from and leading it to that cosmos.
But to the creators of the new symbolism there
is no preexistence, no premise outside the work
itself. The whole symbolic structure is built
up by the artist; it is entirely integrated.
This complete integration is internalization.
(emphasis in original)ib
So, if Kahler is right, Cervantes widens the boundaries,
and so loosens the bonds, of narrative with his masterpiece.
The internalization is, like all the internalizations Kahler
describes, an expansion and also a liberation. In older
narratives, those using "descending symbolism," the story-
teller positions himself between God and man: the story-
teller acts as priest, explaining God's ways to man, putting
revealed truths into other words. The story-teller before
Cervantes stands at a mid-point in the transmission of meaning.
Cervantes stands himself at an end-point in the transmission.
Cervantes is the first to make himself a god in his narrative
world. After Cervantes the story-teller, using "ascending
symbolism," is no longer a message-carrier; he is instead
Kahler's description of exactly how the text of Don
Quixote is its own reality is very sketchy. He makes a few
brief remarks about the conflict between Quixote's chivalric
code and Sancho's bourgeois appetites, but he passes quickly
on. In a study as ambitious and large as The Inward Turn Of
Narrative, of course, there is little time for close textual
explication. Nevertheless, Kahler touches upon issues that
the student of Cervantine criticism is interested in pur-
suing. Although Kahler implies that he is content to discuss
only how Don Quixote means and not what it means, his analysis
leads in only one direction: Kahler would have to take the
side of the "soft" readers of the novel, which is to say
he would have to argue that there is value in Don Quixote's
The crux of the "hard" reader's interpretation is that
Cervantes is only a message-carrier. The hard reader argues,
as Auerbach does, for example, that Don Quixote is only a
Fool, the Fool of old tales, a character who needs only to
regain the usual sense of the world, regain knowledge of
revealed truth, and everything will be all right. The
message of Don Quixote is to the hard reader a message found
in much older narratives: All is right with God's world;
Man must only find his place in it. For the hard reader,
Quixote's deathbed conversion is an essential movement, the
movement of the sinner back to revealed truths, the movement
of the madman back to the Christian way.
To maintain his theory of "internalization" and the
liberation of narrative, Kahler would have to dismiss, in the
style of "soft" readers, the deathbed conversion. If Cervantes
uses a truly different symbolism than Dante does, then his
meaning must be truly different from Dante's meaning. The
hard reader's interpretation is that Cervantes does (in his
final chapter) exactly what Kahler says Dante does in The
Divine Comedy: he leads earthly life to fit into "the pre-
existent divine cosmos." If he did take the time to make
a close textual explication Kahler would have to argue, like
Unamuno, that the deathbed conversion is undercut by the
rest of the book, that the "immortal Quixote" is the mad
Quixote; or, like Ortega y Gasset, Dostoyevsky, and Auden,
Kahler would have to argue that Cervantes presents in Don
Quixote a modern Christianity and that the deathbed conver-
sion is ironically undercut by the fact that the ideals
Quixote has espoused in his madness and rejected in the end
is true Christianity. (Ortega y Gasset calls Quixote a
"Gothic Christ." ) If Cervantes is truly an end-point in
the transmission of meaning, then his meaning must be a new
meaning. His meaning cannot be the meanings of any of the
narratives that came before the Quixote, because otherwise
Don Quixote is not "superindividualized" but only one more
This matter of "superindividualization" in the Quixote
is another aspect of Kahler's theory that the student of
Cervantine criticism would be interested in seeing Kahler
pursue. Unfortunately, Kahler has his critical sights set
higher. One can easily agree that the figure of Don Quixote
is a special individual especially in comparison with the
figures in the narratives most closely related to Don Quixote,
the heroes of the chivalric romances. It is not so easy to
explain exactly how this effect is achieved.
Michael Foucault's theories about Don Quixote make an
interesting contrast to Erich Kahler's. Where Kahler be-
lieves Don Quixote is the first example of a new kind of
narrative, a kind of narrative that we can call "modern,"
Foucault believes Don Quixote is one of the last examples
of a kind of narrative that died near the turn of the
seventeenth century. He argues that Cervantes's master-
piece is a moral satire about madness which reflects an
understanding of madness that has disappeared in Western
In Madness And Civilization: A History Of Insanity In
The Age Of Reason Foucault presents an analysis of Western
society's understanding of madness and traces that under-
standing through its changes in the sixteenth, seventeenth,
and eighteenth centuries. His history is very useful to
the student of Cervantine criticism. Foucault's reading of
Don Quixote's madness, in particular, is intr-iguing.
According to Foucault, Cervantes with Don Quixote
presents a medieval man's moral understanding of insanity.
In works like Don Quixote,
Madness deals not so much with the truth and
the world, as with man and whatever truth about
himself he is able to perceive. It thus gives
access to a completely moral universe. Evil
is not punishment in the end of time, but only
fault and flaw.21
Foucault describes four kinds of madness that appear
in the narratives up to Cervantes's time. In each case
madness is a fault, a moral illness, which cripples the
victim. There is, first of all, madness by romantic fix-
ation. "Its features," Foucault explains, "have been fixed
once and for all by Cervantes."
In appearance in Don Quixote and close imitations
there is nothing but the simple-minded critique
of novels of fantasy, but just under the surface
lies an enormous anxiety concerning the relation-
ships, in a work of art, between the real and the
The second kind of madness is madness of vain presumption
in which "it is not with a literary model that the madman
identifies; it is with himself...Poor, he is rich; ugly, he
admires himself; with chains on his feet, he takes himself
for God." Foucault names other literary characters as
exemplars of this form of madness (e.g., Chateaufort in
Le Pedant joue, M. de Richesource in Sir Politik), but, of
course, Don Quixote exhibits this second form of madness
also. He is especially vain in the early part of the book--
in the incident (1,4) when he "saves" Andres from the
whipping, for example.
Foucault's third kind of madness, the madness of just
punishment, does not appear in Don Quixote. This is the
madness of the victim driven insane by the horror of the
crimes he has committed and by the pangs of his conscience.
Lady Macbeth's madness, according to Foucault, is the mad-
ness of just punishment. Shakespeare also has a moral under-
standing of madness.
The fourth kind of madness is, it seems to me, present
off-stage in Don Quixote. This is the madness of desperate
passion: "Love disappointed in its excess, and especially
love deceived by the fatality of death, has no other re-
course but madness." Ophelia's last song and King Lear's
final madness are Foucault's examples, but this desperate
passion exists as a possibility standing just outside Cer-
vantes's narrative. This is the madness of Beltenebros
which Don Quixote tries to imitate in the mountains of the
Sierra Morena. Of course, it is used as a joke here because
Don Quixote has not lost his lover, as Sancho points out in
one of the best exchanges in the novel:
"It strikes me," said Sancho, "that those
knights had...some cause for such foolish
penances, but what reason has your Grace for
going mad,...what signs have you found that lead
you to think the lady Dulcinea has been up to
"That," said Don Quixote, "is the point of
the thing; that is the beautiful part of it.
What thanks does a knight-errant deserve for
going mad when he has good cause? The thing is
to go out of my head without any occasion for
it, letting my lady see, if I do this for her
in the dry, what I would do in the wet." (p. 199)25
But the humorous point is made at the expense of Don Quixote's
twisted logic, not at the expense of desperate passion. It
is Don Quixote who is undercut by the joke, not the madness
of the desperate, disappointed lover. That might be too fine
a distinction to make an argument from, but the madness of
desperate passion is a presence even closer to the narrative
line in other places. In several interpolated tales the
madness of desperate passion exists as a route narrowly
avoided. An example in Part I is "The Curious Impertinent"
where Anselmo comes close to going insane when he discovers
that his wife, Camila, and best friend, Lothario, have left
him (I, 35). An example in Part II is the "Roque Guinart"
episode where Claudia Jeronima also comes close to losing
her mind when she discovers she has fatally wounded her
innocent lover (II, 60). Both characters avoid insanity,
but must give up their lives as the price. Anselmo dies
as he pens a note of apology and forgiveness to Camila (in
what is obviously and, perhaps, intentionally a fore-
shadowing of Quixote's deathbed drawing up of his will).
Claudia dies figuratively as she takes herself to a nunnery
to live out her days in isolation. In both cases the mad-
ness of desperate passion is no joking matter; it is a real
possibility. Both narratives end as the characters veer
away from that madness to literal (Anselmo's) death and
figurative (Claudia's) death.
These two narratives fit neatly into Foucault's theory
(although he does not mention them). Foucault believes Cer-
vantes and Shakespeare are characterized by the way they
position madness in their stories. Madness always comes at
the end of the line of Cervantes's and Shakespeare's tales,
according to Foucault:
In Shakespeare or Cervantes, madness still
occupies an extreme place, in that it is beyond
appeal. Nothing ever restores it to truth or to
reason. It leads only to laceration and thence
to death... Madness] is already the plentitude
of death; a madness that has no need of a physi-
cian but only of divine mercy.26
Madness to these two writers, because they understand it
in the way of the sixteenth century, is an ultimate state.
It exists on-stage (in Lear) and off-stage (in "The Curious
Impertinent") as an end: "With such overwhelming news as
this that Lothario and Camila had run off together Anselmo
was not only near to losing his mind but to ending his life
as well," (p. 321). In later narratives the change of the
understanding of madness is represented by a displacement
of its place in the narrative line. After the sixteenth
century, Foucault notes,
madness leaves these ultimate regions where
Cervantes and Shakespeare had situated it; and
in the literature of the early seventeenth cen-
tury it occupies, by preference, a median place;
it thus constitutes the knot more than the
denouement, the peripity sic] rather than the
Madness moves forward in later narratives, Foucault maintains;
it becomes a stage for the hero to pass through, a problem
for the hero to solve.
But, there is an obvious question: What about the main
line of narrative of Don Quixote? Isn't Don Quixote's madness
a knot rather than a denouement? Isn't his insanity, in fact,
the problem that creates the whole story?
To keep his argument Foucault must adopt a peculiar
"hard" reading of the novel. He focuses, as critical readers
must, on the final chapter. It seems, of course, that this
chapter would disprove Foucault's thesis as regards Cervantes,
because it is in this chapter that Quixote passes through
madness; in his will-writing he demonstrates that reason can
be restored, that his madness is not "beyond recall." "Cer-
tainly," Foucault concedes,
Don Quixote's death occurs in a peaceful land-
scape, which at the last moment has rejoined
reason and truth. Suddenly the Knight's madness
has grown conscious of itself, and in his own
eyes trickles out in nonsense. But is this
sudden wisdom of his folly anything but "a new
madness that had just come into his head" [a
rendering, apparently, of the line in the final
chapter describing the reactions of the barber,
the curate, and Sanson Carrasco to the news that
Quixote had declared his sanity: "Cuando esto
le oyeron decir los tres, creyeron, sin duda,
oue alguna nueva locura le habia tomado."' ?
The equivocation is endlessly reversible and
cannot be resolved, ultimately, excent by death
itself...But death itself does not bring peace;
madness will still triumph--a truth mockingly
eternal, beyond the end of a life which yet had
been delivered from madness by this very end.
Ironically, Don Quixote's insane life pursues
and immortalizes him only by his insanity; mad-
ness is still the imperishable life of death.28
What Foucault does in this passage is adopt the argument of
the softest of readers, Unamuno, to argue for a hard reading
of the novel. It is Unamuno who first points out that the
immortal Don Quixote is the mad Don Quixote, that the con-
version in the last chapter is undercut by the strength of
the earlier madness. Unamuno uses this argument to prove
that Quixote's madness is exemplary, to show that there is
salvation in making oneself ridiculous. Foucault uses the
same argument to prove that there can be no salvation from
madness, at least from the madness in Cervantes's narrative.
The endlessly reversible equivocation between madness and
sanity insures that there can be no escape from madness be-
cause there is no way to declare sanity.
Foucault must be considered a hard reader, because he
does not believe there is value in Don Quixote's madness.
Don Quixote does not learn from his insanity or develop
because of it or represent an ideal to be imitated; but
Foucault's hard reading is interestingly unlike the inter-
pretation of the most classic of hard readers, Erich Auerbach.
Auerbach, too, compares Cervantes and Shakespeare, but
he draws a line of distinction between them where Foucault
ties them together. Auerbach believes the ways the two
authors handle the theme of madness points to an essential
difference in their works. He is in agreement with Foucault
that the madness in Shakespeare (in the case of Hamlet, at
least,) is beyond appeal but he argues that this understanding
of madness is the modern one and he sees the madness of Don
Quixote as an entirely different malady from that which
plagues the Dane.
In his well-known reading of Don Quixote in Mimesis:
The Representation of Reality in Western Literature Auerbach
makes a point of contrasting Hamlet and Don Quixote. He is
interested in charting "the trend toward reality" in Western
literature. Hamlet, he argues, is an important part of that
trend; Don Quixote is not. Cervantes's novel is too much
like the other Spanish literature of the siglo de oro in that
it participates more in the "fanciful, adventurous, and
theatrical" than in "the history of the literary conquest of
modern reality." In that history, Auerbach writes, "the
literature of Spain's great century is not particularly im-
portant--much less so than Shakespeare and even Dante,
Rabelais, or Montaigne." One of Auerbach's postulates is
that the "trend toward reality" has much to do with tragic
complications and "the problematical." Although he thinks
Cervantes of the Spanish authors of the golden century "is
certainly the one whose characters come nearest to being
problematic," the problems of these characters are not
serious or tragic enough to make Cervantes as important an
author as Shakespeare (in "the history of the literary con-
quest of reality").
"If we want to understand the difference between
Cervantes and Shakespeare ," he writes,
we need only compare'the bewildered, easily
interpreted, and ultimately curable madness
of Don Quixote with Hamlet's fundamental
and many-faceted insanity which can never
be cured in this world.31
Auerbach's argument is thus a reversal of Foucault's inter-
pretation. According to Auerbach it is the madness that is
used as a denouement that is modern; when madness is only a
knot it may not be sufficiently problematic. This is the
case of Don Quixote: his insanity is too simple. Quixote's
madness is used only to generate an entertaining tale; it is
not the madness that generates a disturbing vision of the world.
"Don Quixote alone is wrong," Auerbach observes,
as long as he is mad. He alone is wrong in a
well-ordered world in which everybody else has
his right place. He himself comes to see this
in the end when, dying, he finds his way back to
the order of the world. But is it true that the
world is well-ordered? The question is never
Auerbach focuses his attention upon "The Enchanted Dul-
cinea" episode (II, 10). He points out that the situation
which Cervantes invents could have been a tragic one: there
is illusion and disillusionment in the scene; there is the
one confrontation between Quixote and the woman he believes
is his Dulcinea; there is the manipulation of the master by
the servant; there is the denigration of Quixote's ideals
(as he is tricked into kneeling in dedication to an ugly
peasant girl); and, lastly, there is the psychological dis-
turbance Don Quixote suffers. But, Auerbach points out,
Cervantes does not play the scene for its tragedy; instead
Cervantes makes it into a farce. The episode characterizes
the book for Auerbach (and the fact that the novel does not
participate in the history of the trend toward reality)
because Cervantes does not use the opportunity to investigate
any riddles or pose any problems. Instead Cervantes gives his
hero a way to avoid confronting his madness:
It might be supposed that the incident of
the "Enchanted Dulcinea" would bring on a terrible
crisis...It could produce a shock which in turn
could bring on much deeper insanity. But there
is also the possibility that the shock might bring
about a cure, instanteous liberation from the idee
fixe. Neither of these things happens. Don
Quixote surmounts the shock. In his idee fixe
itself he finds a solution which prevents him both
from falling into despair and from recovering
his sanity: Dulcinea is enchanted. This solution
appears each time the exterior situation estab-
lishes itself as in insuperable contrast to the
illusion...The happy ending is a foregone conclusion.
Thus both tragedy and cure are circumvented.33
Don Quixote's ever-available escape clause, that an enchanter
has interfered in any situation in which Quixote's understand-
ing of the world does not match reality keeps the novel from
becoming a representation of the tragic complications of life.
It enables Cervantes to make a mockery of Quixote whenever his
mad hero makes a serious confrontation with reality. Cer-
vantes thus avoids the problematic.
Auerbach does not believe Cervantes was ever interested
in the metaphysical arguments and riddles other readers have
found in the novel:
For centuries--and especially since the romantics--
many things have been read into [Don Quixote] which
Cervantes hardly foreboded, let alone intended.
Such transforming and transcendent interpretations
are often fertile. A book like Don Quixote disso-
ciates itself from its author's intention and leads
a life of its own.3'
Nevertheless, Auerbach believes the novel was intended to be
a honest entertainment and not a metaphysical (or political)
investigation. "I take it as merry play on many levels," he
writes. "This means that no matter how painstakingly I
have tried to do as little interpreting as possible, I yet
cannot help feeling that my thoughts about the book often go
far beyond Cervantes's aesthetic intention."
It is the improvisational quality of the novel that
leads Auerbach to believe he reads more into it than Cervantes
put there. "Whatever the intention may have been," he argues,
"...it most certainly did not consciously and from the begin-
ning propose to create a relationship like that between Don
Quixote and Sancho Panza as we see it after having read the
novel." The relationship between master and servant
developed naturally, organically, as Cervantes improvised his
tale. lie had no plan, Auerbach believes; he depended for his
success on his ability to tell stories:
the two figures were first a single vision,
and what finally developed from them--
singly and together--arose gradually, as
the result of hundreds of individual ideas,
as the result of hundreds of situations in
which Cervantes puts them and to which they
react on the spur of the moment, as the re-
sult of the inexhaustible, ever-fresh power
of the poetic imagination.37
Evidence of the improvisational nature of the text can be
found in the narrative errors:
Now and again there are actual incongruities
and contradictions, not only in matters of
fact (which have often been noted (the theft
of the ass being the best-known] ) but also
in psychology: developments which do not fit
into the total picture of the two heroes--
which indicates how much Cervantes allowed
himself to be guided by the momentary
Auerbach ends his chapter on Don Quixote with a dis-
cussion of the question his hard reading naturally raises:
What did Cervantes contribute to literary history? If it is
not the thematic content or metaphysical dimension of Don
Quixote, what is it that Cervantes gave to the authors who
followed him and read his masterpiece?
Auerbach seems intent on letting some of the air out
of what he considers Cervantes's romantically-inflated
reputation, (perhaps in a response to Unamuno) but he does
not mean to deflate it entirely. Auerbach believes Cervan-
tes did make his mark in literary history--through his
capacity for rendering "everyday reality." The talent was
so great, Auerbach writes, "that almost everything realistic
written before him appears limited, conventional, or propa-
gandistic by comparison." Auerbach admires Cervantes most
of all because he is a superb story-teller. He believes
Cervantes was attracted to the theme of a man deluded by
books not for its thematic possibilities (it is Cervantes's
readers who have realized these) but for the opportunities
the theme created for narrative possibilities, "the mixture
of the fanciful and everyday elements in the subject, its
malleability, elasticity, adaptability." Cervantes recog-
nized the chances the theme would give him to improvise
tales in a way congenial with his nature. And what is that
nature? Cervantes's nature is his quintessence and it is a
quintessence very hard to describe. The "peculiarly Cer-
vantean" is, Auerbach decides,
not a philosophy; it is no didactic purpose;
it is not even a being stirred by the un-
certainty of human existence or by the power
of destiny, as in the case of Montaigne and
Shakespeare. It is an attitude--an attitude
toward the world, and hence also toward the
subject matter of his art--in which bravery
and equanimity play a major part. Together
with the delight he takes in the multifar-
iousness of his sensory play there is in him
a certain Southern reticence and pride. This
prevents him from taking the play very seriously.
He looks at it; he shapes it; he finds it
diverting; it is also intended to afford the
reader refined intellectual diversion. 1
But Cervantes does not (can not) pass on his nature
to literary history. Cervantes is, in Auerbach's analysis,
not a seminal force; he is a unique writer, but one whose
essence is created by his particular age and peculiar
talents and who cannot be imitated. Auerbach observes, "So
universal and multilayered, so noncritical and nonproble-
matic a gaiety in the portrayal of everyday reality has not
been attempted again in European letters. I cannot imagine
where and when it night have been attempted."
There are, of course, many other diachronic analyses of
fiction which place Don Quixote on a continuum describing the
evolution of narrative. Among the best known of general
theories are The Theory Of The Novel by George Lukacs and
The Nature Of Narrative by Robert Scholes and Robert Kellogg.
And there are, too, many specific investigations of Don
Quixote as a source and influence; a good recent example of
the influence of Cervantes's on English fiction may be
cited: Susan Staves's "Don Quixote In Eighteenth-Century
England" (Comparative Literature 24, Summer, 1972, pp. 193-215).
As my intention is to examine the broad opposition between
diachronic and synchronic emphases, however, I shall now turn
to a consideration of certain synchronic theories.
2. Synchronic Analyses
Generally, the critics who adopt the second kind of
understanding of Don Quixote's place in literary history--
that Don Quixote is an archetypal model or unsurpassed
ideal--agree that what Cervantes managed in his novel was a
unity of disunities. The various theorists describe the
disunities in different ways, but all agree that the linking
of opposite, volatile forces is a masterful achievement.
Cervantes's ability to hold the forces together, many argue,
is the distinctive feature of his genius.
Although there are elements of historicism in his
reading, Leo Spitzer is a synchronic theorist in the sense
that he is interested in describing how the novel is a model
of narrative. In an essay entitled "On The Significance of
Don Quixote" he argues that Cervantes is unique in literary
history because he manages "to hold before our eyes a cosmos
split in two separate halves, disenchantment and illusion,
which, as by a miracle, do not fall apart." Spitzer calls
the theme of the man deluded by his reading "the problem of
the book" and points out that Cervantes was the first writer
to understand the dimensions of the problem brought about
by the printing press, a problem, Snitzer believes, of per-
ception. Before the printing press cultural values were
communicated through the ear, "the musical, religious,
communal sense"; after the printing press made reading so
important cultural values came through the eye, "the rational,
analytic, and individualistic sense" and so "there was born
the peril of wrong application of literature in life by in-
dividuals reading alone, severed from society."
Cervantes's original intention, then, was a critical
one: he wanted to expose "the problem of the book," to
describe how a reading man could go wrong reading obsolete
books. Cervantes wanted his novel to be a counter-novel.
However, his sensibility would not let him be satisfied with
such a one-sided presentation. "Cervantes anticipated,"
Spitzer believes, "the feeling of disharmony and incomplete-
ness which would be produced in the reader by an anti-novel
in pure form." Cervantes's sensibility led him to establish
an almost-impossible and never-duplicated balance, "an
equilibration of the critical sense by the beauty of the
fabulous." The interpolated tales are the most obvious
examples of Cervantes's use of his non-critical, illusion-
making sense. So, on the one hand, with his critical sense,
Cervantes is subverting the romances and, on the other hand,
with his fabulist sense, he is re-issuing the romances with
his tales of Dorothea, the shepherdess Marcela, and all the
The whole of the Cervantine novel falls then
into two parts: the one teaches criticism
before imaginative beauty; the other re-
establishes imaginative beauty in the face of
all possible scepticism.46
Many authors have followed Cervantes's use of his critical
understanding of the world (Spitzer names Balzac, Maupassant,
Thackeray, Tolstoy, Faulkner, Mann, Gide, Conrad, Proust,
Joyce, Virginia Woolf, and Pirandello) in adopting a narra-
tive method that holds hard, cold reality up against the
illusion of the fabulous. Not one, however, has managed to
achieve the balance Cervantes achieved. Only Cervantes
simultaneously presents both visions of the world. The
unity is a unity of narrative voice; disenchantment and
illusion are linked in the narrator of Don Quixote:
the novel is a monument to the narrator qua
narrator, qua artist. For, although the
protagonists...seem to be Quixote, with his
continual misrepresentation of reality, and
Sancho, with his sceptical half-endorsement
of Quixotism, they are overshadowed by Cer-
vantes,...who combines a critical and illu-
sionistic act according to his free will. 7
In his essay "The Example of Cervantes" Harry Levin
adopts an approach to the problem of describing Cervantes's
influence on literature which is very close to Leo Spitzer's.
He observes, as Spitzer does, that Don Quixote "in terms of
its intention and impact constituted an overt act of criti-
cism" but that Cervantes recognized the limitations of
that criticism. For Levin, however, the dichotomy that
Cervantes recognized was that between artifice and reality
(rather than Spitzer's disenchantment and illusion). Levin's
Cervantes sees that criticism of bookishness was necessary,
that there is a fundamental difference between words and
deeds ("nalabras and hechos," as Levin puts it), but sees
also that there is no other way to make a criticism of books
than with a book. Levin's Cervantes faces a paradox because
...literary artifice is the only means the
writer has at his disposal. How else can
he convey his impression of life? Only by
discrediting those means, by repudiating
that air of bookishness in which any book
is inevitably wrapped.'9
Where Spitzer points to the interpolated tales as the key
to the understanding of Cervantes's sensibility (for it is
there that Cervantes admits the illusion of romance into
the text), Levin points to the devices of the self-conscious
narrator as the key (for it is there that Cervantes subverts
the reality of Don Quixote's story by noting artifice).
E. C. Riley is another reader who notes that Cervantes
was simultaneously critic and creator. In Cervantes's
Theory Of The Novel Riley attempts to reconstruct the
Spaniard's aesthetic theory, using as the most important
evidence the various statements about literature and life
mouthed by various characters throughout Don Quixote. lie
concludes that Cervantes felt "a deep-rooted ambivalence
which is fundamental to his writing." Cervantes found in
the ironic mode a vehicle to express his fundamental ambiva-
lence. "Cervantes discovered in irony the novelist's most
important tool," Riley argues,
the multiplicity of possible perspectives makes
a fresh and complex view of things possible, a
view more nearly round. It does not pinpoint
the truth of the matter, but circumscribes the
operative area. Irony allows Cervantes to criti-
cize while he writes, and to present different
points of view with considerable impartiality.51
Lowry Nelson has also pointed out that irony is the glue
that enables Cervantes to hold his unity of disunities together.
Nelson lists Cervantes with Shakespeare, Chaucer, Chekhov,
Kafka, and Svevo in the "select company" of "universal
ironists." These are writers who "contemplate the world with
a kind of gentle resignation and compassion in full knowledge
of both the grandeurs and miseries of human life." They are
different from the "tendentious ironists" like Flaubert,
Ibsen, Hardy, and Mann who "view the world from a programma-
tic stance connoting accusation, bitter protest, and meliorist
reformation of human ills."
Nelson points out that the narrative formula that Cer-
vantes discovers "cannot be stated without the great risk of
oversimplifying or overcomplicating it." Like many other
critics, Nelson believes the formula can be described, however,
as a complex of dualities: "...Cervantes managed to create
the central formula of the modern realistic novel in terms of
a vast range of style accommodating illusion and reality,
aspiration and actuality, the dogmatic and the problematic."
Actually, the citation of dualities is one of the most
common features of Cervantine criticism. Spitzer's "disen-
chantment and illusion" and Levin's "artifice and reality" are
only two pairings in what is a long list of pairings critics
have found to lie at the heart of Cervantes's narrative formula.
Nelson adds the three above. Lionel Trilling believes, "There
are two modes of thought in Don Quixote, two different and
opposed notions of reality." He argues that Cervantes
"changed horses in mid-stream" in writing Don Quixote: at the
start he was trying to side with the concrete view of reality,
that "the powerful immediacy of hunger, cold, and pain" out-
weigh in importance of "the past, the future, and all ideas";
after a time, though, Cervantes switched sides, "to show that
the world of tangible reality is not the real reality after
all. The real reality is rather the wildly conceiving, wildly
fantasizing mind of the Don." Robert M. Adams believes the
fundamental duality generating the narrative of Don Quixote
is the duality of "imagination" and "appetite." The plot of
the novel is an oscillation:
Don Quixote makes his forays into the world
of appetite and is brutally or comically
rebuffed; the world of appetite makes counter-
forays into Don Quixote's imagination and meets
with a resistance just as stout and determined.56
Frederich Karl believes the duality at the heart of Don Quixote
is the duality at the heart of all novels; all novels, he argues,
depend for their existence on the dualism in the adversary
relationship between subversive ideals and bourgeoise values.
Because Don Quixote is the first to present this adversary
relationship, it is the archetypal novel.
For whatever other interpretations we place
upon the Don, he is playing with the sub-
stantive existence of the novel...What the
Don suggests is the dualism of the novel:
imagination and chivalric knight-errantry re-
flect a radical distaste for an existing
society confronted by the dull, plodding 57
routine character of that inescapable society.
Helmut Hatzfield finds an entirely different kind of
duality in Don Quixote. He argues that the unity of disunities
in Cervantes's masterpiece is a triumph of prose style: Cervantes
holds together his dichotomous understanding and presentation
of the world by his narrative style which allows him to
"posit....poetical concepts before the eye on the epic ground
or plastic and picturesque observation." Hatzfield observes:
Commentators have appreciated the funda-
mental contrast between Don Quixote and Sancho,
between the ideal and the real, between heaven
and earth, between the spirit and the flesh,
between poetry and prose. But this antithesis
also represents a conciliation, the stylistical
nuances of which are very interesting. The
antithesis becomes mellow when Cervantes the
author and observer speaks....
A reflection of Reality and Tdeality united
harmoniously in the author's spirit is found in
the union of dissimilar concepts, that is, in the
linking of the concrete and the abstract in one
phrase. For instance: "But daylight and the hope
of succeeding in their object failed them;" or
"Maidens and modesty wandered at will alone and
Of course, the most oft-cited pairing in Cervantine criticism
is the pairing of the two heroes, Don Quixote and Sancho Panza.
No matter what label the critics in their different analyses
attach to the different dualities, most observe that the mad
master exemplifies one element in the duality and the earthly
servant the other. "In crudest terms," Nelson observes,
Cervantes's narrative formula "may be expressed as the pairing
of a tall, thin idealist with a short, fat realist and setting
them off on a series of hazards." Adams writes, "No observa-
tion about Don Quixote is less startling than that the novel
is built around an unresolved tension between Don Quixote's
world and Sancho Panza's." Gerald Brenan, one of Cervantes's
closest readers, notes that there is some uncertainty in the
narrative in the chapters before Sancho Panza is created:
The knight alone was not a sufficiently strong
thread on which to string the incidents. It
took a few chapters for LCervantes- to discover
this; then, bringing his hero home, he sent him
out again with Sancho Panza. After that there
were no more hesitations: master and man by
their wonderful powers of conversation are
sufficient to sustain the interest. It is the
duality of heroes that turns what could other-
wise be a short entertaining story into a long
and very great book.61
One other characteristic marks the second group of critics:
they believe Don Quixote is the supreme model for prose fiction.
There is among this second group some disagreement about the
theme of Don Quixote, but there is little disagreement about
the novel's place in literary history. Most of these critics
would subscribe to the analysis of Lionel Trilling who wrote
on the 400th anniversary of Cervantes's birth:
In any genre it may happen that the first
great example contains the whole potentiality
of the genre. It has been said that all philo-
sophy is a footnote to Plato. It can be said
that all prose fiction is a variation on the
theme of Don Quixote.62
3. Rene Cirard's Analysis
Rene Girard's reading of Don Quixote's place in literary
history is an interesting combination of diachronic and syn-
chronic approaches. Like Trilling, Girard believes Cervantes's
novel is the supreme model of prose fiction--"All the ideas
of the Western novel are present in germ in Don Quixote,"
he says at one point--but he also believes that the model is
transformed gradually as it appears in later novels. The
model undergoes an evolution.
The full title of Girard's marvelous history of the
Quixotic Novel is Deceit, Desire, And The Novel: Self And
Other In Literary Structure. Like the synchrcnic critics just
considered, Girard finds a duality in Cervantes's narrative,
but unlike them he places this duality (of self and other) in
a precise scheme and he carefully outlines how the scheme can
be found in the novels of Cervantes's imitators. He argues
that the essence of Cervantes's novel is the same essence as
that structuring the novels of Flaubert, Stendhal, Proust, and
Dostoyevsky and can be represented by a symbolic figure--the
triangle. According to Girard, Don Quixote shares with Emma
Bovary, Julien Sorel, Marcel, Prince Myshkin, and many other
heroes of the "great novels" a particular psychological malady:
all these heroes are unable to act spontaneously; all have
given up their freedom of choice and action to live by imitation
of mediators. Each of these heroes is possessed by mediated
desire. Girard explains the first such case of the malady:
Don Quixote has surrendered to Amadis the
individual's fundamental prerogative: he no
longer chooses the objects of his own desire--
Amadis must choose for him. The disciple pursues
objects which are determined for him...by the model
of all chivalry. We shall call this model the
mediator of desire...
In most works of fiction, the characters have
desires which are simpler than Don Quixote's. There
is no mediator; there is only subject and object...
desire is always spontaneous. It can always be
portrayed by a simple straight line which joins
subject and object...
In Don QuixoteJ the mediator is...above the
line, radiating toward both the subject and object.
The spatial metaphor which expresses this triple
relationship is obviously the triangle. The object
changes with every adventure but the triangle remains.
The barber's basin or Master Peter's puppets replace
the windmills; but Amadis is always present.64
Because he is so dedicated to imitating his mediator,
to living by chivalric ideals, Don Quixote is, to a certain
extent, cut off from life, at least the direct experience of
life; he lives instead at one remove from reality. In every
action in which he participates he must move through his
mediator rather than directly; he must consider how Amadis
would act and imitate that action, rather than acting immed-
iately and spontaneously. Although Girard does not bother to
detail it, Don Quixote's relationship with the Dulcinea is
the perfect example of mediated desire. The truth is the
knight hardly knows the girl to whom he pledges eternal dedi-
cation. In some passages in the novel it is not even clear
whether or not he has ever seen her. The dedication he pledges
is, of course, a dedication to an ideal lady not the "real"
Dulcinea, the ugly peasant girl, Aldonza Lorenzo. Don Quixote
feels no spontaneous desire for the girl; all his desire is
mediated through Amadis: he is dedicated to the Dulcinea be-
cause Amadis and all chivalric knights are dedicated to ladies.
Sancho Panza also lives by imitation, according to Girard.
Sancho's mediator is Don Quixote. "Some of Sancho's desires
are not imitated," Girard explains,
for example, those aroused by the sight of a
piece of cheese or a goatskin of wine. But
Sancho has other ambitions besides filling his
stomach. Ever since he has been with Don Quixote
he has been dreaming of an "island" of which he
would be governor, and he wants the title of
duchess for his daughter. These desires do not
come spontaneously to a simple man like Sancho
It is Don Quixote who has put them in his head.65
But Girard is actually more interested in observing
how the theme of mediated desire can be found in the novels of
Flaubert, Stendhal, Proust, and Dostoyevsky and describing the
evolution of the theme than he is in examining Don Quixote
itself. Girard's unique diachronic understanding of the his-
tory of the Quixotic Novel leads him to ignore actual chronology
and arrange the authors who follow Cervantes in the order above.
The development of the theme in the works of these authors
is a development toward the "internalization" of mediation.
The hero in Cervantes's seminal story declares his imitation
openly; he is, therefore, a hero of a novel about external
mediation. The heroes in Proust and Dostoyevsky are unaware
of their imitating and believe, in fact, that they are living
spontaneously; they are controlled by internal mediation.
Girard explains his distinction:
Romantic works are...grouped into two
fundamental categories...We shall speak of
external mediation when the distance is sufficient
to eliminate any contact between the two spheres
of possibilities of which the mediator and the
subject occupy the respective centers. We shall
speak of internal mediation when the same distance
is sufficiently reduced to allow the two spheres
to penetrate each other more or less profoundly.
...The hero of external mediation proclaims
aloud the true nature of his desire. He worships
his model openly and declares himself his disciple.
...The hero of internal mediation, far from boasting
of his efforts to imitate, carefully hides them.66
Don Quixote can openly declare his allegiance to Amadis
because the mediator occupies a different state of existence
than he, the disciple, does. Amadis cannot be Quixote's
rival or enemy because he does not live in Quixote's world.
In The Red And The Black, on the other hand, mediators and
disciples occupy the same universe, the same town, and so can
become bitter enemies. Thus, the imitator cannot admit his
desire is made for him by his need to imitate; it would be
admitting his inferiority toward his rival. M. de Renal hires
Julien Sorel, Girard points out, not because he wants his
children to be educated, but because he fears that Valenod,
his rival and mediator, intends to hire Julien to tutor his
children. In the end of the novel Julien wins back the desire
of Mathilde de la Mole by paying court to her rival (and her
model) the Marechale de Fervacques. Because the Marechale
seems to desire him, Mathilde desires him. Because it appears
his children, Renal must have Julien tutor his children. But,
Renal and Mathilde could never admit the source of their
desires (in their model) even to themselves and so they are
controlled by internal mediation. Girard finds an example of
internal mediation in the Quixote, also, but first he describes
the development of the theme.
He cites Jules de Gaultier's essay "Bovarysm" as one
source of his theory. Gaultier's description of those charac-
ters in Flaubert who are plagued by "bovarysm" is a description
of all those heroes Girard names as plagued by "triangular
The same ignorance, the same inconsistency, the
same absence of individual reaction seem to make
them fated to obey the suggestion of the external
milieu, for lack of auto-suggestion from within.67
As the mediator comes closer to the disciple the malady grows
more malignant. Thus, Girard arranges the "triangular desires"
in the following order, from external mediation to the most
internal, from the least malignant to the most malignant:
quixotism, bovarysm, Stendhalian vanity, Proustian snobbism,
and the Dostoyevskian underground. It is for this reason that
Girard ignores the actual chronology of the novels to arrange
the authors in his own scheme. (He says at one point, fThe
last stage was reserved for...Dostoyevsky, who precedes
Proust chronologically but succeeds him in the history of
triangular desire." ) Don Quixote can openly declare and
recognize his model and so, although he is affected by mediation,
he is "among heroes of novels...the least mad." Emma Bovary
can declare her allegiance to her model, the Paris styles and
fashions and the women who wear them, but her model is
closer to her than Amadis is to Quixote: at the ball at La
Vaubyessard she meets her mediator in person. For the heroes
in Stendhal, Proust, and Dostoyevsky, the relationship with
the mediator is recognized not as an allegiance and dedication
(which it actually is) but as a rivalry and competition. In
these cases disciples find their mediators in the people closest
to them. (In The Brothers Karamazov, for example, the triangles
of desire are drawn from one family member to another, each son
through his father.) There are cases in these novels of what
Girard calls "double mediation," the situation when two rivals
represent mediators for each other and so are each dedicated
to the pursuit of the desires each imagines the other to have.
(The de Renal-Valenod relationship in The Red And Black is such
an example; Varvara Petrovna and Stepan Trofimovitch in The
Possessed "are equally fascinated with each other."
Girard names Dostoyevsky's The Eternal Husband as a kind
of ultimate in the development of triangular desire because in
that story the mediator is so close to the disciple the object
of desire becomes irrelevant. The husband, Pavel Pavlovitch,
can desire only women who seem to attract the former lover of
Pavlovitch's dead wife, Veltchaninov. He follows the lover
around and even brings him to meet the woman he has chosen to
be his next wife. "The Eternal Husband reveals the essence
of internal mediation in the simplest and purest form possible,"
Confronted with Pavel Pavlovitch we can have
no more doubts about the priority of the Other in
desire...The hero is always trying to convince us
that his relationship with the object of desire is
independent of the rival. Here we clearly see that
the hero is deceiving us. The mediator is immobile
and the hero turns around him like a planet around
Then, in a brilliant turn, Girard notes where else this simplest
and purest form of the story of triangular desire has appeared:
in "The Curious Impertinent" in Don Quixote Girard observes:
The two "extremes" of desire, one [external mediation]
illustrated by Cervantes, the other Linternal media-
tior] by Dostoyevsky seem the hardest to incorporate
in the same structure. We can accept that Pavel
Pavlovitch is a brother to Proust's snob and even
to Stendhal's vaniteux, but who would recognize
in him a distant cousin of the famous Don Quixote?
...How could the creator of this sublime being have
an inkling of the swamps in which the eternal hus-
The answer is to be found in one of the short
stories with which Cervantes padded Don Quixote...
"The Curious Impertinent" portrays a triangular
desire exactly like that of Pavel Pavlovitch.72
Girard's parallel makes a convincing argument. Anselmo is the
"eternal husband" in Cervantes's interpolated tale. He is more
interested, it seems (after we have heard Girard's argument),
in his (Lothario's) model's desire for his wife, Camila, than
he is in loving his wife himself. Anselmo's planet circles
around Lothario's sun. "The only difference between the two
stories," Girard believes, "is in technique and the details of
the intrigue...In both cases only the prestige of the mediator
can certify the excellence of a sexual choice."
It is in this way that Don Quixote contains "all the ideas
of the Western novel...in germ." For Girard all the ideas
of the Western novel are variations on the idea of triangular
No literary influence can explain the points
of contact between "The Curious Impertinent" and
The Eternal Husband. The differences are all dif-
ferences of form, while the resemblances are resem-
blances of essence.
No small distance separates the Cervantes of
Don Quixote and the Cervantes of Anselmo since
it encompasses all the novels we have considered
Don Quixote, Madame Bovary, The Red And Black,
Remembrance of Things Past, A Raw Youth, The
Eternal Husband, The Brothers Karamazov, et al.
Yet the distance is not insuperable since all
the novelists are linked to each other; Flaubert,
Stendhal, Proust, and Dostoyevsky form an un-
broken chain for one Cervantes to the other.75
For lirard, then, the Quixotic Principle is a thematic
principle. Cervantes's major contribution to literary history
and specifically the genre of the novel is the introduction
of the theme of mediated desire, a theme which forms the essence
of the many novels Girard examines. It is not a matter of
direct literary influence, he claims: Dostoyevsky "never
realized" the connections between his work and Cervantes's
(because Dostoyevsky "saw the Spanish masterpiece only through
romantic exegeses and probably had a most inaccurate picture
of Cervantes." ) It, the presence of the theme of mediated
desire, is an apparently inevitable, unintended result of the
great novelists writing novels about heroes with desires.
"The creative force of Cervantes is so great," Girard concludes,
"that it is exerted effortessly throughout the whole 'novelistic'
space." By Girard's argument it seems the theme Cervantes
introduces cannot be avoided. It structures the Western novel
"effortlessly" because mediated desire is the essence of life
in Western culture, because mediated desire is a real presence
structuring all our lives.
One of the most appealing aspects of Girard's theory is
that it enables him to explain, rather than reject, Don Quixcte's
deathbed conversion in the final chapter of the novel. As
Girard notes, many commentators (Girard calls them "the romantics")
consider Cervantes's ending to his masterpiece artificial
and insignificant. Although Girard mentions him only once by
name, Unamuno is probably the best-known of those who reject
the last chapter. In The Tragic Sense Of Life Unamuno argues
against the reality of the deathbed conversion by maintaining
that there are two Quixotes. The first Quixote dies after
disassociating himself from the absurd ideals of chivalry
(absurd ideals which Unamuno, of course, espouses); the
second, the immortal Quixote, the man we remember and must
imitate, is the mad knight of the rest of the book: "this
Don Quixote is not converted," Unamuno writes, "this Don
Quixote continues to incite us to make ourselves ridiculous,
this Don Quixote must never die."
It is interesting to compare Girard's analysis with
Unamuno's because Girard, too, finds a spiritual message at
the heart of the Quixotic Novel, but for him it is in Quixote's
rejection of the absurd ideals that Quixote proves himself.
(Girard calls Unamuno's interpretation, in his one reference
to it, "delirious." ) Girard links Quixote's deathbed
conversion to two similar ones--Stepan Trofimovitch's in
The Possessed and Julien Sorel's in The Red And Black.
Girard observes, "The unity of novelistic conclusions con-
sists in the renunciation of metaphysical desire. The
dying hero repudiates his mediator." "The title of hero
of the novel," he adds,
must be reserved for the character who triumphs
over metaphysical desire in a tragic conclusion
and thus becomes capable of writing the novel.
The hero and his creator are separated through-
out the novel but come together in the conclusion.
Approaching death, the hero looks back on his lost
Every novelistic conclusion is a beginning. 80
Every novelistic conclusion is The Past Recaptured.
Girard has, like Unamuno, a tragic sense of life, but
he believes the end of despair lies in a different direction
than the one Unamuno suggests we follow. Unamuno believes
that man must make himself absurd, live by ridiculous ideals,
to live with the fear of death. Girard believes that man
must face death squarely. For Girard it is only in the
acceptance of death and nothingness that man can escape
despair. Heroes of novels triumph when they realize they
must die and that they are, and have been, nothing. The con-
clusions of Don Quixote, The Possessed, The Red And The Black,
and all the other books Girard calls "novels" are the same:
The conclusions of all novels are reminiscent
of an oriental tale in which the hero is hanging
by his finger-tips to the edge of a cliff; exhausted,
the hero lets himself fall into the abyss. He ex-
pects to smash against the rocks below but instead
he is supported by air; the law of gravity is annulled8 1
The hero floats because in acceptance of death he stops living
by mediated desire at last.
If one wants to dispute Girard's theory, the best approach
is, I think, to argue that the triangle explains too much.
Once Girard gives us his triangle we can see more clearly
than ever the obvious links between Emma Bovary and Don Quixote
and we can grasp more easily the complex and distant link be-
tween Don Quixote and Proust's Marcel. The triangle is so
useful, however, that Girard can use it as the basis for many
arguments that intuition should not allow us to accept. The
following passages are taken out of context, but reading them
out of context is the best way to appreciate the dimensions
of Girard's claims. "The movement toward slavery," he says
at one point,
is one of the basic principles of novelistic
structure. Every authentic development in the
novel, no matter how broad its scope, can be
defined as a transition from mastery to
Girard leaves it unclear what can be considered an "authentic
development in the novel" but one must reject, intuitively
rather than logically, any theory that claims to explain every
authentic development in the novel. Girard displays throughout
his book an ability for making the kind of hyperbolic comment
that leaves the reader's mind reeling so that he can't find
any way to grasp the thought. One example comes early in the
book when Girard gives a brief history of triangular desire:
"Romantic revulsion, hatred of society, nostalgia for the
desert, just as Dsic gregariousness, usually conceal a mor-
bid concern for the Other." (Can one live a life style
that reveals one's lack of concern for the Other? Not if
Girard is right.) Another example appears in Girard's analysis
of Remembrance Of Things Past:
Recapturing the past is to welcome a truth
most men spend their lives trying to escape, to
recognize that one has always copied Others in
order to seem original in their eyes and in one's
own. Recapturing the past is to destroy a little
of one's pride.8L
There are also many statements throughout the text which,
no matter how admirable one finds Girard's critical energy,
one must reject as oversimplifications: "It is a fact that
Don Quixote and Sancho Panza are never happy unless they are
being beaten black and blue" ; "Great novels always spring
from an obsession that has been transcended," ; "All heroes
of novels hate themselves..." ; "The ultimate meaning of
desire is death..."
Finally, Girard's explanation of the mechanics of
Cervantes's influence on Western literature must be questioned.
His description of the model in Don Quixote is very precise:
he gives us that single figure, the triangle of mediated de-
sire. But, how does this model come to be used by the other
authors? How exactly does Cervantes's model move "effortlessly
throughout the whole 'novelistic' space?" How does Dostoyevsky
come to use it when he could not have read Don Quixote in
an accurate translation? Is the truth really as Girard im-
plies, that the phenomenon of mediated desire is such a part
of Western culture that authors after Cervantes cannot help
making it the central theme of their work? Or, is the truth
4. Looking Ahead
For the analyst interested in examining Don Quixote
from the point of view of the narrative there is something
to be gained from all three of the approaches sketched above.
All suggest directions for further analysis.
Among the diachronic critics there is agreement that
Quixote's madness is a central element in the text. Both
Foucault and Auerbach understand this element in a way that is,
I think, interestingly close to how Propp and Barthes would
understand it. They both understand (implicitly) that Quixote's
madness serves the discourse's "instinct for preservation."
Foucault calls Quixote's madness "the knot"; Auerbach realizes
that Quixote's madness generates the narrative action. Indeed,
Auerbach's reading of "The Enchanted Dulcinea" episode looks
like Barthes's reading of "The Story's Interest" (Chapter One)
turned inside out. Auerbach's point is that Quixote's "solu-
tion" prevents the hero from coming to grips with reality,
but his analysis demonstrates that Quixote's escape clause
serves the same function served by Sarrasine's ignorance: the
story's interest. When Quixote and Sarrasine learn they are
deluded and come to grips with reality, their stories must
Propp and Barthes might be interested in turning the
analyses of the synchronic critics inside out, also. They
would be interested by Spitzer's, Levin's, Nelson's, and the
others' attempts to reduce the complexities of the novel to
a simple model, a unity of disunities, but they would also
be interested in demystifying the characters of Quixote and
Sancho. There is much to be said for the argument that the
novel is generated by a synthesis of "the ideals of Quixote"
and "the appetites of Sancho" or of "a tall, thin idealist"
and a "short, fat realist." However, Propp and Barthes, and
any one else interested in examing the novel from the point
of view of the narrative, might want to ask whether it is the
opposition of dualities that generates the tale or the tale
that generates the opposition.
Girard's reading is, I think, the most intriguing to the
analyst interested in examining Don Quixote from the point of
view of the narrative. Girard gives us the triangle of
mediated desire, shows us how the figure can be seen as a dy-
namic model structuring the novels of Flaubert, Stendhal,
Proust, and Dostoyevsky, and leaves us with the question about
the mechanics of influence: How does the triangle move "effort-
lessly" through literary history?
1. Erich Auerbach, Mimesis: The Representation Of
Reality In Western Literature, trans. Wiliard R. Trask
(Princeton: Princeton University Press, 1953) p. 358.
2. Salvador de Madariaga, "Don Quixote": An Introductory
Essay In Psychology (London: Oxford University Press, 19bl;
originally published 1934).
3. Erich Kahler, The Inward Turn Of Narrative, trans.
Richard and Clara Winston (Princeton: Princeton University
Press, 1973), p. 57.
4. Michel Foucault, Madness And Civilization: A History
Of Insanity In The Age Of Reason, trans. Richard Howard (London:
Tavistock, 1965, pp.31-32.
5. Auerbach, p. 358.
6. Leo Spitzer, "On The Significance Of Don Quixote,"
Cervantes: A Collection Of Critical Essays, ed. Lowry Nelson,
Jr. (Englewood Cliffs, N. J.: Prentice-Hall, 1967), p. 91.
7. Harry Levin, "The Example of Cervantes," Cervantes:
A Collection Of Critical Essays, p. 36.
8. Gerald Brenan, "Cervantes," in Cervantes: A Collection
Of Critical Essays, p. 16.
9. Kahler, p. 57.
10. Ibid., p. 49. It should be noted that Kahler means
"superindividual significance" in his own special sense, not
as the term might be used to describe, for example, Aeneas
or an Old Testament patriarch.
11. Ibid., p. 5.
12. Ibid., p. 67.
13. Ibid., p. 20.
14. Ibid., p. 35.
15. Ibid., p. 38.
16. Ibid., p. 40.
17. Ibid., p. 57.
19. Jose Ortega y Gasset, Meditations On Quixote, trans.
Evelyn Rugg and Diego Marin (New York: Norton, 1961), p. 51.
20. Foucault also discusses Cervantes in Les Mots et
les choses (Paris: Gallimard, 1966) translated as The Order
Of Things (New York: Pantheon, 1971).
21. Foucault, p. 27.
22. Ibid., p. 28.
23. Ibid., p. 29
24. Ibid., p. 30.
25. Page numbers throughout this study refer to the
English translation by Samuel Putnam. The Ingenious Gentleman
Don Quixote de la Mancha (New York: Modern Library, 1949).
The Spanish In brackets and notes is from the Editorial
Juventud edition (Barcelona: 1971).
26. Ibid., p. 31.
27. Ibid., p. 32
29. Auerbach, p. 331.
30. Ibid., p. 332.
32. Ibid., p. 357.
33. Ibid., p. 340.
34. Ibid., p. 354.
39. Ibid., p. 355.
42. Ibid., p. 358
43. Spitzer, p. 86. Spitzer's diachronic emphasis is
clear enough in what is paraphrased here. It is even more
pronounced in his bold summary of Don Quixote's unity: "Modern
anarchy checked by a classical will to equipoise: the baroque
attitude:" (p. 95). His discovery of a formal equilibrium of
disunities, however, has, as I argue, strong synchronic impli-
44. Ibid., p. 91.
47. Ibid., p. 92.
L8. Levin, p. 30.
49. Ibid., p. 36.
50. E. C. Riley, Cervantes's Theory Of The Novel (Oxford:
Clarendon Press, 1962), p. 11.
51. Ibid., p. 31.
52. Lowry Nelson, Jr. "Introduction,"
Collection Of Critical Essays, p. 10.
53. Ibid., p. 3.
54. Lionel Trilling, "Manners, Morals,
The Liberal Imagination (Garden City, N. Y.:
And The Novel,"
55. Ibid., p. 202.
56. Robert M. Adams, "Two Lines From Cervantes," Strains
Of Discord: Studies In Literary Openness (Ithaca, N. Y.:
Cornell University Press, 1958) p. 73.
57. Frederick R. Karl, A Reader's Guide To The Eighteenth
Century English Novel (New York: Noonday Press, 1974) p. 60.
58. Helmut Hatzfield, "The Style," trans. Edith Mead,
The Anatomy of Don Quixote, ed. M. J. Bernardete and Angel
Flores (Port Washington, N. Y.: Kennikat, 1969) pp. 87-88.
59. Nelson, p. 4.
60. Adams, p. 73.
61. Brenan, p. 16.
62. Trilling, p. 203.
63. Rene Girard, Deceit, Desire, And The Novel: Self
And Other In Literary Structure, trans. Yvonne Freccero
(Baltimore: Johns Hopkins Press, 1965), p. 52.
64. Ibid., pp. 1-2
65. Ibid., p. 3.
66. Ibid., p. 9.
67. Jules de Gaultier in Girard, p. 5.
68. Op. cit., p. 41.
69. Ibid., p. 192
70. Ibid., p. 172
71. Ibid., p. 46
72. Ibid., p. 49.
73. Ibid., p. 50.
74. Ibid., p. 51.
75. Ibid., p. 52.
76. Ibid., pp. 51-52
77. Ibid., p. 52
78. Miguel de Unamuno, Tragic Sense Of Life, trans.
J. E. Crawford Flitch (New York: Dover, 1950I), p. 323.
79. Co. cit., p. 79.
80. Ibid., pp. 296-297
81. Ibid., p. 291.
82. Ibid., p. 170.
83. Ibid., p. 15.
84. Ibid., p. 38.
85. Ibid., p. 180.
86. Ibid., p. 300.
87. Ibid., p. 55.
88. Ibid., p. 290.
SANCHO PANZA FROM THE POINT
OF VIEW OF THE NARRATIVE
This chapter and the next are given to an examination
of the narrative structure of the first Quixotic Novel. The
present chapter attempts to do two things: to explain the
nature of Sancho Panza's "character" and to describe two of
the three kinds of narrative sequences in the novel. The
next chapter focuses on Don Quixote and discusses the third
kind of narrative sequence.
There are, it seems to me, three kinds of sequences in
Don Quixote. All work by the principles outlined by Propp
and Barthes: they are cause and effect sequences whose motion
is a movement from imbalance to the restoration of balance.
The reader is led to move through every sequence by the invoca-
tion of a deficiency (a disequilibrium) and by the implicit
promise of narrative chronology that some event will bring the,
sequence to an end: the deficiency will be liquidated. The
differences in the three kinds of sequences are differences of
length. The shortest sequences can be called the first-order
or atomic sequences; of the three kinds of sequences to be ex-
amined here they most closely resemble the sequences Barthes
calls the actional code. They are brief, two-step processes;
the reader, unless he is analyzing, is not even aware that a
cause has been invoked (implicitly promising an effect) and
an effect given (capping the sequence). All the atomic
sequences can be named and recognized as generic entities, but
the ordinary reader does not usually stop to examine them.
(It should be noted that neither Propp nor Barthes divides
narrative into three kinds of sequence. Both their systems
of analysis will be modified in this discussion. I do not in-
tend to reproduce either system exactly. Rather, by imitating
their spirit and general principles, I intend to find my own way
through the narrative of Don Quixote.)
The next longer sequences are the sequences ordinarily
recognized as narrative sequences by Cervantine critics: these
are the second-order sequences; they can be called the episodic
sequences. They work by the same principles of cause and effect,
but they are longer than the atomic sequences. The "Windmills"
episode is an example of an episodic sequence.
The third kind of sequences can be called the novelistic
sequences. As far as I can tell, there are only five of these
in Don Quixote. These are the loosest structures of all the
sequences and run throughout the novel. They will be discussed
in Chapter Four.
The present chapter is divided into four parts. The first
section is a brief summary of a few of the different interpre-
tations of Sancho Panza's role in Don Quixote. The second
section is an analysis of the narrative of one chapter in
the novel, (II, 28). Five atomic sequences are found and the
function of Sancho in each sequence is examined. The third
section below is an analysis of some of the episodic sequences
in the novel: the "Windmills" episode, the "Enchanted Dulcinea"
episode, and others. A formula is proposed to describe the
structure of these sequences, the formula Q --'V-- 2Q/V.
Sancho's functions in the sequences are again examined.
The analyses of the atomic sequences and the episodic
sequences lead to the same understanding of Sancho's role
in the novel: Sancho has no set role to play; he is a
flexible character who can perform different functions and
fits into the narrative structure in different ways. The
reason critics disagree about Sancho's character is, I will
argue, because Sancho Panza is not a homogeneous character;
he is a complex of various functions. The fourth section
below is a brief answer to an imagined argument, an argument
which might be made by a reader unsympathetic to the methods
adopted in this chapter.
1. Sancho's Role
Sancho Panza is probably the most famous sidekick in
all of literary history. Clearly, he is an important element
in the texture of Don Quixote. Several commentators have
observed that his role is an essential part of Cervantes's
narrative formula. It is interesting, therefore, that there
seem to be several, quite different readings of that role.
As a term in sociological and colloquial contexts out-
side Cervantine criticism "Sancho Panza" has come to represent
the antithesis of "quixotism": a "Sancho" is what a "Quixote"
is not. Usually, in these contexts "Sancho Panza" is a man
of common sense, an earthly realist. A "Sancho" is in touch
with mundane reality; a "Quixote" is deluded by the lure of
Although there are a few analytical readers who argue
for this understanding of Sancho in the text (Turgenev, for
example), this is more often, I think, the understanding of
those who have not read the book and know it only at second-
hand. One cannot read far into the novel without wondering
about Sancho's common sense. In his first appearance, in
fact, (I, 7), Sancho reveals that he has decided to join
Quixote in chivalric adventures because he believes serving
as a squire will earn him a governorship. In the next chapter,
the Adventure of the Windmills, Sancho does seem to be in
touch with mundane reality--he can see the windmills are not
giants--but witnessing Quixote's mad charge and the following
absurd, uncalled-for attack on the Biscayan does not cure
Sancho of his delusion or shake his misplaced faith in
Quixote. After the knight defeats the Biscayan, Sancho
kisses his master's hand and says,
"May your Grace be blessed, Senor Don Quixote,
to grant me that governorship of that island which
you have won in this deadly fray; for however large
it may be, I feel that I am indeed capable of govern-
ing it as well as any man in the world has ever
done." (p. 75)
It is moments like this and Sancho's loyalty to Quixote
throughout the novel that give substance to Girard's claim
that Sancho is not Quixote's antithesis at all. In Deceit,
Desire, And The Novel Girard argues that Sancho is a victim
of the same "triangular desire" that disables Quixote. For
Girard, Sancho is a kind of shadow-Quixote: Sancho is
imitating Quixote in the same way Quixote is imitating
Amadis. "Don Quixote is Sancho's mediator," writes Girard.
The effects of triangular desire are the same in
the two characters. From the moment the mediator's
presence is felt, the sense of reality is lost and
Girard thinks those who see "little more than the contrast
between Don Quixote the idealist and Sancho the realist"are
"romantic readers." The contrast is real," he concedes,
Salvador de Madariaga is another reader who argues
against the ordinary understanding of Sancho. His study
Don Quixote: An Introductory Essay In Psychology is probably
the best-known attempt to explain Sancho's role in the novel.
Madariaga calls the reading that sees Sancho only as Quixote's
antithesis "superficial tradition":
Superficial tradition has reduced the novel's
marvelous psychological fabric to a line of simplest
melody: Don Quixote, a valiant knight and idealist:
Sancho Panza, a matter of fact and cowardly rustic.
What tradition does not see is that this design
which, on first impression, is based on contrast,
resolves itself into a complicated and delicate
parallel. Sancho is...a transposition of Don
Quixote in a different key. (emphasis in original)5
Although from this passage one may think Madariga's reading
is like Girard's, it is, actually, crucially different.
Girard implies that Sancho is a stable entity, a character
never able to live life directly because all his desires must
he mediated, because he can not live life spontaneously. The
essence of Madariaga's reading, on the other hand, is that
Sancho is a character in flux, a character whose nature
changes as he travels through the adventures of the novel.
Madariaga proposes an elegant (and famous) formula to explain
the structure of the novel: Sancho's development is a psycho-
logical movement away from his original earthly, mundane
sensibility toward his companion's quixotic sensibilities;
Quixote's development is a complementary one, a movement from
romance and idealism toward earthliness and realism. Don
Quixote and Sancho Panza begin as opposites, according to
Madariaga, but then they
draw gradually nearer, attracting each other
by virtue of a slow and mutual influence which
is, in its inspiration and its development, the
great charm and achievement of this book.6
Sancho, by this argument, changes his personality as he
learns to imitate Quixote.
...while Sancho's spirit rises from reality to
illusion, Don Quixote's descends from illusion
to reality. And the two curves cross in that
saddest of adventures, one of the cruelest in
the book, when Sancho enchants Dulcinea...7
"The Enchanted Dulcinea" episode marks the spot where Sancho
and Quixote's characters cross because in that scene it is
Sancho who deludes Quixote and it is Quixote who can see
only mundane reality. (Sancho covers up his failure to carry
a message from his master to the Dulcinea by pretending that
three peasant girls are the beautiful Dulcinea and her hand-
A fourth reading of Sancho's role has been provided
recently by John J. Allen in an essay entitled "The Governor-
ship Of Sancho And Quixote's Chivalric Career." Allen, like
Madariaga, believes Sancho is a character in flux, a charac-
ter whose nature changes in the course of the novel, but,
unlike Madariaga, he sees Sancho's development as a movement
away from illusion. Although Madariaga calls his own outline
a "parallel" it is Allen's reading that draws the parallel.
In Madariaga's scheme the characters move toward each other,
Sancho from reality toward Quixote, Quixote from illusion
toward Sancho: it is a crossing pattern. Allen, the most
careful and painstaking critic of Cervantes I have read,
finds and describes a true parallel arrangement in the
stylistic and psychological elements in Sancho's quest for
the governorship and Quixote's quest for chivalric fame.
He believes the two quests take the two characters from
the same kind of delusion, in overconfidence and vanity, to
the same kind of epiphany, in humility and self-understanding:
Both Sancho and Don Quixote [live]through a
process beginning with pride and presumption and a
consequent unawareness of their limitations moving
toward self-discovery through suffering, and cul-
minating with confession and repentance.8
Sancho begins his participation in the adventure, Allen
observes, believing he can govern "as well as any man has
ever done"; he ends his quest for the governorship after the
Duke and Duchess give him his island and he realizes power and
wealth are not for him. He learns his hunger for power was
"I know no more about governing islands
than a buzzard does, and if I thought for a min-
ute that in order to be a governor the devil would
have to carry me off, then I would rather go to
Heaven as Sancho than go to Hell as a governor."
Sancho's realization foreshadows Quixote's deathbed discovery
that he has been deluded. Quixote proves he has made the
same discovery Sancho made on the island when he says, after
the chivalric fever has left him, in response to Sanson
Carrasco's and Sancho's urgings that he continue with their
plans to become shepherds, "Not so fast, gentlemen. In last
year's nests there are no birds this year." (p. 986)
(Allen makes an observation about Quixote I have seen nowhere
else: Quixote is ethical only because his role requires it;
"Don Quixote's initial goal is fame.....The initial attraction
of the books of chivalry is....esthetic, not ethical, and his
desire to right wrongs is simply a necessary consequence of
this attraction.") Allen's reading is unlike Madariaga's
then, because Madariaga believes the two characters change
in different ways while Alien believes they change in the
same way. Specifically, Madariaga argues that "Sancho's
spirit rises from reality to illusion" while Allen argues
that Sancho (like Don Quixote) goes from illusion to reality.
There are ways to reconcile these four readings, but it
better serves my purposes to underline their differences by
presenting the following diagram:
Reader Illusion Reality
"Superficial Quixote Sancho
Madariaga Quixote -
The variety of interpretations is interesting. How can there
be such a disagreement about the fundamental nature of one of
the most famous partnerships in literature? Why haven't
critics, after all these years, come to some uniform under-
standing of the two heroes?
Much has been written of late, by Norman Holland,
Harold Bloom, and others, about the intrinsic plurality of
texts: reading is essentially a subjective activity and so
there can (or must) be as many interpretations as there are
readers. We could use this truth to explain away the con-
tradictions diagrammed above, but a more valuable procedure
would be to attempt to find in the text the sources of the
disagreement. This is what the next section of this paper
2. II, 28: A Sample Analysis
One problem with adapting Propp's and Barthes's methods
to longer narratives is that the going is slow. The analyst
interested in tracing every narrative sequence must proceed
word by word through the text. In the case of Don Quixote
the analyst has a one-thousand page text to consider. I think
the only practical solution is for the analyst to examine a
significant sample of the text and to see what conclusions
he can reach from an examination of that fragment. That is
what I intend to do in this section. I will examine one short
fragment of the novel, Chapter 28 in Part II.
The chapter is not the most memorable in the book. In-
deed, any reader who can remember what happens between the
Braying Adventure (II, 27) and the Adventure of the Enchanted
Bark (II, 29) has an admirable knowledge of Don Quixote. But
the chapter does make a good sample for the purposes of this
study: it is relatively brief; it depends for its existence
on the presence of Sancho (without him, there could be no
chapters in Don Quixote like II, 28); and it is made of
dialog. Several critics have argued that the interludes of
dialog, like II, 28, are the essence of the novel. Criado
de Val has argued, in fact, that the dialog is the only
masterful element in Don Quixote. Gerald Brenan has argued
that it is the dialog between Sancho and Don Quixote that
gives the novel its unity:
Perhaps the relationship between the pair
may best after all be compared to that most
intimate of partnerships, marriage. Their
dialog is made up of the same inconclusive
wranglings, the same recriminations, and tu
quoques, the same fixed recollections and
examples dragged out again and again from the
past to clinch the argument...
And this has the effect of lacing together
in an extraordinary way the various incidents.
One of the most admirable things about this
novel, which at first sight seems to be composed
of a number of separate episodes, strung together
like beads on a thread, is that few things in it
are really finished with when they have occurred.
On the contrary, they are taken up into the minds
of the two protagonists and reappear later on as
a part of their argument. This not only gives the
plot more unity, but it also makes it more subtle.
Every striking event has...a succession of echoes
and it's these echoes that make the book what
Chapter Twenty-seven ends with Don Quixote running from
a battle for the first time since his knighting. Sancho has
stirred the anger of the town of the brayers by braying in
their presence, an act they consider mockery. They knock
Sancho senseless and attack Quixote with stones. Don Quixote
retreats with his heart in his mouth.
The reader is asked here to consult the complete text
of II, 28.
I have broken the chapter into what seem to me to be the
five atomic sequences that make up the chapter. The numbers
on the skeletal outline below represent the five sequences;
the letters mark the paragraphs (in the Putnam translation)
within each sequence for easier reference.
Chapter XXVIII. Of things that, Benengeli says, the reader
will come to know if he reads attentively.
l.a.) When the brave man flees, it means...
l.b.) This truth was brought out in the case of Don
l.c.) "An evil hour it was, Sancho, when you learned to
bray! Where did you ever hear that it was a good thing to
mention the rope in the house of the hanged man?"
l.d.) "I'm in no condition to answer...I'll not bray
anymore, but I can't help remarking that knights-errant appear
to run away and leave their faithful squire in the hands of
I.e.) "He who retires," said Don Quixote, "does not
flee...The histories are full of such instances, but as it
would do you no good to refer you to them, I shall spare
myself the trouble for the present."
There are a few complexities here which hide the funda-
mental structure somewhat, but there is a generic sequence
here. It can be named: A witness confronts a pretender with
evidence of his duplicity; the pretender finds an excuse to
justify his duplicity and so continue his pose. The pretender
is, of course, Quixote. Sancho's role in this sequence is a
role he plays at many other places in the narrative: he is a
witness; he has heard Quixote extol his knightly courage and
has seen him run like a coward. His comment "I can't help
remarking..." initiates the cause and effect sequence (cause=
confrontation; effect=justification). The sequence is one
the reader recognizes from the already-read; it is a scene he
has encountered elsewhere, perhaps in Act II, Scene iv of
Henry The Fourth, Part I, where Hal and Falstaff perform a
sequence much like the above. (Hal, playing the witness, con-
fronts Falstaff, the pretender, with evidence that the story
Falstaff has been telling about dueling with a dozen men is a
lie since it was Hal himself, in disguise, who defeated him.
Like Quixote, Falstaff quickly finds an excuse: "By the
Lord," he exclaims, "I knew ye as well as he that made ye...
Was it for me to kill the heir-apparent?")
It is worthwhile looking at the complexities of this
sequence for a few moments because in this particular move
Cervantes uses several narrative elements besides the cause-
effect sequence to generate his narrative. First of all, he
uses the momentum of the previous chapter to carry his reader
for a few paragraphs into this chapter; in fact, he repeats
himself. Paragraph l.b.) is a repetition of the last para-
graph of the previous chapter. It is an awkward moment in
the narrative because the repetition is unnecessary and it is
in moments like this that we sense Cervantes struggling, his
pen is moving across the page, putting black on white, while
he waits for his narrative genius to come up with something.
The moment sets the tone for the whole chapter: it is an
improvisation. It is moments like this that Auerbach is
thinking of when he observes that Cervantes had no real narra-
tive plan for Don Quixote but only allowed himself "to be
guided by the momentary situation, by the demands of the
adventure in hand." The narrative does not unravel under
inspection, however. Cervantes may be improvising, but he is
improvising like a genius. He ties the sequence to the rest
of the narrative by the echoing proverb (1.c.) "Where did
you ever hear it was a good thing to mention the rope..."
etc., a proverb Sancho used hundreds of pages earlier.
Cervantes also gives the sequence a symmetry by having Quixote
repeat Cid Hamete's explanation of why brave men flee. The
novel is filled with little ironic touches like this. Here
the Cid's commentary sounds true enough until it is subverted
by Quixote's self-justifying use of it. Quixote's argument
is undercut, to add another irony, by the fact that he has
just argued for the opposite opinion in "The Adventure Of The
Lions" when he observes, "It is better for the brave man to
carry his bravery to the point of rashness than for him to
sink into cowardice." (p. 619) All of these complexities
deserve more analysis and commentary, but they are not part
of the atomic structure. What makes the opening sequence
work is the question and the answer. Sancho poses the question
that the reader wants asked (the reader wants to know how
Quixote will justify himself this time) and Quixote gives the
answer that must be given (for the narrative to continue the
quixote must continue to delude himself).
The other four moves in the chapter may be more quickly
2.a.) Sancho was by now once more on his gray's back...
Every so often Sancho would heave a deep sigh or moan...it
was nearly driving him crazy.
2.b.) "The reason for that," remarked Don Quixote, "is
undoubtedly the fact that the club they used was a long one..."
2.c.) "By God," exclaimed Sancho, "your Grace has taken
a great load off my mind..."
2.d.) "I'd do a lot better," he went on, "...by going
home to my wife and young ones...l'd like to see the one who
started this knight-errantry business burned to ashes..."
Quixote initiates this sequence by throwing salt in
Sancho's wounds; his gratuitous analysis ("...the club they
used was a long one...") is the cause; Sancho's angry reply is
the effect. The sequence can be generically named: The
callous master treats his suffering servant too cruelly; the
servant reacts with harsh insubordinate words. Sancho plays
in this move the role of the anti-Quixote: his words are
the words of the earthly realist, of a man of common sense,
"I'd do a lot better by going home to my wife..."
The next two sequences are interlocked; there are two
separate sequences, but the second, #4, is initiated before
#3 has reached its conclusion:
3.a.) "I would lay a good wager with you, Sancho..."
3.b./2.e.) "When I worked for Tome Carrasco...I earned
two ducats a month...The work in the fields may be hard, but
at the worst...we have our olla and a bed to sleep in--
and that's something I haven't had since I've been serving
3.c.) "I admit that everything you say is true, Sancho..."
3.d.) "As I see it, if your Grace would give me a couple
of reales more for each month, I would consider myself well
3.e.) "Very well, it is now twenty-five days...
4.a.) "Oh body of me!"
4.b.) "Well and how long has it been..."
t.c.) "If I'm not mistaken...more than twenty years and
4.d.) At this Don Quixote slapped his forehead...
4.e./3.f.) "Why with my wanderings in the Sierra and all
the rest barely two months have gone by...But tell me, you
perverter of the laws of chivalry...where have you ever heard
of a squire who made such terms with his lord...?
3.g.) "Turn, then, the reins...and go back to your home..."
3.h.) Sancho...was so smitten with remorse that the tears
came to his eyes.
3.i./5.a.) "Master, I will grant you that all I lack is
a tail and I would be an ass..."
The third sequence is the longest of the five in Chapter
Twenty-eight; it can be named The wit lays a trap for the
fool; the fool is trapped. Quixote tricks Sancho in #3:
he leads him into revealing his unchivalrous materialism so
he can abuse him by pointing to it; Quixote uses the trap
to contrast his idealism to Sancho's commonness. The se-
quence is difficult to extricate from the narrative because
#4 is intertwined with it and there is a residue of #2: in
3.b./2.e.) the third sequence is already in motion--Don
Quixote has lured Sancho into calculating his pay and Sancho
has begun to do that--but, there is still some anger to be
spoken from sequence #2. In this paragraph then Sancho's
speech is part of two sequences: it is part of #2 because
the anger is the effect of sequence #2, and it is part of
#3 because Sancho is in this speech stepping into the trap
that forms the structure of this sequence.
Sequence #4 is inserted inside #3. It can be titled:
The madman speaks his madness; the sane man laughs a hearty
laugh. Cervantes invokes Sancho's madness at this point to
generate a move inside a move. The madness of his two heroes
is from the author's point of view an invaluable element.
The author can create a move at any moment by having either
hero demonstrate his madness. No justification, introduction,
or explanation (other than that the heroes are madmen) is
necessary, and the madness usually generates a sequence: the
madman may be wondered and laughed at (as here), or abused
(e.g. by the Duke and Duchess), or dealt with more sympa-
thetically (e.g. by Sanson Carrasco, Dorothea, and the
curate and the barber). In any case, madness is a conven-
ient cause which can be inserted at any time. Sancho has,
obviously, no reason to say here that the journey has lasted
"twenty years and three days, more or less" (4.c.); the
reader accepts the statement into the narrative, however,
because he knows Sancho is a madman; the reader forgets
that the real reason Sancho speaks this absurdity is be-
cause Miguel de Cervantes needs to put black on white:
Sancho's madness is one technique Cervantes can use to fill
an empty page. This is the truth behind Sancho's behavior:
his madness is not some analyzable function of a certain
psychological personality; his madness is a narrative ele-
ment. This truth is further revealed by a close examination
of sequence #3. Why does Sancho apologize? Why is he
suddenly so smitten with remorse? The words he spoke in
anger were sincere and accurate: he has had to endure some
difficult conditions and it does seem that his squireship
is not leading him to the rewards he was promised. His
sudden apology does not make "sense," then. But the reader
does not stop to find fault with the narrative here; the
reader is, instead, carried along by the cause and effect
sequence. Sancho's apology makes "sense" because it is part
of the operation of the sequence, because the narrative gave
him a chance, the narrative provided a space, for apology.
The fool's apology makes "sense" because the wit has tricked
An analysis of #5 shows that this sequence is slightly
different from the other sequences. The skeleton of #5 is:
3.i./5.a.) "Master, I will grant you that all I lack
is a tail and I would be an ass..."
5.b.) "I should have been surprised, Sancho if in the
course of your speech you had not rung in some proverb. Very
well, I forgive you...you should take heart..."
5.c.) Sancho assured him that he would do so...
It is difficult to name this sequence in the way the other
four have been named, because here the initiation of the se-
quence is mutual: Sancho's "Master, I will grant..." and
Quixote's "I should have been surprised..." are both implicit
offers of reconciliation and form together the cause in this
fifth sequence. The sequence is completed mutually also:
Quixote's "I forgive you..." and Sancho's assurance to his
master that he will take heart form together the effect in
the sequence. The name of the sequence, then, must be some-
thing like Two friends (or lovers) who have abused each
other make mutual offers of reconciliation; the reconciliation
is accepted and the friendship restored. In each of the
other sequences the cause was initiated by one member of the
pair of the heroes and the effect was completed by the other
member. Quixote initiates #2 and #3 and completes il and #4;
Sancho initiates #1 and #4 and completes #2 and #3. The
balance in #5 in which both characters participate in the
initiation and in the conclusion suggests the truth: the
fifth sequence caps the sequential movement of the whole
chapter. Chapter Twenty-eight is itself structured by a
cause and effect sequence. It can be entitled: Two friends
fail each other; they discover the means to forgive each
other. With this generic title the chapter can be analyt-
ically described. It is an interlude between adventures
and its structure indicates its nature: it is a pause, a
delay, a hesitation in the narrative. The pair of heroes
pause in their adventures to argue; the bond between them
is tested but not broken because the narrative must continue.
With the bond restored to its strength by the final move
in the chapter the narrative can lead the pair to the next
If the sequences have been properly identified, the
chapter can now be diagrammed:
1. Witness observes Pretender
duplicity finds excuse.
2. Master abuses Servant
servant speaks up.
3. Wit lays trap Fool falls for Friends fai:
Sane man laughs.
ion--L accept reconcile
The student interested in an empirical understanding of
narrative could at this point set out to prove each of the