Authors: RENÉ GIRARD
imaginary element, but it is the very specific imagination of people who crave violence. As a
result, among the textual representations there is a mutual confirmation. This correspondence
can be explained by only one hypothesis. The text we are reading has its roots in a real
persecution described from the perspective of the persecutors. The perspective is inevitably
deceptive since the persecutors are convinced that their violence is justified; they consider
themselves judges, and therefore they must have guilty victims, yet their perspective is to
some degree reliable, for the certainty of being right encourages them to hide nothing of their
massacres.
Faced with a text such as Guillaume de Machaut's, it is legitimate to suspend the general rule
by which the text as a whole is never worth more, as far as real information goes, than the
least reliable of its features. If the text describes circumstances favorable to persecution, if it
presents us with victims of the type that persecutors usually choose, and if, in addition, it
represents these victims as guilty of the type of crimes which persecutors normally attribute
to their victims, then it is very likely that the persecution is real. If this reality is confirmed by the text itself then there is little scope for doubt.
When one begins to understand the perspective of the persecutors, the absurdity of their
accusations strengthens rather than compromises the informational value of the text, but only
in reference to the violence that it echoes. If Guillaume had added stories of ritual infanticide
to
-102-
the episodes of poisoning, his account would be even more improbable without, however, in
the least diminishing the accuracy of the massacres it reports. The more unlikely the
accusations in this genre of text the more they strengthen the probability of the massacres:
they confirm for us the psychosocial context within which the massacres must have taken
place. Conversely, if the theme of massacres is placed alongside the theme of an epidemic it provides the historical context within which even the most precise scholar could take this
account of poisoning seriously.
The accounts of persecutions are no doubt inaccurate, but in a way they are so characteristic
of persecutors in general, and of medieval persecutors in particular, that the text can be
believed in all the areas in which conjectures are prompted by the very nature of the
inaccuracy. When potential persecutors describe the reality of their persecutions, they should
be believed.
The combination of the two types of characteristics generates certainty. If the combination
were only to be found in rare examples we could not be so certain. But its frequency is too
great to allow doubt. Only actual persecution seen from the perspective of the persecutors can
explain the regular combination of these characteristics. Our interpretation of all the texts is
confirmed statistically.
The fact that certainty is statistically verifiable does not mean it is based only on an
accumulation of equally uncertain documents. All documents like Guillaume de Machaut's
are of considerable value because in them the probable and improbable interact in such a way
that each explains and justifies the presence of the other. If there is a statistical character to
our certainty it is because any document studied in isolation could be forged. This is unlikely,
but not impossible, in the case of a single document. And yet it is impossible where a great
number of documents are concerned.
The modern Western world chooses to interpret "texts of persecution" as real, this being the only possible way to demystify them. This solution is accurate and perfect because it makes
allowance for all the characteristics found in this type of text. Solid intellectual reasoning is
the basis, rather than humanitarianism or ideology. This interpretation has not usurped the
almost unanimous agreement granted it. For the social historian reliable testimony, rather
than the testimony of someone who shares Guillaume de Machaut's illusions, will never be as
valuable as the unreliable testimony of persecutors, or their accomplices, which reveals more
because of its unconscious nature. The conclusive document belongs to persecutors who are
too naïve to cover the traces of their crimes, in contrast to modern persecutors who are too
cautious to leave behind documents that might be used against them.
I call those persecutors naïve who are still convinced that they are
-103-
right and who are not so mistrustful as to cover up or censor the fundamental characteristics
of their persecution. Such characteristics are either clearly apparent in the text and are
directly revealing or they remain hidden and reveal indirectly. They are all strong stereotypes
and the combination of both types, one obvious and one hidden, provides us with information
about the nature of these texts.
We are all able today to recognize the stereotypes of persecution. But what is now common
knowledge scarcely existed in the fourteenth century. Naïve persecutors
are unaware of what
they are doing
. Their conscience is too good to deceive their readers systematically, and they present things as they see them. They do not suspect that by writing their accounts they are
arming posterity against them. This is true of the infamous "witch-hunts" of the sixteenth
century. It is still true today in the backward regions of the world. We are, then, dealing with the commonplace, and my readers may be bored by my insistence on these first obvious facts.
The purpose will soon be seen. One slight displacement is enough to transform what is taken
for granted, in the case of Guillaume de Machaut, into something unusual and even
inconceivable.
My readers will have already observed that in speaking as I do I contradict certain principles
that numerous critics hold as sacrosanct. I am always told one must never do violence to the
text. Faced with Guillaume de Machaut the choice is clear: one must either do violence to the
text or let the text forever do violence to innocent victims. Certain principles universally held
to be valid in our day, because they seem to guard against the excesses of certain
interpretations, can bring about disastrous consequences never anticipated by those who,
thinking they have foreseen everything, consider the principles inviolable. Everyone believes
that the first duty of the critic is to respect the meaning of texts. Can this principle be
sustained in the face of Guillaume de Machaut's work?
Another contemporary notion suffers in the light of Guillaume de Machaut's text, or rather
from the unhesitating way we read it, and that is the casual way in which literary critics
dismiss what they call the "referent." In current linguistic jargon the referent is the subject of the text; in our example it is the massacre of the Jews, who were seen as responsible for the
poisoning of Christians. For some twenty years the referent has been considered more or less
inaccessible. It is unimportant, we hear, whether we are capable or not of reaching it; this
naïve notion of the referent would seem only to hamper the latest study of textuality. Now the
only thing that matters is the ambiguous and unreliable relationships of language. This
perspective is not to be rejected wholesale, but in applying it in a scholarly way we run the
risk that only Ernest Hoeppfner, Guillaume's editor in the venerable
Société des anciens
textes
, will be
-104-
seen as the truly ideal critic of that writer. His introduction does in fact speak of courtly
poetry, but there is never any mention of the massacre of the Jews during the plague.
The passage from Guillaume provides a good example of what I have called in
Things
Hidden since the Foundation of the World
"persecution texts."
4. B
y that I mean accounts of real violence, often collective, told from the perspective of the persecutors, and therefore
influenced by characteristic distortions. These distortions must be identified and corrected in
order to reveal the arbitrary nature of the violence that the persecution text presents as
justified.
We need not examine at length the accounts of witch trials to determine the presence of the
same combination of real and imaginary, though not gratuitous, details that we found in the
text of Guillaume de Machaut. Everything is presented as fact, but we do not believe all of it,
nor do we believe that everything is false. Generally we have no difficulty in distinguishing
fact from fiction. Again, the accusations made in trials seem ridiculous, even though the
witch may consider them true and there may be reason to suspect her confession was not
obtained by torture. The accused may well believe herself to be a witch, and may well have
tried to harm her neighbors by magical proceedings. We still do not consider that she
deserves the death sentence. We do not believe that magic is effective. We have no difficulty
in accepting that the victim shares her torturers' ridiculous belief in the efficacy of witchcraft but this belief does not affect us; our skepticism is not shaken.
During the trial not a single voice is raised to reestablish or, rather, to establish the truth. No
one is capable of doing so. This means that not only the judges and witnesses but also the
accused are not in agreement with our interpretation of their own texts. This unanimity fails
to influence us. The authors of these documents were there and we were not. We have access
to no information that did not come from them. And yet, several centuries later, one single
historian or even the first person to read the text feels he has the right to dispute the sentence
pronounced on the witches
. 5.
Guillaume de Machaut is reinterpreted in the same extreme way, the same audacity is
exercised in overthrowing the text, the same intellectual operation is in effect with the same
certainty, based on the same type of reasoning. The fact that some of the details are imagined
does not persuade us to consider the whole text imaginary. On the contrary, the
____________________
4. Girard,
Things Hidden
, 126-38.
5. J. Hansen,
Zauberwahn, Inquisition und Hexenprozess im Mittelalter und die Entstehung
der grossen Hexenverfolgung
( Munich, Leipzig: Scientia, 1900); Delumeau,
La Peur en
Occident
, vol. 2, chap. 2. On the end of the witchcraft trials, see Robert Mandrou ,
Magistrats et sorciers
( Paris: Plon, 1968). See also Natalie Zemon Davis,
Society and
Culture in Early Modern France
( Stanford, Calif.: Stanford University Press, 1975).
-105-
incredible accusations strengthen rather than diminish the credibility of the other facts.
Once more we encounter what would seem to be, but is not, a paradoxical relationship
between the probable and improbable details that enter into the text's composition. It is in the
light of this relationship, not yet articulated but no less apparent to us, that we will evaluate
the quantity and quality of the information that can be drawn from our text. If the document is
of a legal nature, the results are usually as positive or even more positive than in the case of
Guillaume de Machaut. It is unfortunate that most of the accounts were burned with the
witches. The accusations are absurd and the sentence unjust, but the texts have been edited
with the care and clarity that generally characterize legal documents. Our confidence is
therefore well placed. There is no suspicion that we secretly sympathize with those who
conducted the witch-hunts. The historian who would consider all the details of a trial equally
fantastic, on the excuse that some of them are tainted by the distortions of the persecutors, is
no expert, and his colleagues would not take him seriously. The most effective criticism does
not consist in rejecting even the believable data on the ground that it is better to sin by excess
rather than lack of distrust. Once again the principle of unlimited mistrust must give way to
the golden rule of persecution texts: the mind of a persecutor creates a certain type of illusion
and the traces of his illusion confirm rather than invalidate the existence of a certain kind of
event, the persecution itself in which the witch is put to death. To distinguish the true from
the false is a simple matter, since each bears the clear mark of a stereotype.
In order to understand the reasons behind this extraordinary assurance evidenced in
persecution texts, we must enumerate and describe the stereotypes. This is also not a difficult
task. It is merely a question of articulating an understanding we already possess. We are not
aware of its scope because we never examine it in a systematic fashion. The understanding in question remains captive in the concrete examples to which we apply it, and these always
belong to the mainly Western historical domain. We have never yet tried to apply this
understanding beyond that domain, for example, to the so-called ethnological universe. To
make this possible I am now going to sketch, in summary fashion, a typology of the
stereotypes of persecuti
on. 6.
____________________
6. This sketch of stereotypes of persecution, chapter 2 of
The Scapegoat
, is presented in the next chapter of the Reader. -
J. W.