Research

in
Phenomenology
Research in Phenomenology 39 (2009) 184–203 brill.nl/rp

Miracle and Machine:

The Two Sources of Religion and Science in
Derrida’s “Faith and Knowledge”

Michael Naas
DePaul University

Abstract
This essay attempts to lay out the three principal theses of Jacques Derrida’s 1994–1995 “Faith
and Knowledge,” Derrida’s most sustained but also most challenging work on the nature of
religion and the relationship between religion and science. After demonstrating through these
three theses that religion and science not only share a common source—or have a common
genesis—but are in what Derrida calls an autoimmune relationship to one another, the essay
puts these theses to the test by reading a brief passage near the middle of the essay where Derrida
recounts the genesis of “Faith and Knowledge” itself. Derrida’s seemingly anecdotal recounting
of this genesis is thus shown to reflect the three theses of “Faith and Knowledge,” the way in
which, in a word, the breath of creation, or the miracle of religion, is always doubled, supple-
mented, and thus contaminated by the machine of science and tele-technology.

Keywords
Derrida, religion, science, faith, genesis, autoimmunity

In any debate over whether, when, or to what extent Derrida took that now
infamous “theological turn” in French phenomenology, the 1994–95 essay
“Faith and Knowledge” turns out to be not just central but absolutely pivotal.1
Though Derrida had, of course, addressed many theological questions in texts
as early as Glas (1974),2 though he frequently addressed during the 1980s the

1)
I am referring here, of course, to Dominique Janicaud et al., Phenomenology and the “Theo-
logical Turn” (New York: Fordham University Press, 2000).
2)
Glas (Paris: Éditions Galilée, 1974) translated by John P. Leavey, Jr., and Richard Rand as Glas
(Lincoln: University of Nebraska Press, 1986). Hereafter cited as Glas, followed by French then
English page numbers.
© Koninklijke Brill NV, Leiden, 2009 DOI: 10.1163/156916409X448166
 M. Naas / Research in Phenomenology 39 (2009) 184–203 185

relationship between deconstruction and negative theology,3 and though his

own writing in the early 1990s—in “Circumfession” and Memoirs of the Blind,4
for example—not only evoked religious themes but engaged the religious
genres of confessional writing and conversion narrative, it is in “Faith and
Knowledge” that Derrida takes on most explicitly the question of the nature
of religion itself and, perhaps most importantly, its place in the world today.5
To think Derrida’s relationship to the question of religion, then, it is essential
to spend some time with this highly provocative and illuminating though also
extremely dense and sometimes cryptic sixty-page essay around which so
much revolves and upon which so much hinges.
Readers have generally acknowledged the importance of “Faith and Knowl-
edge” in Derrida’s work, even if the sometimes elliptical style of this essay has
made it difficult to articulate precisely what is being argued there. The essay
has thus already provoked a great deal of discussion about, for example,
the precise relationship between a general structure of religiosity and various
determinate, revealed religions, or else the role played by the Greek khōra in a
text that is ostensibly about the three Abrahamic monotheisms, or else the
relationship between Derrida’s and Kant’s views on religion or, to a lesser
extent, Bergson’s, but little attempt has been made to spell out the general
argument about religion in this essay and the way in which Derrida’s writing
contributes to and exemplifies that argument. In the pages that follow I would
like to move from the greatest level of generality, from three theses that, on my
reading, run beneath the entirety of “Faith and Knowledge,” to a single, brief
passage near the middle of the essay where Derrida recounts the genesis of
“Faith and Knowledge” and where, as we will see, this genesis reflects the three
theses of which it is the origin. Whether read as an expression of Derrida’s
most poignant revelations about the nature of religion or, as I will argue, as a

3)
See, for example, “How to Avoid Speaking: Denials,” trans. Ken Frieden and Elizabeth
Rottenberg, in vol. 2 of Psyche. Inventions of the Other, ed. Peggy Kamuf and Elizabeth Rottenberg
(Stanford, CA: Stanford University Press, 2007), 143–95. This essay was originally written for a
conference in Jerusalem in 1986.
4)
“Circumfession,” in Jacques Derrida, with Geoffrey Bennington (Chicago: University of Chicago
Press, 1993); Memoirs of the Blind, trans. Pascale-Anne Brault and Michael Naas (Chicago:
University of Chicago Press, 1993).
5)
“Faith and Knowledge,” translated by Samuel Weber, can be found in a couple of different
places: Acts of Religion, ed. Gil Anidjar (New York: Routledge, 2002), 42–101, and Religion, ed.
Jacques Derrida and Gianni Vattimo (Stanford, CA: Stanford University Press, 1998), 1–78. In
order to facilitate reference to either edition, as well as to the original French, I will refer through-
out to section rather than page numbers, preceded by the abbreviation FK.
186 M. Naas / Research in Phenomenology 39 (2009) 184–203

little textual machine that makes the truth around which it turns and that
affirms a faith that is inseparable from that machine, “Faith and Knowledge”
is a text that makes serious and unavoidable demands upon us. It demands to
be read and reread, interpreted and studied as the unique and unrepeatable
inscription or event that it is, but also, and especially, to be taken apart and
broken down, analyzed and reassembled, so that we might understand the
machinery behind this event, the machine, as we will see, that is at the origin
of the miracle, and the miracle that will have made possible the machine.
Now, in speaking here of both the miracle and the machine and of each as
the condition or source of the other, I am already evoking the fundamental
duplicity that Derrida (borrowing a phrase from Bergson—already a first
duplicity!) will find at the origin of religion. The subtitle of “Faith and Knowl-
edge” is thus “The Two Sources of ‘Religion’ at the Limits of Reason Alone,” a
subtitle—a second title—that not only names two sources of religion but con-
denses two of the secondary sources that form the backdrop of Derrida’s anal-
ysis, Kant’s 1793 book Religion Within the Limits of Reason Alone and Henri
Bergson’s 1932 work The Two Sources of Morality and Religion. But that is really
just the beginning of the duplicity. Apart from the fact that “Faith and Knowl-
edge” is really just a translation or double or mere repetition of the title of
Hegel’s 1802–3 essay “Glauben und Wissen,”6 almost every “source” divides
in two—at least two—from the very beginning: hence Derrida marks the
essay with two places, one European and one American, the Italian island of
Capri where an early version of “Faith and Knowledge” was first delivered and
Laguna Beach in California where the essay was completed; and these two
places correspond to two times, February 28, 1994, in Capri, and April 26,
1995, in Laguna Beach. As for the form of the essay itself, in addition to its
two titles, and its subtitle that speaks of two sources and combines two sec-
ondary sources, the essay has fifty-two sections that are themselves divided
into two sets of twenty-six, the first set printed in italics and the second not,
with this second (entitled Post-Scriptum) itself divided into two unequal sec-
tions with two distinct though seemingly related headings. This is just a small
sampling of the many doublings and moments of duplicity in this at once
improvised and highly constructed and organized text. By means of this
duplicity at the level of form, Derrida seems already to be suggesting that any

6)
Hegel’s “Glauben und Wissen” is translated into English by Walter Cerf and H. S. Harris as,
precisely, Faith and Knowledge (Albany, NY: State University of New York Press, 1977). Derrida
cites the entire final paragraph of this work in Glas, 111a/96a, and refers to the same paragraph
at FK §18.
 M. Naas / Research in Phenomenology 39 (2009) 184–203 187

analysis that attributes a single source to religion, a single source that would
remain intact, indemnified, or immune from all duplicity and contamination,
will miss what has always been the case about religion and what is most strik-
ing about it today. Here too, then, two times (always and today), the archaic
and the contemporary, appear necessary to think religion “itself,” that is, reli-
gion as it has been from the beginning, divided from itself because of its two
distinct sources and because of its complicity with that which might appear to
oppose it—reason, science, technology, telecommunications, and so on. In a
word, in two words, religion must be understood today in relation to both
faith and knowledge.
Though Derrida seems to make a point not to develop his argument in
“Faith and Knowledge” in any kind of a straightforward or linear fashion,
preferring instead to scatter his thoughts through the fifty-two sections like a
good game of fifty-two card pick-up, I would like to argue that there are essen-
tially three main theses underlying the essay and that these can and must be
ordered in a particular way. All three of these theses express the fundamental
duplicity of religion or else the fundamental complicity and duplicity of reli-
gion and science.
First thesis, then, first thesis of duplicity—religion has not one but two
sources. On the one hand, argues Derrida, religion has its source in an experi-
ence of sacrality or holiness, the indemnified or the unscathed, the safe and
sound—terms that in “Faith and Knowledge” Derrida will not tire to link
together into a single network. The French word salut, which can mean either
health and salvation, restoration or redemption, would perhaps be the best
single word to identify this network or configuration. It is difficult to imagine
a religion, Derrida claims, that does not promote or promise a restoration,
redemption, or indemnification of the self or the community through various
kinds of ritual, sacrifice, or prayer. There is no religion, Derrida suggests, with-
out some promise of health or redemption, without an attempt to keep the
self or the religious community safe from contamination or desecration.
But this reference to the promise already suggests another source for reli-
gion, one related not to the object of that promise (indemnification, sacraliza-
tion, health, or redemption) but to the act of the promise itself. There is, writes
Derrida (FK §30), “no religion without the promise to keep the promise to say
the truth in promising to say it . . . in already having promised to say it. . . . No
religion without the act of the promise.” In its most elementary sense, bearing
witness has to do with promising to say the truth, promising that what I have
already promised is to tell the truth. Hence Derrida argues throughout “Faith
and Knowledge” that religion’s other source is to be found not in the promise
188 M. Naas / Research in Phenomenology 39 (2009) 184–203

of health, redemption, or salvation, but in the promise that would precede,

exceed, and condition such a promise in the form of an originary gage or
engagement to the other, an experience of faith that has to do not with the
indemnified community but with credit, confidence, and the good faith of
witnessing. This second source must be thought before any particular promise
of health or redemption as an originary greeting of the other, an originary turn
toward the other or address of the other before any attempt even to recognize
the other and affirm them within a community. This second source would
thus correspond to yet another meaning of the French word salut: no longer
understood as health or salvation, no longer a noun to be referred to, invoked,
or promised, the second source of religion would have to be thought in rela-
tionship to the salut! as an originary performative greeting of the other, a
threshold greeting that would precede and condition the constitution of any
religious community.
On the one hand, then, religion would have its source in an experience of
something that must remain intact or that must be restored, protected, safe-
guarded, indemnified, while on the other it has to do with a kind of faith in,
promise to, or engagement with regard to the other. While one source is thus
turned, we might say, toward a protection or indemnification of the self or the
community, an immunization of it from outside aggression or contamination,
the other source is turned precisely toward this outside, or toward this outside
within, toward all the resources that, as we will see, threaten the self or the
community but also make these possible in the first place.
This leads Derrida to his second thesis in “Faith and Knowledge.” In order
for religion or a religious community to protect and promote the first of these
two sources, namely, the experience of salut as health, security, and indemnifi-
cation, it must rely upon and enlist the resources of technoscience and tele-
communications. An analysis of what is happening in religion today makes
very clear what has always been the case: while religion often rejects and tries
to get beyond or before science, it can do so only by appropriating the very
means of science—the abstraction or deracination of teletechnoscience, tele-
communications, and so on (FK §37). There is, thus, as Derrida will call it, an
autoimmune relationship between religion and science: in order to protect or
indemnify itself, in order to return to some nature, idiom, family, or filiation
that would remain immune from abstraction, translation, universalization,
and so on, religion must appropriate the very things it opposes and court that
which compromises and contaminates it.
We will look in more detail in a moment at how Derrida develops this
second thesis, but before doing so let me introduce the third and final thesis
 M. Naas / Research in Phenomenology 39 (2009) 184–203 189

of “Faith and Knowledge.” If religion turns to science in order to indemnify

itself, if it turns, in an autoimmune fashion, to the very thing that will com-
promise what it is trying to protect, then that is because there is an even more
intimate relationship between religion and science or faith and knowledge.
That is, religion at once needs science and is in open antagonism against it in
spite of the fact—because of the fact—that the two share a common source. For
in addition to having a source in reason or technological thinking, science has
another source, a second source, in the promise or performative faith that
makes science precisely performative. For science, like religion, requires faith,
trust, credit, and so on, an originary faith that is anterior to every science and
is the quasi-transcendental condition of all knowledge. Hence a certain sci-
ence has always been associated with religion both insofar as religion enlists
science in order to secure and indemnify one of its two origins and insofar as
science, like religion, relies upon a kind of originary faith—that is, upon a
source it shares with religion.7
Every time one bears witness (see FK §49), Derrida argues, and even in sci-
ence, the truth is promised beyond all proof, all perception, all imitative mon-
stration. Even when I lie, Derrida argues, I ask the other to believe the other
who I am. This performativity or promise to tell the truth thus “conditions”
like a quasi transcendental not only all sincere declarations but all lies, and not
only all professions of faith within religion but every empirical claim in sci-
ence, and it amounts to saying, says Derrida: “Believe what I say as one believes
in a miracle” (FK §49). Thus we are called upon to believe every testimony—
every claim to the truth, every claim that one is telling the truth about what
one knows, believes, or sees—as an “extraordinary story” or a “miracle.” In an
essay in Sovereignties in Question entitled “Poetics and Politics of Witnessing,”
Derrida makes even clearer this relationship between belief and testimony.

“I bear witness”—that means: “I affirm (rightly or wrongly, but in all good faith, sincerely)
that that was or is present to me, in space and time (thus, sense-perceptible), and although
you do not have access to it, not the same access, you, my addressee, you have to believe me,
because I engage myself to tell you the truth, I am already engaged in it, I tell you that I am
telling you the truth. Believe me. You have to believe me.”8

7)
See the interview with Derrida published in Paper Machine, trans. Rachel Bowlby (Stanford,
CA: Stanford University Press, 2005), 116–17.
8)
“Poetics and Politics of Witnessing,” in Sovereignties in Question, ed. Thomas Dutoit and Outi
Pasanen (New York: Fordham University Press, 2005), 76.
190 M. Naas / Research in Phenomenology 39 (2009) 184–203

Hence religion has two sources, one of which is the common source of both
science and religion, a source that automatically, immediately, mechanically
divides at the origin into the two sources of religion, as well as into religion
and science. In order to indemnify and protect itself, religion must appropri-
ate the science that delocalizes it, abstracts it, and takes it away from “itself,”
that is, from what it considers to be its ipseity or its indemnified purity,
but not necessarily, notice, from one of its two sources, the one that is
common to both it and science. From a single common source, then, comes
two—the two sources of religion as well as the relation/reaction of religion to
science, an autoimmune reaction that can always be read as a reaction against
either the universalizing, abstracting tendencies of technoscience or their
common source.
Now, these three theses concerning the nature of religion, that is, concern-
ing religion as the object of Derrida’s analysis, cannot but have implications for
the form and writing of Derrida’s essay. The many textual and graphic dou-
blings and divisions we noted earlier, from the two titles and the two second-
ary sources to the two fonts and the two subheadings, can all be read as
reflecting the fundamental complicity and duplicity between religion and sci-
ence. But this fundamental duplicity within religion and this irreducible
complicity between religion and science also influence Derrida’s method or
approach to the question of religion “itself.” Note, for example, how Derrida
begins both the essay itself (FK §1) and the first section of the second set of
twenty-six sections (FK §27). In both places Derrida asks about the possibility
of talking “of religion? Singularly of religion, today” (FK §1), the possibility of
speaking “here and now, this very day,” of the “essence” of religion “with a sort
of religio-sity” that would not “introduce anything alien, leaving [religion]
thus intact, safe, unscathed ” (FK §27). But Derrida’s emphasis in both places
on the day, on what is happening to religion today, already suggests that the
“essence” of religion, the seemingly ahistorical essence of religion, must be
broached by means of the way religion is manifesting itself today, which is to
say, by means of the question of the relationship between religion and science.
To speak about religion, it might have been thought, one must speak with
respect, with scruple (religio), with a kind of religiosity that would leave reli-
gion, in accordance with one of its two sources, intact, safe and sound, indem-
nified or unscathed. But in what follows Derrida will precisely not leave religion
intact but will introduce all kinds of things into the discussion that might
seem to be foreign, indeed even antithetical, to it, beginning with science and
technology. Derrida seems to be suggesting that in order to speak of religion
today it is necessary precisely to interrupt the religiosity (though always in the
right way), to introduce what may seem foreign to religion in order to under-
 M. Naas / Research in Phenomenology 39 (2009) 184–203 191

stand everything that is happening to religion and in the name of religion

today. Derrida will do this by trying to pull together the two themes or two
sources we have identified, the indemnified presence of religion or of the reli-
gious community, an emphasis on the sacrosanct, the safe and sound, on the
one hand, and faith or belief, reliability and credit, the promise or the gage that
is at the origin of science as well as religion, on the other.
It is here that Derrida reminds us of something we all know about religion
today but that we might write off as inessential to religion itself, religion in its
indemnified essence, namely, that today’s “wars of religion” rely upon teletech-
noscience to an unprecedented degree. The participants in these wars thus
fight, as we know, not only over how one is to imagine, represent, or speak
about the celestial, about the nature of the heavens or the starry skies above,
but over control of those very skies. In other words, today’s wars of religion are
cyberspatialized as they have never been before, not simply played out or fought
out and then reported and broadcast across the world but waged through and
by means of teletechnoscience and its media. Digital culture is thus not simply
the means of reporting on these wars but the battleground itself. Derrida in
1994 gives several examples of this fundamental complicity between media
and cyberculture and today’s religious “manifestations,” from the Pope’s travels
around the globe to the worldwide distribution of his encyclicals to the Rushdie
Affair and the increasing use of cyberspace by fundamentalist groups. These
examples could be supplemented by an endless series of our own, from the
powers (political and otherwise) of televangelist churches in the U.S. and else-
where to the worldwide debates propagated through the media and the inter-
net over the publication in Denmark of what were taken to be offensive
cartoons of Mohammed, to cite just one example. In a particularly apt meta-
phor, Derrida speaks of the way in which “a celestial gaze—like a giant eye of
CNN—whether bestial or divine, monstrous, is always watching” (FK §27,
note), as if God’s Cyclopean eye in the sky, his celestial and synoptic vision,
had been replaced by a global network of satellites that ensure not simply that
“the whole world will be watching” every battle in these new wars of religion
but that this technological network will be the front line for these new wars.
This analogy between CNN and a divine, celestial eye and these references
to cyberculture as the place of religion’s—let me now emphasize the word—
manifestation allow us to read the contemporary relationship between religion
and technology in two directions; while such manifestation has never been so
worldwide, while teletechnoscience has never allowed religious manifestations
to be disseminated to such an extent, such manifestation, or spectrality
even, has always been the lifeblood of religion. The figures religion now takes
(tele-techno-media-scientific, capitalistic, politico-economic) are thus not at
192 M. Naas / Research in Phenomenology 39 (2009) 184–203

all original and without precedent. While cyberculture has amplified the
virtual powers of religious manifestations to a previously unknown degree,
this virtuality is not something absolutely new. As Derrida powerfully puts it,
“the ether of religion will have always been hospitable to a certain spectral
virtuality” (FK §27, note; my emphasis). Hence the so-called “return” to
religion that was the object of so much media attention during the early to
mid-1990s was not a simple return at all, argues Derrida, insofar as science has
always encrypted within it the original faith it shares with religion.
In what is perhaps the central argument or thesis of “Faith and Knowledge”
(my second thesis above), Derrida claims that religion must court the delocal-
izing and deracinating techniques and processes of technoscience in order to
protect, purify, and indemnify one of its two sources. In these new wars of
religion, teletechnoscience is used and put in the service of a return to what
would claim to come before science, namely, the original community, an
authentic and original relation to the divine, blood, nation, and autoch-
thony—all those things that the intrinsically universalizable movement of
teletechnoscience tend to disturb or dislocate. In order to immunize itself,
then, against what is considered to be outside it and threatening to it (the
West, modernity, Enlightenment, science), the religious community uses—
and oftentimes with extraordinary skill—the very instruments of teletechno-
science and cyberculture it is fighting against.
In FK §37 Derrida tries to give further definition to this relationship
between religion and science, that is, to the logic that leads religion to reject
science by means of the very techno-scientific means that religion rejects. He
speaks there of a machinal, automatic, unreflective movement that, like a
reflex, repeats a double movement of abstraction and attraction, a movement
that at once abstracts or uproots, that deracinates and attempts to universalize,
and that attracts or is attracted to the literal and the idiom, that reattaches onto
a putatively pure and indemnified identity. Yet this second moment of attrac-
tion or rerooting can take place only through a repetition and intensification
of the movement of abstraction and rerooting. Derrida will thus call this an
auto-immune auto-indemnification, that is, a self-indemnification, self-affirma-
tion, or self-protection that, through the very gesture of self-affirmation and
self-protection, opens the autos or the self up in a universalizing movement
to an outside that goes beyond the self and penetrates or compromises the self-
protection that was supposed to be reinforced. In this auto-immune auto-
indemnification, the immunization of the autos or the self is undone or
countered in the very movement whereby the autos or the self is exposed to the
other in its attempt to protect itself from the other.
 M. Naas / Research in Phenomenology 39 (2009) 184–203 193

This is the autoimmunity of religion: the attempt by religion to protect

itself by means of what always threatens to destroy it. Every auto-protection of
the indemnified, the safe and sound, must thus protect itself against its own
protection; religion must protect itself against the techno-science that protects
it. It is, Derrida concludes, “this terrifying but fatal [ fatale] logic of the auto-
immunity of the unscathed that will always associate Science and Religion”
(FK §37). This autoimmune indemnification of religion through science is
thus not something religion has a choice to do or not to do: it is, as Derrida
says in the sentence I just cited, fatale, that is, as Sam Weber has translated it,
“fatal,” that is, (at least potentially) destructive, disastrous, even deadly, though
also, since this is the other sense of fatale, fated or fateful, which is to say,
inevitable, unavoidable, and ineluctable.
It is important to note that this notion of autoimmunity, which had been
used by Derrida in earlier texts, such as Specters of Marx and Politics of Friend-
ship, is really given its first full treatment only here in “Faith and Knowledge.”9
Much of this work is done in a long footnote to FK §37 where Derrida first
recalls the ecclesiastical context of the term “immunity,” the notion of offering
someone safe haven, the status of exemption or inviolability, in a church, tem-
ple, or synagogue. Before diplomatic immunity or biological immunity, there
would have been the immunity offered by religion or by religious communi-
ties to those being pursued or persecuted. As for autoimmunity, this is obvi-
ously a much more recent term of the biological sciences. As Derrida will go
on to develop in even greater detail, in texts such as Rogues, in relationship to
democracy (and this is hardly a coincidence), autoimmunity has to do with
the way a living organism protects itself by attacking its own self-protection
and destroying its own immune defenses, thereby making it vulnerable to
what it might have otherwise combated. This attack on or protection against
one’s own mechanisms of protection is thus fatale—inevitable and always
potentially deadly—though also sometimes, for example in the case of
immuno-depressants, essential to the organism’s survival, essential to its accep-
tance of a graft or transplanted organ that will allow it to survive. Indeed
without autoimmunity, without this breach in the immunitary and self-
protective systems of the organism, the organism would have no future, no
possibility of a supplement that could either destroy or save it, that is, bring it
to an end or allow it to live on. Autoimmunity, we might say, is another name
for the aporia of the salut we saw earlier. Without the salut! as greeting and

9)
For a fuller development of the notion of autoimmunity in Derrida, see chap. 7 of my Derrida
From Now On (Bronx, NY: Fordham University Press, 2008).
194 M. Naas / Research in Phenomenology 39 (2009) 184–203

reception, without the welcoming of the supplement or graft of the other, the
self would remain within itself, safe and sound, absolutely protected, which is
to say, absolutely dead and without a future. Though the phantasm of absolute
immunity remains—indeed this is, in the end, the only phantasm—such
absolute immunity would be absolute death.
Starting, then, from the strictly ecclesiastical notion of immunity and then
the biological notion of autoimmunity, Derrida argues that a “sort of general
logic of autoimmunization” seems indispensable for thinking today the rela-
tion between faith and knowledge, religion and science, and the duplicity
of sources in general. For there are, it seems, always two sources or two
tendencies—a move toward indemnifying self-protection and another toward
autoimmune self-destruction, or put otherwise, a move toward complete self-
indemnification that, if successful, would put an end to the life of the self or
the community that is being protected, and an autoimmune movement that
at once threatens any organism or community at the same time as it allows it
to live on. These two movements are of course linked insofar as every attempt
by an autos or a self to protect itself ends up producing an attack upon the
self ’s own defenses or protections—whether for good or ill. This then means
that the self is never itself without its self-expropriation into the other and the
incorporation of that other into the self. In every attempt to purify, protect,
and indemnify itself, the self lives off what it is not, projecting a phantasm of
life protected from all death, and yet living on only by means of the supple-
ments of death: writing, science, telecommunications, the graft, iterability,
abstraction, the machine, and so on. Hence the immune/autoimmune reac-
tion of the self to what is outside it, and of religion to science, is absolutely
ineluctable, “automatic and machinal” (FK §37).
By allying itself with the “enemy,” by becoming hospitable to the antigens
that oppose it, religion grows and “swells [se gonfle],” says Derrida, with the
adversary’s powers (FK §37). Religion’s power thus grows and swells as it
appropriates the very technoscience that threatens it. In order to indemnify
itself, it must take on even more of what it opposes. Only this immunity and
autoimmune double reactivity can account for the resurgence of religion today,
that is, to use Derrida’s word, this déferlement, a word that is often used to
describe the breaking of waves or the spread of violence or a surge of troops.
Derrida thus speaks not of a “return” but of a “resurgence” of religion, a new
wave or surge in what has been a constant force or movement in European
cultures. Religion is thus allied today as it has always been with techno-
science, but because of the growth and multiplication of the latter, the former
 M. Naas / Research in Phenomenology 39 (2009) 184–203 195

has had to react with even greater force. Religion is thus at once what Derrida
calls mondialatinisation, that is, a globalizing process or, better, a “worldwide”
process (monde = world) that is linked, in a first moment, to the language in
which Christianity first spread across the globe, namely, Latin, and then in a
second moment to religion’s new lingua franca, namely, Anglo-American, and
the reaction, the war declared against that which gives religion this new power
only by dislodging it from its proper place, disrupting the relationship between
truth and place.
That which is truly “proper” to religion—its ineradicable allegiance and
allergy to science—is thus that which makes it fundamentally improper. If this
autoimmune auto-indemnification is indeed, as Derrida suggests, inevitable,
unavoidable, in a word, fatale, then what is “proper” to religion is both its
attempt to indemnify the first of its two sources, and the unavoidable expro-
priation of religion into what it is not. It is this expropriating movement, this
improper propriety—this autoimmunity—that leads to various attempts on
the part of religion to indemnify those things that are traditionally considered
proper to the self or the community: property, language, family, nature, blood,
and soil, and finally and perhaps most importantly for Derrida, life. If the
French word salut appeared earlier as the best way to identify one of the two
sources of religion, it is in the name of life that this salut is always sought, life
as what is restored in health or life as what is redeemed or saved. Religion thus
attempts to protect, indemnify, and augment life by means of the technologi-
cal supplement against which religion then reacts with the automaticity of a
machine. Derrida writes: “The reaction to the machine is as automatic (and
thus machinal) as life itself ” (FK §37). In the end, it is not only religion but
life itself that is autoimmune, life itself that reacts in an automatic, machinal
way, life itself that must now be thought in relation to the machine. Autoim-
munity is the place where life and the machine cross, where a biological pro-
cess happens with the regularity and automaticity of a machine, that is, with
a regularity and automaticity that is fatale, inevitable, predictable like a
machine, and deadly to any concept of a life before the machine, of a self with-
out the other, or of a community that would be absolutely pure and protected
in its self-identity.
Let me now try for a moment to put this in a very different context before
returning to “Faith and Knowledge” in order to see how the autoimmunity of
life and the three theses I have identified influence everything in Derrida’s
essay, right down to the syntax, letter, even the punctuation of his text. I turn
to this other context very briefly because it will help us understand, I think,
196 M. Naas / Research in Phenomenology 39 (2009) 184–203

the stakes of what is happening in “Faith and Knowledge” and, perhaps,

Derrida’s method and project more generally.10 Let us consider, therefore,
some of the things Derrida says about writing in his 1968 essay “Plato’s Phar-
macy,” that is, some of the accusations made against writing by Plato and
others following him.11 Each time, as we will see, the accusation against writ-
ing is made in the name of life, an accusation against what is mechanical and
deadly in the name of what is living, animated, and full of breath, an accusa-
tion against what is automatic and mechanical in the name of what is living
and spontaneous. If, as we will see, one of the essential attributes of life and
particularly human life is the living breath of speech, we will want to see how
Plato opposes this speech and this life, the spontaneity of speech and life, not
simply to death but to the automaticity of the machine. Now, to forestall a
potential objection, what I am looking for here is not some strict comparison
between Plato’s notions of speech and writing, on the one hand, and religion
and science on the other, but a structural analogy in Derrida’s reading. In both
cases what we will see is a contamination of something that would claim to be
“proper,” in the end, a necessary and unavoidable contamination of life, by
means of a technoscientific supplement that comes to inscribe repetition,
duplicity, and death into the heart of life or the living present.
In “Plato’s Pharmacy” Derrida makes a case that writing is condemned in
the Platonic corpus because it threatens the living present of speech, because
what is proper to living speech is contaminated, exposed, and expropriated by
means of the lifeless signs of writing, the living breath taken right out of speech
by the supplement—what Derrida shows to be a necessary supplement—of
writing. Whereas, on Derrida’s reading of Plato, logos or speech is “a living,
animate creature” (PP 79/89), writing “substitutes the breathless sign [le signe
essoufflé] for the living voice [la parole vivante]” (PP 92/104), “the passive,
mechanical ‘by heart’ for the active reanimation of knowledge, for its repro-
duction in the present” (PP 108/124). Writing is thus condemned for its
“breathless impotence [impuissance essoufflée]” (PP 115/130), for its substitu-
tion of the dead letter for the living voice. Writing is thus dangerous and to be

10)
Derrida himself encourages us from very early on to read him in precisely this way. See,
for example, his comments in Positions about stapling Writing and Difference into the middle of
Of Grammatology, and vice versa (Positions, trans. Alan Bass [Chicago: University of Chicago
Press, 1981], 4).
11)
“Plato’s Pharmacy,” in Dissemination, trans. Barbara Johnson (Chicago: University of Chicago
Press, 1981), 61–171; originally published as “La pharmacie de Platon,” in La dissemination
(Paris: Éditions du Seuil, 1972), 69–198. Hereafter cited as PP, followed by first the English
pagination and then the French.
 M. Naas / Research in Phenomenology 39 (2009) 184–203 197

avoided because of the way it threatens speech with thoughtless and impotent
repetition, in a word, with death, though also, and especially, because it is a
simulacrum that gives the appearance of life. Though it is but a “breathless
impotence,” writing is “something not completely dead: a living-dead, a
reprieved corpse, a deferred life, a semblance of breath [un semblant de souffle]”
(PP 143/165).
As Derrida will go on to argue in “Plato’s Pharmacy,” Plato condemns writ-
ing in order to protect and indemnify speech, the living breath of speech. In
order to do so, however, Plato must borrow the resources of writing, namely,
a system of differential and non-univocal signs that can never be controlled by
the living speech that Plato is trying to protect or indemnify. Thus, just as the
protection and indemnification of life in religion takes place only through the
autoimmune appropriation of the technoscientific supplement, so the life and
spontaneity of the logos is protected and promised only through the lifeless,
mechanical supplement of writing. In this text of 1968, then, life is related to
speech, breath, fertility, and importantly, spontaneity, while death is related to
writing, breathlessness, impotence, and interestingly, automaticity and the
machine: “Writing would be pure repetition, dead repetition that might always
be repeating nothing, or be unable spontaneously [spontanément] to repeat
itself, which also means unable to repeat anything but itself: a hollow, cast-off
repetition” (PP 135/155). And elsewhere, “Writing would indeed be the sig-
nifier’s capacity to repeat itself by itself, mechanically [machinalement], with-
out a living soul to sustain or attend it in its repetition, that is to say, without
truth’s presenting itself anywhere” (PP 111/127).
It is as if Derrida were saying back in 1968 in “Plato’s Pharmacy” that live
speech, inner speech, the living breath, is autoimmune; in order for speech to
express itself, in order for it to be itself, that is, in order for a signifier to be
understood as the same, it must court the powers of repetition and technique,
that is, in short, the powers of writing and the machine. Hence the only way
for speech to live on, the only way for it to grow, swell, or multiply beyond
itself, is for it to open itself up to what it is not, that is, to the powers of death
that go by the general name of “writing.” The only chance for living, organic
speech is thus contamination by the expropriating powers of writing; the only
chance for the organism, the powers of the machine. While this comparison
should be read, as I said a moment ago, less for what it says about the relation-
ship between speech and writing in Plato, on the one hand, and religion and
science, on the other, and much more for what is says about Derrida’s approach
to both, it is hard not to be impressed by a common matrix of terms and hier-
archies. In the end, both the Platonic discourse against writing and religious
198 M. Naas / Research in Phenomenology 39 (2009) 184–203

discourse against technoscientific modernity oppose life and the living breath
to the machine. Just as Derrida in “Plato’s Pharmacy” will thus end up figuring
a notion of archē-writing at the origin of both speech and writing, so, as we
will now see, he will in “Faith and Knowledge” figure a kind of mechanical
repetition at the origin—at the genesis—of religion.
In “Faith and Knowledge” Derrida suggests that the source of religion, the
“living” source of religion, cannot be thought without automaticity, repeti-
tion, and the machine, that is, without an immediate, automatic duplicity. At
the heart of religion, then, at its source, Derrida has identified not unity but
duplicity, and not life or purposeful creation, not living spontaneity and the
indemnification of life, but an automatic repetition and reaction, an autoim-
munity that turns every indemnifying movement against itself automatically
or mechanically—like a machine.
These references to life and the machine bring us back to the very beginning
of “Faith and Knowledge,” that is, as I have suggested, to its genesis. To con-
clude, let me move from what has been up until now a rather general overview
of the essay “Faith and Knowledge” to a much closer look at one brief passage.
Though it has the look of mere anecdote and background to the theses argued
in the essay, this passage can and, I believe, really must be read in the context
of the three theses I have just outlined and the comparison to “Plato’s Phar-
macy” we have just seen. The passage recounts a meeting in Paris between
Derrida and Maurizio Ferraris to discuss the theme of the upcoming confer-
ence in Capri where Derrida would deliver a first version of “Faith and Knowl-
edge.” Here are the crucial—though at first seemingly pedestrian—lines, first
in French and then in English translation:

Or il faut bien répondre. Et sans attendre. Sans trop attendre. Au commencement,

Maurizio Ferraris au Lutétia. “Il faut, me dit-il, il nous faut un thème pour cette rencontre
de Capri.” Je souffle, sans souffler, presque sans hésiter, machinalement: “La religion.” Pour-
quoi ? D’où cela m’est il venu, et oui, machinalement ? (§35)
But, one still must respond. And without waiting. Without waiting too long. In the
beginning, Maurizio Ferraris at the Hotel Lutétia. “I need,” he tells me, “we need a theme
for this meeting in Capri.” In a whisper, yet without whispering, almost without hesitating,
machine-like, I respond, “Religion.” Why? From where did this come to me, and yes,
mechanically? (FK §35)

Everything about this seemingly simple, deceptively simple passage is dou-

ble—at least double—and duplicitous. Notice, first, that even before the
beginning, there is response. Before the beginning there is a reference to
 M. Naas / Research in Phenomenology 39 (2009) 184–203 199

responding that is itself already divided—il faut bien répondre: on the one
hand, one must respond, one well has [bien] to respond, and on the other
hand, one has to respond well [bien]. One must respond, and one must
respond well, respond, for example, to the invitation to go to the Capri con-
ference and respond well to a request for a theme.
The beginning is thus preceded by a response, by a certain responsibility to
respond, a certain engagement or promise even before the beginning. Hence
the beginning is itself already a response to what will have come before it, a
response to or a call for a response that will have thus been the beginning. In
the beginning, then, was the response. Second, because of this irreducible
delay in the beginning, because the beginning is always already a response to
itself, just about everything in this passage is doubled—repeated, iterated, but
in each case with a difference: one must respond, says Derrida, “without wait-
ing / without waiting too long [trop]”; Ferraris says, “I need [il faut: what
is needed] / we need [il nous faut] a theme for this meeting in Capri,” and
Derrida responds, “in a whisper / yet without whispering,” “almost without
hesitating, machine-like [machinalement],” “Religion.” And Derrida then asks
himself, or responds to himself, “Why? From where did this come to me and,
yes, mechanically [machinalement]?”More or less everything is thus doubled
here, everything save the response “Religion,” which, as we have seen, will be
defined by Derrida in the conference to come in terms of two sources and an
irreducible relationship to the machine.
But how are we to understand the one repetition where a word is not simply
qualified but negated—“in a whisper, yet without whispering [ je souffle, sans
souffler]”? Everything revolves, everything pivots, around what appears to be a
way of qualifying the response and then negating that qualification. Since we
are talking about a certain genesis of religion, it appears that Derrida is repeating
here—in an almost machine-like way—the language of negative theology,12 or
else the kind of logic he finds in Blanchot’s narratives, where, as he writes

12)
In both “Denials” and the essay “Post-Scriptum,” Derrida evokes the necessary possibility of
the language of negative theology being mechanical. He writes, for example, “What we are iden-
tifying under these two words [i.e., negative theology], today, isn’t it first of all a corpus, at once
open and closed, given, well-ordered, a set of statements recognizable either by their family
resemblance or because they come under a regular logicodiscursive type whose recurrence lends
itself to a formalization? This formalization can become mechanical. . . .” Hence the discourse
of negative theology can always be thought—and we will recognize these terms from “Plato’s
Pharmacy” and other texts—to be “on the side of the empty and then of mechanical, indeed
purely verbal, repetition of phrases without actual or full intentional meaning” (“Post-Scriptum,”
200 M. Naas / Research in Phenomenology 39 (2009) 184–203

elsewhere, a “syntax of the without comes so frequently in Blanchot’s text

to neutralize (without positing, without denying) a word, a concept, a term
(X without X).”13 Je souffle, sans souffler might thus indeed be heard as a way of
qualifying Derrida’s response in one breath and then negating that qualifica-
tion in the next. But between these two breaths, between souffle and souffler, in
the silent but graphically visible scansion of this phrase in the form of
“X without X,” one might also hear not only negation but displacement and
dissemination. For when used with the preposition à and an indirect object,
as in “je souffle un mot à quelqu’un,” “je lui souffle un mot,” souffler does
indeed mean to whisper, to whisper, in this example, a word to someone. But
the verb souffler can also mean to “blow,” “blow into,” “breathe,” or even “ani-
mate,” while the noun souffle, as we just saw in “Plato’s Pharmacy,” can mean
“breath” or “living breath,” such that writing might be called but “a semblance
of breath [un semblant de souffle],” that is, a technical supplement that looks
like a living being but has had all the wind, all the breath, and all the life
knocked out of it.
Je souffle, sans souffler might thus suggest not whispering but breathing, per-
haps even the giving of life and breath, and then the immediate negation or
withdrawal of that life and breath—the ones with which, for example, accord-
ing to another tradition, it will have been infused at creation. As one popular
French translation of Genesis has it, God “created man from the dust of the
ground and breathed into his nostrils the souffle [the breath] of life.” Given the
fact that Derrida begins this entire passage, or almost, with the phrase “In
the beginning,” this reference to Genesis does not appear misplaced. Indeed
Derrida’s response to the question of Ferraris appears to be nothing less than a
replay of Genesis, but one that puts the machine in the place of the living
breath, technique in the place of live speech, a machine-like reaction in the
place of a spontaneous and purposeful response. If Derrida positions himself
at the origin or as the author of the theme of religion at the Capri conference,
he will have provided the theme without creating it or without being the ani-
mating force behind it. One might thus translate Derrida’s response: “I whis-
per, but without breathing, that is, without breathing life into anything or
animating anything”—“Religion.” That helps explain Derrida’s use—indeed
his double use—of the word machinalement. Because, for Derrida, there is no
such thing as live speech, an animating intention, or a spontaneous response

in Derrida and Negative Theology, ed. Harold Coward and Toby Foshay [Albany: State University
of New York Press, 1992], 295–96).
13)
Jacques Derrida, Parages (Paris: Éditions Galilée, 2003), 140.
 M. Naas / Research in Phenomenology 39 (2009) 184–203 201

that is not already from the beginning contaminated by repetition and tech-
nique, that is, in a word, by reactivity and the machine, there is no response
and no beginning, no genesis even, that is not in some way machine-like—and
thus double.
Hence Derrida repeats Genesis, this book that is so central to all three of the
West’s monotheisms, and he repeats it not once but twice, giving us, it seems,
both versions of the creation of mankind, the one that speaks of man’s creation
by means of the breath and the one that does not.14 For if, in Genesis 1, “God
created humankind in his image” (Gen. 1:27), he did so, it seems, without
breath, sans souffler, while in Genesis 2, he “formed man from the dust of the
ground, and breathed into his nostrils the breath of life; and the man became
a living being” (Gen. 2:7).15
Though Derrida seems to be recounting a more or less spontaneous, unre-
hearsed conversation with Maurizio Ferraris in the Hotel Lutétia, I think we
can see that this passage has been carefully crafted or staged around the two
versions of creation in Genesis, the one with breath and the other without, the
one with souffle and the other sans souffler. Derrida imitates Genesis, he repeats
it, he restages it, and perhaps—like the title “Faith and Knowledge” itself,
which, as we saw, comes from Hegel—he makes off with it, takes it as his own,
doubles and plagiarizes it. For as Derrida reminds us in his 1965 essay “La
parole soufflée,” souffler means not only to whisper, blow, or breathe but to
steal, swipe, or run off with, repeat, double, or spirit away. Writing in the

14)
My thanks to Louis Ruprecht for drawing my attention to this doubling.
15)
Here are the two versions of creation in Genesis in both a common French translation and an
English one: “Au commencement, Dieu créa les cieux et la terre. La terre était informe et vide, les
ténèbres couvraient l’abîme et le souffle de Dieu planait sur les eaux. . . . Dieu créa l’homme à son
image; à l’image de Dieu il le créa, homme et femme il les créa” (Gen. 1:1–2, 27). [In the begin-
ning when God created the heavens and the earth, the earth was a formless void and darkness
covered the face of the deep, while a wind from God [or the spirit or breath of God] swept over
the face of the waters. . . . So God created humankind in his image, in the image of God he cre-
ated them; male and female he created them.]
Hence the souffle, the breath, wind, or spirit [all possible translations of the Hebrew ruah] is
present at Gen. 1:2 but does not enter into the actual creation of the world, and particularly
not of man. But at Gen. 2:7 this breath or souffle plays an essential role in man’s creation: “Le
Seigneur Dieu forma l’homme avec la poussière du sol, et il lui insouffla dans les narines un
souffle de vie, et l’homme devient un être vivant” (Gen. 2:7). [The Lord God formed man from
the dust of the ground, and breathed into his nostrils the breath of life; and the man became a
living being] (La Sainte Bible [Paris: Turnhout], 1973; New Revised Standard Version, The New
Oxford Annotated Bible [New York: Oxford University Press, 1991]).
202 M. Naas / Research in Phenomenology 39 (2009) 184–203

context of a reading of the theatre of Antonin Artaud, Derrida brings several

of these meanings of souffler together in just a couple of sentences:

This derivation of force within the sign divides the theatrical act, exiles the actor far from
any responsibility for meaning, makes of him an interpreter who lets his life be breathed
into [insoufflé] him, and lets his words be whispered [soufflé] to him, receiving his delivery
as if he were taking orders. . . . To let one’s speech be spirited away [soufflé] is, like writing
itself, the ur-phenomenon of the reserve.16

One can begin to see the abyssal semantic resources of the word souffler. It is
as if someone in the wings of the text has whispered all these possibilities at
once to Derrida—possibilities within the French language but also in excess of
that language. As Derrida reminds us in “La parole soufflée,” the one who
whispers forgotten lines to an actor from offstage, what is called in English a
“prompter,” is called in French a souffleur, that is, one who speaks, who
breathes, who whispers, who simply reads or learns lines by heart and then
repeats them in a mechanical way—like a kind of writing. Words like, for
starters, “In the beginning . . .”
In this little passage of “Faith and Knowledge,” one word or one meaning
comes to supplement, augment, and make off with the others. What presents
itself as a mere repetition, a simple doubling, a word whispered from the wings
by a souffleur or prompter on the stage of the Lutétia, comes to take away the
breath of live speech and open up the uncontrollable dissemination of mean-
ing that Derrida associates with writing and survival. Though we might be
tempted to ask what meaning Derrida, the author of these words, “intended”
us to hear in this doubled and negated souffler, the “life” of this passage—like
the entirety of “Faith and Knowledge”—is beyond his control. It is impossible
to say for certain, it has always been impossible to say for certain, and reading
is at once condemned to this impossibility and freed for itself as a result. What
thus can be said, and said without hesitation, is that the undecidability of
souffler comes automatically, in a machine-like way, to contaminate the origi-
nal intention, the animating breath, with these various possibilities. Every-
thing about this single passage speaks about the duplicity of religion and its
necessary relationship to repetition and the technical supplement.
In the beginning, then, there was already the end of univocal meaning, the
multiplication of tongues, and the deconstruction of the Tower of Babel
(which is why, for example, I have been looking here so closely at the relation-
16)
“La parole soufflée,” in Writing and Difference, trans. Alan Bass (Chicago: University of
Chicago Press, 1978), 189; my emphasis on like writing itself.
 M. Naas / Research in Phenomenology 39 (2009) 184–203 203

ship between the French and the English translation of “Faith and Knowl-
edge”), the end of a single source and the beginning of repetition and
dissemination. In the beginning was the machine, and the machine will have
separated the author from the very beginning from the beginning, that is—all
duplicity intended—from himself and his words. The task, then, Derrida
seems to suggest, is to think religion in relationship not only to the originary
faith that is at the origin of every response but to the machine that always
conditions that faith and repeats every response. Hence the miracle and the
machine, originary faith and its repetition, life and the survival or living on
that conditions that life, the living on, for example, of a response that said yes
to an invitation, and the survival of that trace that repeats itself and, in repeat-
ing itself, engenders yet another response, yet one more response, one more
still, up until the end.