Bachelard, Lacan and the Impurity

of Scientific Formalization
TOM EYERS
Abstract:
This essay examines the conjunction of French historical epistemology
and Lacanian theory in postwar France. In particular, Lacan's account of
scientific formalization is scrutinized insofar as it develops aspects of the
prior epistemological research of Gaston Bachelard, whose innovative approach
to the problem of the nature and limits of scientific knowledge proved so
influential on the subsequent field of French structuralism. Lacan's reflections
on formalization will be shown, in contrast to Bachelard, to place an emphasis
on the constitutive and limiting role of language in its interaction with logical
and scientific projects. In asking how Lacan's structural psychoanalysis extends
and subverts the rationalist emphasis of French philosophy of science, I hope to
provide a new optic through which to assess the role of formalization in critical
theory today.
Keywords: epistemology, psychoanalysis, philosophy of science, Lacan,
Bachelard, French philosophy
i

This article explores Lacan's conception of scientific formalization

in its relationship to his wider reconstruction of psychoanalytic
theory. Lacan approached the question of formalization via a certain
conception of language and writing, and his ideas are implicated in
the broader project, conceived in post-war France and associated with
the early work of Alain Badiou, Jacques-Alain Miller and others,
to synthesize the insights of rationalist philosophy of science with
the concept of the subject of the unconscious as developed by
psychoanalysis. The work of the French philosopher Gaston Bachelard
is interrogated as a key precursor to Lacan's epistemological innovations
and to the post-Lacanian structuralism that, for a time, exerted a
Paragraph 35.3 (2012): 320-337
DOL 10.3366/para.2012.0062
Edinburgh University Press
wvirw.euppublishing.com/para

Bachelard, Lacan and Sdentifc Formalization

321

defining influence on French thought.^ In recent years, the work of

Alain Badiou in particular has resurrected many, of the questions that
underlay French structuralism, and attention to the early exchange
between French epistemology and psychoanalysis may help to clarify
the continuing centrality of such debates in theory today.
Bachelard held the inaugural chair of the History and Philosophy
of Science at the Sorbonne until his death in Paris in 1960, and his
legacy can be read as a crucial variable in the rethinking of questions
of scientific legitimacy in the years after the waning of existentialism
in France. If, for post-Heideggerian phenomenology and Sartrean
existentialism, the 'technical' objectifications of scientific practice
formed an object of critique, for the parallel field of French philosophy
of science it was the phenomenological emphasis on experience and
consciousness that represented a block to clear thinking. In particular,
Bachelard's influence can be said to have reinstalled at the very centre
of French intellectual life a rationalist concern for formalization.
Michel Foucault, and Bachelard's successors in the philosophy of
science, including Alexandre Koyr and Georges Canguilhem, all
took different things from Bachelard's rationalist theory of scientific
knowledge, but all three followed Bachelard in rejecting an empirical
or experiential account of the formation of knowledge. Instead, an
emphasis was placed on the constitutive role of theory in rendering
objects proper to the epistemological structures that different sciences
construct and reconstruct in perpetuity.
While reference is frequently made to Althusser's borrowing of
Bachelard's idea of the 'epistemological rupture', my focus here is
on the less-interrogated relation of Bachelard to Lacan. My aim is
to understand both the inheritance of Bachelard's rationalism within
the revision of psychoanalysis proposed by Lacan, and the new model
of formalization that emerges with Lacan out of that inheritance, a
model that provides a retrospective critique of Bachelard's insistence on
a rigid distinction between the ordinary language of communication,
and conceptuality and formalization as best rendered within the terms
of mathematics.
References to Bachelard, and particularly to his successors
Canguilhem and Koyr, pepper Lacan's seminar, and Lacan's increasing
concern in the 1950s and 1960s with the mathematical formalization
of analytic concepts found a sturdy and supportive context in
Bachelard's own insistence on the centrality of mathematics to
any modern philosophy of science. At one and the same time,
however, Lacan's project of a return to Freud seems to imply

322

Paragraph

a necessary coexistence between the progressive rationalization of

analytic concepts, and an account of how both language and the
subject ground, and perhaps disrupt, that process of rationalization.
In what follows, I will explore Jacques-Alain Miller's own, 'postLacanian' attempt to encompass the psychoanalytic subject within
the formalism of structure, while also signalling an alternative way,
present in the very late Lacan, to complicate the inviolability of
the conceptual surface of scientific knowledge so often imputed to
Bachelard, an inviolability that would find its first points of doubt
within Bachelard's own recognition of the impurity of scientific
objects. First, however, it is necessary to gain a firm grip on the
powerful ambiguities of Bachelard's epistemology, ambiguities that
provide the route through which Lacan could both appropriate and
subvert the model of formalization therein.
Bachelard Between Object, Concept And Signifier

In much of Bachelard's philosophical w^ork, the question of language,

and of writing, seems secondary, if not irrelevant. Bachelard had two
intellectual projects, received as distinct contributions to discrete sets
of questions. The central project that consumed Bachelard for most of
his professional life was an historical and epistemological enquiry into
the foundations of the physical sciences, perhaps especially chemistry.
But Bachelard also concerned himself with poetic imagination,^ and
it is in these works that one would normally look for his particular
account of language. Nonetheless, whilst problematic, Bachelard's
account of language in his epistemological writings has pertinence
to the conjuncture of formalization and the 'logic of the signifier' in
Lacan.
Responses to Bachelard have often taken the form of enquiries
into his relative debt to Descartes, and the question provides a useful
way into any more general account of Bachelard's brand of rationalist
epistemology. Mary Tiles, in her important study Bachelard: Science
and Objectivity, teases out the implications of Bachelard's shifting
relationship to Descartes, as when she writes:
If it can be presumed that the medium through which one is looking is perfectly
transparent and non-distorting, there is no problem in sorting out what part of
what is seen is due to the medium and what to the object viewed through it;
it can be assumed, with the naive realist, that things are just as they are seen to
be. If it cannot be presumed that the medium through which one is looking is

Bachelard, Lacan and Scientific Formalization

323

perfectly transparent and non-distorting, one cannot extract from wbat one sees
information concerning tbe sbape, size, color or surface features of tbe objects
viewed tbrougb it witbout learning wbat tbe cbaracteristics of tbe medium are.-*

With necessary caveats that I shall outline below, Bachelard adopts the
second perspective, whereby the philosopher of science must account
for both the delineation of a theoretical object of knowledge
distinct from any mere empirical object and the particular lens that
provides knowledge of it. That lens, for Bachelard and in contrast to
much Anglophone philosophy of science, is irrevocably historical and
situated. Moreover, the particular 'psychology' of the scientist or group
of scientists, inevitably immersed as they are in the vagaries of nonscientific influences, must also be accounted for, contributing as they
do to what Bachelard refers to as 'epistemological obstacles', obstacles
that may militate against the emergence of an epistemological shift or
rupture in the development of a science.
If, for Descartes, the reliability of the subject of enquiry is
guaranteed by the self-transparency of thought, for Bachelard thought
is inherently mediated both by sense experience in its potential
for mystification, and by the technical and epistemological lenses
through which the scientist defines his or her object. Bachelard
emphasizes, again in sharp distinction to the purity of a post-Popperian
Anglophone epistemology, the importance of the technical media
microscopes, ever more sophisticated laboratories that allow
scientists to sharpen the contours of their objects of inquiry. Against
the empiricist assumption that best scientific practice proceeds from
the observable and the given to the development of hypotheses and
theories, Bachelard, in typically rationalist fashion, considers the job
of the scientist to lie in complexifying the empirical by rendering it
amenable to conceptual qualification.
What is less typical, however, for a rationalism so often polemically
defined through its negative relation to empiricism, is Bachelard's
insistence that the subject or object is always-already mixed in with
its ostensible opposite. That is to say, there is no clean separation
for Bachelard between the empirical object of experience and
the theoretical object of science; their dialectical articulation and
distinction is the work in progress of science itself. Here, we get the
first glimpses of what will characterize the distinctly po5i-rationalist
character of Lacan's take on epistemology, but while for Bachelard this
muddying of the purity of the theoretical object occurs as a result of
the givenness of experience, for Lacan it is principally the logic of

324 Paragraph

the signifier that complicates the formation of any discrete object of

inquiry.
For Bachelard, nonetheless, a general dualism persists, organized, as
Tiles has most explicitly shown, around the categories of subjectivity
and objectivity. It is important to stress, nonetheless, that this duality
is not mappable on to the distinction found in Karl Popper and others
between science and pseudo-science, in so far as the latter distinction
presupposes a continuum between common sense as revealed in
experience and the domain of the scientific.'* To reiterate, it is
rather in the break with the empirically given that the beginnings of
any scientific process must be found for Bachelard. He renders this
starkly in his The Formation of the Scientific Mind (1938): 'Immediate

objective knowledge is necessarily incorrect by virtue of the fact that

it is qualitative. It produces error that must be rectified. It lays an
inevitable burden of subjective impressions on the object; objective
knowledge must be unburdened.'^ There are a number of intriguing
ambiguities in this short passage. For what is the status, precisely, of the
'object' referred to here? If, as Tiles insists, 'the pivotal distinction in
Bachelard's contrast is not between the empirically testable (falsifiable)
and the empirically irrefutable, but between subjective and objective',^
what status does incorrect, unscientific, intuitive but nonetheless
objective knowledge have for Bachelard, knowledge described above
as 'objective knowledge' to be 'unburdened'? Bachelard makes things
a little clearer when he adds: 'The object may not be designated as an
immediate "objective"; in other words, a march towards the objective is
not initially objective.'^
As Dominique Lecourt has observed, at play here is a deliberate
equivocation between different senses of the term 'object', between
the given object of experience with its burden of empirical
mystification and the object of science as it is produced through its
realization in scientific theory and praxis; or, alternatively, between
the different aspects of a single object, one empirical and the other
the product of scientific and theoretical labour. And as Lecourt also
notes, Bachelard will further underline the split in the object with the
coining of neologisms, such as 'super-object', defined as follows: 'The
super-object is the result of a critical objectification, of an objectivity
which only retains that part of the object which it has criticized.'^ It
seems to me that we must complicate the sense, then, of an absolute
break between the subjective and the objective in Bachelard if we
are to account for the persistence of error, or that 'part' of the
object that will eventually be jettisoned through a process of critical

Bachelard, Lacan and Sdentific Formalization

325

objectification, even as that part of the object persists in the domain

of the objective. That persistence, moreover, what Bachelard
considers the 'burdened' part of objective knowledge finds an
intriguing echo in Lacan's recognition of the continuity of truth and
misrecognition.
For Lacan, the Imaginary (with its object, the ego) and the Symbolic
(defined by the movements of the signifier) are replete with necessary
illusions. The Imaginary, as the seat of specular identification, is
premised on the constitutive misrecognition of the mirror image as
evidence of the subject's self-mastery, while the Symbolic relies on
an illusory sense of wholeness in order to function as the subject's
condition of possibility. For Bachelard above however, it is the object,
rather than the subject, which is split between its potential as a
conceptually or theoretically produced nexus of scientific labour and
its burden as being situated within the domain of the empirical.
Bachelard hints as much when he comments: 'It is also very difficult
to establish a hierarchy of error and to describe in an orderly way
the disorders of thought.' (FSM, 31) We can infer that any 'hierarchy
of error' would have to presuppose a clean separation between the
empirical and theoretical object, something that Bachelard in the
quotes above seems to refuse by splitting the object, even in its
'objective' guise, between its rational conceptual core and its empirical
burden. Tellingly, the 'epistemological obstacles' that Bachelard, in
the same book, argues must be overcome for the 'rational syntheses'
of scientific knowledge to be possible, are described in a similar
language of impurity: 'it is the nature of epistemological obstacles
to be intermixed and polymorphous' (FSM, 31). I'll turn now to
Bachelard's The New Scientific Spirit to further refine the 'impurity'
in question.
Published in 1934, The New Scientific Spirit seeks to interrogate
the implications for philosophy of science of the supersession of
Newtonian mechanics, among other leaps forward in geometry,
chemistry and elsewhere. Central to the book is the conviction that
developments in science pose a challenge to philosophy that should
be met by a nuancing of otherwise over-simple oppositions such as
that between rationalism and realism, between the observer and the
observed. Bachelard writes:
What does it mean to say that science can 'rectify' metaphysics? As an example of
what I have in mind, consider how 'realism' changes, loses its naive immediacy,
in its encounter with scientific skepticism. Similarly, 'rationalism' need not be a

326 Paragraph
closed system; a priori assumptions are subject to change (...). Science in effect
creates philosophy. Philosophy must therefore modify its language if it is to reflect
the subtlety and movement of contemporary thought.'"

One might certainly take issue with Bachelard's simplification of the

tradition of philosophical rationalism here, a simplification that seems
set up to allow science its role of constitutive clarification. Nonetheless,
it seems clear that Bachelard is not content with advancing an a priori
rationalism unsuited to the complexity and constitutive impurity of
the scientific objects of his time. As he writes of developments in
physics, 'even notions whose essence is geometric, such as position
and simultaneity, carmot be grasped in any simple way but only in
composite (...). Physics becomes a geometrical science and geometry
a physical science.' {NSS, 47) Underpinning this recognition of the
intermixing of previously discrete practices is a sense, touched on
above, of the theoretical and conceptual interchange between different
domains of scientific knowledge; thus, any reduction of the process of
scientific discovery to the constraints of empirical observation, or the
setting up of a symmetry between the theoretical and the empirical, is
ill-suited to understanding the complexities of the post-Newtonian
scientific context. Such an attempt, moreover, would implicitly
subscribe to the possibility of there being a general account of Science,
rather than a theoretically attuned attention to the particular regimes
of knowledge proper to each particular scientific practice.
I'd like to underline a certain vacillation in Bachelard here, between
his rejection of empiricism and his generally rationalist approach to
the predominance of theory on the one hand, and his complex but
suggestive account of the impurity of the objects of scientific enquiry
on the other, an impurity that points towards an ultimate rejection
of the very terms of the opposition empirical/conceptual. If nothing
else, the consequences of this vacillation render highly problematic
the reduction, made most recently by Paul Thomas in his critique of
Althusser, of Bachelard's epistemology to a thoroughgoing and easily
assimilated rationalism built on a clean opposition between science
and ideology. As Thomas sees it, 'Science according to Bachelard (...)
cannot be reached or judged by ideological means, and no ideological
path is ever about to lead to science, for the latter cannot be so much as
identified by any ideological mechanism.''^ While it is no doubt true
that 'absolutely' scientific practice, as a goal to be worked towards, must
be rigorously distinguished for Bachelard from the 'epistemological
obstacles' of non-scientific or ideological thinking, the quotes above

Bachelard, Lacan and Scientific Formalization 327

suggest, at the least, that the process of separating the two is one of
coming up against the persistent impurity of the objects of possible
knowledge.

Between Bachelard and Lacan

In a manner formative of Lacan's later fascination with mathematical

formalization, Bachelard insists on the sovereignty of mathematics
as the basis for rational knowledge. In this sense, mathematics, for
Bachelard, should be seen less as a medium for priorly constituted
scientific inquiries, and more as that which is formative of the
scientific as such. As he writes in The Formation of the Scientific Mind,

'Mathematism is not descriptive but formative. The science of reality is

no longer content with the phenomenological how: the mathematical
why is what it seeks.' (17) Under the sway of developments in physics,
and perhaps especially quantum mechanics, where the reliability of
observation was ever more under question, Bachelard emphasized the
constitutivity of mathematics over and above its role in expressing or
quantifying that which is observed.
I'd like to draw an initial, tentative parallel here with Lacan's
treatment of the signifier. If, even in the structural linguistics of
Saussure, the ultimate function of the signifier is to couple with a
signified, Lacan's innovation was to insist on the material isolation of
the signifier from sense, expressed most famously in the concept of
the 'letter'.^^ In so far as the signifier has the capacity to uncouple
from its representative function, it assumes much the same role as
Bachelard imparts to the mathematical, forming what it is only
expected to describe. For Lacan, that is, the signifier constitutes the
immovable horizon of all human activity, assuming a quasi-ontological
status, and there is, particularly in the critique of the psychological
subject common to Bachelard's reflections on mathematics and Lacan's
reflections on the signifier, a shared concern for the structural
constitutivity of elements previously only considered reflective of
prior experience or observation. If Lacan will most obviously absorb
Bachelard's influence through his attempt to formalize psychoanalytic
ideas through the creation of 'mathemes', this broader sense of
structural constitutivity is crucial too in assessing the crosscurrents
between psychoanalysis and philosophy of science.
The most obvious shared concern in both Bachelard's philosophy of
science and Lacan's return to Freud is, nonetheless, the critique of the

328 Paragraph

subject as formulated by psychology and philosophical anthropology,

a critique that, for Bachelard especially, emerges around the abovementioned question of mathematics in its relation to scientific
knowledge. In an especially suggestive passage, Bachelard writes:
It bas been repeated endlessly tbat matbematics is a language, a mere means of
expression. People bave grown used to tbe idea tbat matbematics is a tool wielded
by a self-conscious mind, mistress of a set of ideas endowed with prematbematical
clarity. (...) Tbe new science sbuns naive images, bowever, and bas in a sense
become more homogenous: It stems entirely from matbematics. {NSS, 55)

Here, perhaps, one of the most important, if underlooked, potential

sources for what would become Lacan's critique of the subject is
crystallized. But despite this obvious common ground, Lacan will also
insist on the limits of formalization, expressed most pungently in the
20'*^ Seminar with the claim that 'The real can only be inscribed
on the basis of an impasse of formalization.' At one and the same
time, mathematical formalization is essential for Lacan in rendering
the a-priority and non-empirical status of psychoanalytic concepts.
Formalization, in turn, becomes an object of psychoanalytic interest
in its very failure. What indexes this failure for Lacan is the status
of mathematical formalization as a form of writing, as interlaced with
the logic of the signifier. As he notes, 'That is why I thought I
could provide a model of it [the real as impass] using mathematical
formalization, inasmuch as it is the most advanced elaboration we have
by which to produce signifierness.'^^
Here, the gulf between the thinking of the relation between
mathematization, language and knowledge as undertaken by Bachelard
and Lacan becomes clear. For while Bachelard and Lacan both explore
the non-expressivity of mathematics, its inherent complication of any
mirror model of representation, they differ in the precise relationship
of that complication to the question of language more generally.
For Lacan, the signifier, as the immovable horizon of the subject
of all knowledge, is not simply a mystifying tool of intuition. Both
mathematical formalization and the signifier, rather, reveal for Lacan
the inherent, and indeed constitutive, impasses that de-totalize, render
impure, both the seeming purity of mathematical number and the grip
of the signifier qua 'big Other'. As Lacan continues in his 20* Seminar,
'it was not in vain that I eventually came up with the inscriptions a,
the $ of the signifier, A (...). Their very writing constitutes a medium
[support] that goes beyond speech, without going beyond language's
actual effects.' (93)

Bachelard, Lacan and Scientific Formalization 329

'
^'
In Other words, there is a domain for Lacan, that of 'writing', that
can encompass both the centrality of formalization and its necessary
impasse, and that, further, might figure their very coincidence. I will
return to this problem of writing in Lacan presently. Bachelard, by
contrast, often seems to consider language to be a mere nuisance, in
need of perpetual rectification. In The Formation of the Scientific Mind,
Bachelard refers to the 'artifices of language' and links this artifice to
the mere sense or intuition of scientific progress that may, in fact, act
as a block to genuine advances. In a striking critique of the pretensions
of psychology to scientific status, be writes: 'Whenever a known
phenomenon is designated by a learned name, lazy thinking gets a
real sense of satisfaction. Certain medical diagnoses and psychological
insights that make play witb synonyms could easily provide us w^ith
examples of those verbal satisfactions.' (104)
One may benefit from playing especial attention to Bachelard's
own choice of words bere. Not only is language figured as a
salve to lazy thinking, as tbe recourse of tbinking predicated on
a lack of innovation, but it is described in terms of satisfaction,
of enjoyment, of 'play' in sbort, as sometbing frivolous. Tbe
reduction bere of language to science's opposite belies Bacbelard's
wider urge to destabilize canonical distinctions in pbilosophy of
science. Tbis tension, between tbe reductive reification of tbe
oppositions subjective/objective and empirical/rational oppositions,
and recognition of tbeir constitutive impurity, is nicely exemplified
in tbe following:
We must constantly strive towards desubjectification if we are to Uve and relive
the instant of objectification, if we are to remain forever in the nascent state of
objectification. The mind that psychoanalysis hasfreedfromthe twofold slavery of
subject and object can savour the heady delight of oscillating between extraversion
and introversion. An objective discovery is at once a subjective rectification. {FSM,
246)
Wbat migbt seem initially to be a simple plea for tbe reduction of tbe
subjective influence on tbe search for objective trutb is, in fact, mucb
more productively ambiguous. First, Bacbelard seems to abandon bis
implicit recognition, as commented on above, of tbe impurity of tbe
objective, tbe potential for error to persist witbin tbe field of tbe
objective. Instead, we get a sense of immanent contradiction, a kind
of seesawing between tbe objective and tbe subjective. Also present
in tbe quote, bowever, is sometbing of tbe worked upon impurity
outlined above. Tbus, objectification is a 'nascent state', one constantly

330 Paragraph

worked towards, and a state implying a symmetrical rectification in

the subjective. And yet, desubjectification, the total evacuation of
the subjective, is posited, in the very first sentence, as a seeming
condition of the 'instant' of objectification. At one and the same time,
then, Bachelard seems to wish for the disappearance of the subject
as a condition of the objective, even as the two poles are situated
in a position of mutual interdependence. We can ask of Bachelard,
therefore, how the objective could persist if 'desubjectification' were
ever fully achieved. Might this be a process without end, a perpetual
oscillation between poles that, given the comments above on the
impurity of the object of objectivity, are never finally and absolutely
distinguishable?
Bachelard, in sum, provides both a rationalist account of the
conceptual autonomy of science, and the beginnings of what I would
like to call a post-rationalist critique of the sustaining binaries of
rationalism itself It is necessary to read Bachelard against himself,
to perform a 'symptomatic reading' of his writing, in order to
extract this critical destabilization of rationalist approaches to the
justification of knowledge. It is precisely this second stream of thinking
that sets the stage for the attempt within the broader project of
French psychoanalytically informed structuralism to provide a more
theoreticaUy rigorous account of the signifier and the subject in
their disjunctive relation with the 'objective' domain of scientific
knowledge. Next, I'd like to focus on how aspects of Lacan's late
thinking significantly further, and yet subvert, the ambiguities of
Bachelard's epistemology as underlined above.

Lacan and the Impurity of the Signifier

If Bachelard's partial recognition of the persistence, even constitutivity,

of subjective error within the 'objective' is undermined by his
reductive accounts of language and the subject, Lacan would transpose
the sense of a non-contradictory and constitutive impurity into every
facet of his metapsychology. Through a reading of key passages
in his unpublished 24'^'^ seminar, I hope to demonstrate here how
the complexity of Lacan's account of the priority of the signifier
further undermines the distinction between the subjective and the
objective already rendered shaky in aspects of Bachelard's epistemology.
While, for Bachelard, the figurative character of language renders it a
more than imperfect means of rendering the conceptual autonomy

Bachelard, Lacan and Sdentifc Formalization

331

of scientific knowledge, Lacan recognizes the inevitabihty of the

interpntration of the vicissitudes of the signifier in the domain of
the objective, an interpntration that is not, in itself, a threat to the
'scientific' as such.
What links Lacan's reflections on the coincidence of the 'impasse'
of the signifier and the centrality of formalization to Bachelard, and
what makes a comparison of their approaches so suggestive, is their
shared concern for the mathematical as the site of formalization. Before
drawing out the fuU implications of the comparison with Bachelard,
it is necessary to outline the ways that Lacan, in his very late work,
rethought the relationship between language, formalization and a
certain conception of writing.
Lacan's 24 seminar, one of his very last, builds upon the insights
of the previous few years, where the psychoanalytic concept of the
'symptom' as a knot of occluded meaning to be interpreted was
replaced by the 'sinthome', as a material signifier lending consistency
to the subject. Early comments from the 1950s on the symptom
significantly prefigure the concept of the sinthome, emphasizing as
they do the particular problem that the symptom poses for the
production of sense. In 1957, Lacan wrote: 'The fact that symptoms are
symbolic is not the whole story (...), their use as signifiers distinguishes
them from their natural meaning.'^'* Their 'natural meaning' refers to
what, elsewhere, I have called the signifier's being 'in-relation', as being
in a situation of co-determination.^^ With the symptom, by contrast,
there is a certain disconnection of the signifier from its determining
others, leaving it in isolation, and thus somewhat askance from the
'natural meaning' that it is so often assumed it is the signifier's role
to facilitate. To recognize Lacan's insistence on the materiality of the
isolated signifier is also, crucially, to recognize his transcendence of the
influence of Saussure, whose account of the generation of linguistic
meaning relies on the idea of the signifier as only ever existing through
its relations; as Lacan comments in the 24 seminar, 'what's annoying
is that all we ever do is involve linguistics. I passed that way, but I didn't
stop there.'^^
In the second half of the seminar, given the title 'Towards a New
Signifier' by Jacques-Alain Miller, this materiality of the signifier, its
disconnection from the relationality of sense, is in turn interrogated
as the specific 'object' of a psychoanalytic theory of writing. There,
Lacan comments: 'A new signifier that wouldn't have any kind of
meaning, that would perhaps be what would open us up to what, in my
clumsiness, I call the real. Why couldn't we try to formulate a signifier

332 Paragraph

that, contrary to the usage that we currently make of it, would have
an effect?' It is crucial to note here the immanence of the signifier,
especially in its dimension of meaninglessness, to the Real, as what
Lacan elsewhere designates as the conjunction of contingency and
impossibility. Against any tendency to cleanly separate the Real from
the Symbolic, here the Real exists within the Symbolic as the signifier
in its isolated, non-sensical state. Whereas for Bachelard the subjective
persists within the objective as that aspect of the object 'burdened'
with the empirical, the domain of contingency and non-sense not only
persists within the 'objective' domain of structure for Lacan but, rather,
exists in a state of definitional dependence upon it. We can furnish this
point by underlining the continuum between this aspect of the signifier
and mathematics for Lacan; whereas Bachelard will posit mathematics
as a formative conceptual domain distinct from the impurity of the
subjective and the experiential, for Lacan the signifier as it figures
in writing and the mathematical are fundamentally interlaced. As he
writes, 'One tries to reach language by writing. And writing doesn't
give us anything but mathematics, where it's a matter of working by
formal logic, that is, by the extraction of a certain number of things
that we define as fundamental axioms. Thus we extract letters.'
When read together with his attempt to articulate the 'impasse'
of formalization with a new conception of writing as distinct from
speech, this quote emblematizes the singularity of Lacan's thinking
on the formal linkage of formalization and language. Lacan, it seems
to me, offers here a more expansive treatment of the constitutive
impurity of the objective/subjective distinction, bridged by a focus
on the signifier or 'letter', than is present in latent form in Bachelard's
reflections on the impurity of the production of objective knowledge,
even as the latter's emphasis on impurity seems decisive in laying the
ground for Lacan's innovations. When Lacan writes of 'the extraction
of a certain number of things that we define as fundamental axioms',
he raises psychoanalytic conceptuality to the level of the axiom in
mathematics, whereby any act of interpretation is grounded in a
priori constructs that are particular to psychoanalysis as a domain of
knowledge. At the same time, Lacan associates this axiomatic character
of psychoanalytic conceptuality with 'letters', or signifiers as they are
abstracted from relations of meaning. The 'letter', defined by Lacan
twenty years earlier as material and indivisible, is the signifier as it exists
outside the production of sense, and it is this dimension of language,
its material persistence, that renders the link between the matheme
and language concrete, while gesturing towards the generation of a

Bachelard, Lacan and Scientific Formalization

333

'new' signifier co-extensive witb a new take on tbe stakes of scientific

formalization.
.
t\
Instead of subscribing to tbe either/or of formalization as the
key to any successful grounding of psycboanalytic claims, or the
alternative move away from tbe sciences in favor of a bermeneutic
or purely textual interpretivism, Lacan locates witbin the process of
formalization itself its own immanent, and constitutive, failure, indexed
to tbe signifier as its formative, and deformative, ground. In tbe
24* Seminar, Lacan w^ill evoke formalization tbrougb tbe concept of
'metalanguage', or a language tbat might step outside the vicissitudes
of tbe signifier; Bacbelard's dream, we migbt say. Lacan writes: 'There
is an embryo of metalanguage, but it always goes off tbe skids for the
simple reason tbat all I know about language comes from a series of
actual [incarnes] languages.' Any attempt to surpass language in favour
of a pure conceptuality, tbat is to say, must contend witb tbe elements
tbat would make up tbat conceptuality, namely signifiers composing
'actual languages', tbe movements of wbicb resist any transcendental
purity.
Here, it is useful to briefly compare tbe argument made by Miller
in 'Suture'. Tbis article, published in tbe Cahiers pour l'Analyse,
foregrounds the destabilizing yet formative insistence of tbe nonidentical subject, witb this seemingly non- or a-subjective insistence on
the signifier in the Real as tbe point of impossibility in processes
of formalization. Miller argues, tbrougb a reading of Frege, tbat tbe
succession of self-identical numbers must rely on zero as its nonidentical foundation. For tbe very self-identity of numbers to be
meaningful, that is to say, tbey must refer back to a non-identical
element tbat acts to negatively determine tbem. Extrapolating from
tbis specific example. Miller writes:
In effect, wbat in Lacanian algebra is called tbe relation of the subject to tbe field of
tbe Otber (as tbe locus of trutb) can be identified witb tbe relation wbicb tbe zero
entertains witb tbe identity of tbe unique as tbe support of trutb. Tbis relation,
in so far as it is matrical, cannot be integrated into any definition of objectivity
tbis being tbe doctrine of Lacan. Tbe engendering of tbe zero, from tbis notidentical witb itself under wbicb no tbing of tbe world falls, illustrates tbis to
you."

Tbus, Miller draws a formal relation between tbe function of zero as

tbe non-identical ground for tbe self-identical or 'unique' succession
of numbers, and tbe function of tbe subject in Lacan, conceived of
bere as providing tbe non-identical ground for tbe 'Otber (as tbe locus

334 Paragraph

of trutb)', or tbe domain of tbe signifier. Wbile it is certainly true

tbat, especially in tbe 11* Seminar, Lacan discusses tbe subject in tbe
terms of an occluded 'cause' of tbe Symbohc, punctuating tbe battery
of signifiers wbile forever 'fading' before tb signifiers tbat will come
to represent it,^^ Lacan's remarks in tbe 24* Seminar seem to suggest a
displacement of tbis 'eccentric element' onto tbe signifier itself, wben
taken in its self-identical, abstracted, isolated guise as 'letter'.
A little further on from bis comments on tbe possibility of a new
signifier apart from meaning, Lacan's remarks bear on tbis seeming
replacement of tbe non-identical subject as constitutive cause and tbreat
to formalization witb tbe signifier as it is abstracted from meaningful
relation. Lacan writes, eUipticaUy, tbat
There's only one case in which I risk working in the direction of metalanguage.
The metalanguage in question consists of translating Unbewusst by une-bvue. It's
absolutely not the same meaning. But it's a fact that as soon as he sleeps, man
blunders [une-bvue] with all his might (...). What Freud said, and what I mean, is
this there isn't, in any case, a waking up. Science can only be invoked indirecdy
in this case. It's a waking up, but a difficult and a suspect one. One is only sure
that he is woken up if what is presented and represented doesn't have any meaning
at all.

Leaving aside tbe elaborate wordplay tbat bad become a mainstay

of Lacan's seminar by tbis point, wbat is initially striking in tbis
passage is tbe reference to science as a 'waking up', even if a difficult
and 'suspect' one. At least superficially, Lacan is close to Bacbelard
bere, whose empbasis on tbe production of scientific knowledge as a
process of unburdening tbe object of knowledge from its encrusted
mystifications bears a similar sense, perversely Heideggerian despite
tbe opposite intentions, of a waking up into trutbfulness. But Lacan
frames bis own account of a scientific waking up witb a reference
to metalanguage, sometbing be bad previously deemed impossible. If
science, we infer, can be considered a metalanguage, tben it is only
balf possible, or 'indirect', and subject ultimately to tbe fact tbat, as
Lacan suggests was already stated in Freud, 'tbere isn't (...) a waking
up', or tbere isn't any final clean break from tbe 'blunders' of tbe
unconscious. Crucially, I think it's important to read Lacan bere not
as a standard bumanist sceptic, cautioning science in its ambitions
to acbieve tbe purity of metalinguistic clarity; ratber, tbe insigbt of
Lacan's comments resides in tbe idea tbat, if science is to approacb
a position of sucb self-transcendence, it may only do so 'if wbat is
presented and represented doesn't bave any meaning at all', if, in otber

U - . .."-^ -s-, v s . 'I

Bachelard, Lacan and Sdentific Formalization 335

words, it recognizes simultaneously its total reliance on the signifier

that which presents, and represents and the possibility, internal to
signification, of signifiers, those axiomatic 'letters' referred to above,
that do not have a relation to meaning as it is normally conceived.
To reiterate, Lacan's comments here retrieve an aspect of language
for formal conceptuality in the face of Bachelard's earlier dismissal of
language, a retrieval crucial to the 'objectivity' that Miller above argues
must rest on the non-identical subject. Here, however, it is less the
non-identical subject that might ground this conceptuality, and more
the self-enclosed 'letter', an element that is, by definition, a-subjective,
even as it forms the material ground upon which the subject of the
unconscious may cohere. One year prior to the 24* Seminar and as
mentioned briefly above, Lacan had developed his striking concept of
the 'sinthome', as the symptom abstracted from any regime of meaning
or analytic interpretation, persisting as a knot of sense-less jouissance
at the eccentric centre of the subject. But instead of reverting to a
Freudian energetics, and thus conceiving of this binding agency of/in
the subject in terms of energy or libido, Lacan insists on the role of the
signifier as it exists in isolation, detached from the psychic architecture
of meaning, in providing the vehicle for this subjective consistency.
In other words, an element, the signifier, normally associated, as
in Bachelard above, with the confusions of meaningful, empirically
directed discourse, becomes the very abstract (even objectai) condition
for the persistence of the subject itself. As Lacan puts it in his 23'^''
Seminar, this consistency can only be understood as 'an ex-sistence
(...) which for its part belong to the Real which is its fundamental
character'. ^^
The concept of the Real, at this stage in Lacan's teaching, had
become associated with the particular kind of formalization proper
to psychoanalytic conceptuality, a formalization that recognizes the
coincidence of Symbolic consistency and its immanent tendency
towards dissolution. Lacan's concept of the sinthome places this
coincidence of formation and deformation, of creation and potential
destruction, at the centre of the subject. A subject's 'sinthome', that
is to say, coheres as a result of the isolated signifier's self-consistency,
its abstraction from relation, but that very isolation is also a condition
of great precarity; in analytic practice, by consequence, displacing a
subject's 'sinthome', far from being an unproblematically therapeutic
manoeuvre, places that subject at great risk of psychic collapse.
For our purposes, Lacan's conceptualization of this 'new signifier'
significantly complicates Bachelard's relegation of language, while

336

Paragraph

fundamentally reintroducing the subject as a condition and consequence

of formalization, in precisely the way that Bachelard, in his emphasis
on the desubjectification proper to scientific conceptuality, rejects.
In a theoretical move redolent of the topological complexity
that defined his seminar at the time, Lacan loops the 'formal'
elements of psychoanalytic conceptuality, the axiomatic letters so
closely related to the matheme, back into the concept of the
subject, such that the very clean separation of the subject from
the formal movement of conceptual elements, a separation that
underlies Bachelard's epistemology, is problematized. Miller's attempt
to ground logical consistency on the non-identity of the subject, while
suggestive, can be supplemented by reference to this late attempt
by Lacan to theorize a formally sense-less signifier; in so doing,
Lacan recognized the self-identical, 'formal' elements present in the
very subject itself, thus significantly complicating Miller's dualistic
attempt to think the reliance of self-identity on the non-identical,
and extending the productive ambiguities in Bachelard's theorization
of the impurities of the scientific object. When read critically
together, Bachelard and Lacan, and the post-Lacanian extension of
psychoanalytic ideas represented by Miller, provide an internally
heterogeneous but nonetheless related attempt to undermine clean
distinctions between formalized knowledge and its hidden subject. It
is, nonetheless, their crucial differences, especially around the relative
status of language, that pose anew the determining problems of French
structuralism.

NOTES
1 It was the journal Cahiers pour l'Analyse, published in Paris between 1966 and
1969, that hosted much of the most experimental writing at the intersection
between psychoanalytic theory and philosophy of science, including early
work by Alain Badiou, Jacques-Alain Miller, Jean-Claude MOner and others.
See http://cahiers.kingston.ac.uk for the complete text of the journal in
French.
2 Gaston Bachelard, The Poetics of Space [1957], translated by Maria Jolas
(Boston: Beacon Press, 1994) and On Poetic Imagination and Reverie, translated
by Colette Gaudin (New York: Spring Publications, 1994).
3 Mary Tiles, Bachelard: Science and Objectivity (Cambridge: Cambridge
University Press, 1994), 39.
4 Karl R. Popper, Conjectures and Refutations: The Growth of Sdentifc Knowledge
[1963] (New York: Routledge, 2003).

Bachelard, Lacan and Scientific Formalization 337

5 Gaston Bachelard, The Formation of the Sdentijk Mind : A Contribution to a
Psychoanalysis of Objective Knowledge (1938), translated by Mary McAUester
Jones (Manchester, Clinamen, 2002), 210. Further page references are
included in the text,'nd foUow the abbreviation FSM.
6 Tiles, Bachelard, 53.
7 Bachelard quoted in Dominque Lecourt, Marxism and Epistemology: Bachelard,
Canguilhem and Foucault, translated by Ben Brewster (London: New Left
Books, 1975), 52.
8 Lecourt, Marxism and Epistemology.
9 Bachelard quoted in Lecourt, Marxism and Epistemology, 52.
10 Gaston Bachelard, The New Scientific Spirit, translated by Arthur Goldhammer
(Boston: Beacon Press, 1984), 3. Further page references are included in the
text, and follow the abbreviation NSS.
11 Paul Thomas, Marxism and Scientific Socialism: From Engels to Althusser (New
York: Routledge, 2008), 117.
12 Jacques Lacan, 'The Instance of the Letter in the Unconscious, or Reason
Since Freud' in crits, translated by Bruce Fink (New York: W. W Norton,
2006), 412-41.
13 Jacques Lacan, The Seminar of Jacques Lacan Book XX: Encore, edited by
Jacques-Alain Miller, translated by Russell Grigg (New York: W. W. Norton,
1998), 93.
14 Lacan, 'Psychoanalysis and its Teaching' in crits, 36483.
15 See my 'Psychoanalytic Structuralism and the Cahiers pour l'Analyse' in
Angelaki : Journal of the Theoretical Humanities, forthcoming 2012.
16 Jacques Lacan, session of 17 May 1977, in Seminar 24: L'insu que sait de I'unebvue, s'aile mourre, edited by Jacques-Alain Miller, unofficial translation by
Dan Collins for personal use.
17 Jacques-Alain Miller, 'Suture: Elements of the Logic of the Signifier' [1966]
in The Symptom: Online Journal for Lacan. Com (Winter 2007).
18 Jacques Lacan, The Seminar ofJacques Lacan Book XI: The Four Fundamental
Concepts of Psychoanalysis, edited by Jacques-Alain Miller, translated by Alan
Sheridan (New York: W. W. Norton, 1998), 53-67.
19 Lacan, session of 9 December 1975, Seminar 23: Joyce and the Sinthome,
unofficial translation by Cormac Gallagher.

Copyright of Paragraph is the property of Edinburgh University Press and its content may not be copied or
emailed to multiple sites or posted to a listserv without the copyright holder's express written permission.
However, users may print, download, or email articles for individual use.