Journal for the Theory of Social Behaviour 38:4

0021–8308

Social Representations and Repression: Examining

Michael
Original
Social
Blackwell
Oxford,
Journal
JTSB
© 2008Representations
1468-5914
0021-8308
XXX Billig
Article
The
UK
for
Publishing
the
Author
Theory ofand
Journal
Ltd Repression
Social
compilation
Behaviour
© The Executive Management Committee/Blackwell Publishing Ltd. 2008

the First Formulations of Freud and Moscovici

MICHAEL BILLIG

Questions of historical origins are central to the theory of social representations.

The
ABSTRACT
these
first
Key
Repression,
the concept
formulations
words:
English
notions,
of
social
edition
especially
“complex”.
ofrepresentations,
the
of Moscovici’s
key
in
Moreover,
psychoanalytic
relationclassic
to
nominalization,
psychoanalytic
Moscovici’s
work
concept
on claim
the
of
psycho-analysis,
theory
“repression”.
social
thatcould
psychoanalytic
representation
never
Some
Freud,
becomparisons
univocal
complex
theory
of psychoanalysis
is
forSocial
toitare
be
incorporated
made
found
Representations
enables
between
in Freud’s
theusvoices
Freud’s
tofirst
and
reflect
formulations.
ofRepression:
first
patients
on formulations
the historical
andItexamining
is
their
suggested
oforigins
representations
“repression”
firstof
that
formulations
psychoanalytic
some
andofMoscovici’s
of
thethe
world.
processes,
ideasAn
first
and
examination
formulations
of
which
social
Moscovici
representation
ofStudies
of “social
attributes
onrepresentation”.
Hysteria
theory
to the
itself.
reveals
passage
Moscovici
further
of psychoanalytic
multivocality
claimed thatas
science
ideas
Freud
touses
was
commonsense,
both
bothunivocal
the “action”
canand
belanguage
abstract
found occur
and,
of ordinary
in
in the
these
early
liferespects,
and
history
the reified,
itofdiffers
psychoanalysis.
nominalized
from the social
This
language
can
representations
beofseen
science.
in Freud’s
ofExamples
commonsense.
criticisms
are given
ofThis
the
in way
paper
relation
Jung
explores
toused
the

Moscovici (1984), in his essay “The phenomenon of representations”, wrote that

if one wanted to understand a particular representation, it was necessary to start
with the representation or representations “from which it is born” (p. 13). This is
what makes the English translation of Moscovici’s great work on psychoanalysis
both untimely and timely. In an obvious sense, it is untimely because it is so long
overdue. Since 1961, when the book first appeared, various French psychology
books of far less intellectual importance have been translated into English. But,
in a deeper sense, the delay is timely. Its current re-publication encourages us now
to reflect historically on the representations from which the theory of social
representations was born.
At the minimum this means reflecting on the main thesis of Moscovici’s
Psychoanalysis—namely, the representation of psychoanalysis in popular culture. To
follow Moscovici’s own recommendation, we should seek to understand this
representation historically. This means we should reflect upon the origins of
psychoanalysis and, in particular, examine what representations of the world were
involved in these origins. As will be suggested, it is possible to find processes,
which Moscovici located in the passage of psychoanalysis into commonsense,
occurring within psychoanalysis before it entered popular culture. This has
significance for Moscovici’s thesis about the relations between science and
commonsense in contemporary society.
In addition, the re-publication is timely, because it should encourage us to reflect
on the origins of the theory of social representations itself. If social psychologists
are to be truly historical, they should do more than examine the historical
origins of the ideas that they study. They should also self-reflexively examine the
historical origins of their own ideas (Billig, 2008). When Moscovici wrote Psychoanalysis,
there was no network of researchers self-consciously promoting the study of
“social representations”. There were no “social representation” summer schools,
conferences or doctoral programmes. Today the editors of Papers on Social Repre-
sentations can address “the social representation community” (Editors, 2003,

© 2008 The Author

Journal compilation © The Executive Management Committee/Blackwell Publishing Ltd. 2008. Published by Blackwell
Publishing, 9600 Garsington Road, Oxford, OX4 2DQ UK and 350 Main Street, Malden, MA 02148, USA.
356 Michael Billig
p. 3.1). The journey of “social representation” from a nominal, denoting a supposed
entity in the social world, to “social representation” as an adjective, describing a
particular community, approach or theory, had not yet been made.
Moscovici’s arguments about the diffusion of scientific ideas point reflexively to
his own first book as a resource for examining the birth of an idea that has
become more than an idea—that has become the identity for a community of
academics. Accordingly, the tardy re-publication of Psychoanalysis, appearing as it
does in a very different intellectual climate from its first edition, offers the opportu-
nity, not just for looking at the origins of psychoanalysis and social representation
theory, but for comparing the historical transformations of both.

SCIENCE AND SOCIETY

Moscovici’s Psychoanalysis is a deeply historical work. It concerns the movement of

science into common-sense, as ideas pass from the world of scientists into the
wider society. The book is also a historical study of contemporary society. Moscovici
suggested that today commonsense is infused with concepts that originated in
scientific theories. This makes contemporary representations different from those
of traditional societies which had no formal sciences. It is not merely the content
of today’s representations that is entirely new, but so are the means of their
diffusion. Radio and magazines—and when Moscovici conducted his original
research, television was not the cultural force that it has now become—promote
entertaining controversies. They are not disseminating a unified view of the world,
in the way that propagandists attempted. Disagreement and debate sell news-
papers and attract listeners. We live in an age of atomised opinions, rather than
structured ideologies. In consequence, commonsense contains new forms of entity—
namely, social representations which differ from the collective representations that
Durkheim studied in traditional societies.
The evidence for this bold, historical thesis lay in Moscovici’s analysis of
psychoanalytic terminology. Moscovici took psychoanalysis as an example of a
science whose ideas were spreading to the population at large. He was arguing
that, in the years following the Second World War, commonsense in France had
incorporated concepts from psychoanalysis. This incorporation involved more
than a transfer of ideas from one domain to another. Psychoanalytic ideas were
transformed as they entered public debate, for the journey from science to social
representation made the abstract concrete and the strange familiar.
At the root of this thesis lay a contrast between science and common sense. In
the “Preliminary remarks”, Moscovici declared that “gradually or suddenly,
depending upon the country, regime or social class, psychoanalysis descended
from the heaven of ideas and entered into the life, thoughts, behaviour, habits and
the world of conversations of a great number of individuals” (2008, p. xxv, emphasis
in original). The metaphor echoed Marx and Engels’s claim in The German Ideology

© 2008 The Author

Journal compilation © The Executive Management Committee/Blackwell Publishing Ltd. 2008
 Social Representations and Repression 357
that the philosophy of the German idealists “descends from the heavens to earth”
(1846/1970, p. 47). Whereas Marx and Engels were arguing that any philosophy
descending from the heavens would misunderstand the nature of ordinary life,
Moscovici was suggesting that, by descending from the heavens, science enters
into everyday life—it becomes a part of what Marx and Engels referred to as
the “life-process”. This image of descent contains two elements: science as an
abstraction, emerging from the unworldly heavens, and commonsense rooted into
worldly life of people. In the “Preliminary remarks”, Moscovici comments of
social representations that “their role is to shape something that is given from the
outside” (p. xxx). In this way, Moscovici was depicting science as something that
existed separately from commonsense.
In an obvious sense, this was an exaggeration. The world of science is not
divorced from the world of common sense. As has been repeatedly shown, scientists
must use ordinary language when they deal with colleagues, run their laboratories,
explain away the theories of rivals, even when they write their technical papers
etc (e.g., Davis and Hersh, 1990; Gilbert and Mulkay, 1984; Latour and Woolgar,
1986; Mulkay, 1991; Woolgar, 1988). Vaihinger (1935) observed a number of
years ago that science needs fictional exaggerations in order to advance. Most
notable theories in psychology start their existence as exaggerations—Freudian
psychoanalysis, Skinnerian behaviourism, cognitive dissonance, Piagetian theory
etc. Later researchers can add the qualifications and limitations, rather like
accountants balancing the books. Similarly, the theory of social representations
begins with a creative exaggeration. As if to counterbalance his exaggeration,
Moscovici (1984) later commented that science draws on commonsense, making
it less common. He might equally have said that science makes commonsense less
“sensible” (in both meanings of the word). Nevertheless, the original emphasis on
the movement from science to commonsense has continued. Volklein and
Howarth (2005) observe that later social representation researchers have rarely
examined how social representations influence science, rather than vice versa.
Moscovici did not define the differences between science and commonsense.
Indeed, he has consistently resisted offering definitions for his key concepts,
suggesting that definitions typically constrain the theoretical imagination (Moscovici,
1985). This is one reason why he was unconcerned whether psychoanalysis really
was or was not a science. Nevertheless, in Psychoanalysis Moscovici made a number
of distinctions between science and commonsense. He claimed that scientific ideas
stand in need of verification, while social representations, being embodied in
social life, produce their own concrete examples (p. 112f ). He also suggested that
sciences are abstractions, although typically they do not originate in abstract
thinking (pp. 21–2). Social representations, by contrast, move in the opposite
direction by translating the abstract into the concrete (pp. 67–8). There is a further
difference of great significance. Science, according to Moscovici, is univocal: it
does not permit the co-existence of contradictions (p. 178). On the other hand,
commonsense—and especially the commonsense of contemporary societies—is

© 2008 The Author

Journal compilation © The Executive Management Committee/Blackwell Publishing Ltd. 2008
358 Michael Billig
multivocal. It is characterised by “cognitive polyphasia” (pp. 190ff; for a discussion
of this concept, see Jovchelovitch, 2006). The idea that commonsense contains
opposing, contradictory themes is not new. It can be found in ancient rhetorical
theory. Opposing topoi—such as those of justice/mercy, or courage/foolhardiness—
belong to the sensus communis, or shared sense of the community (Billig, 1996).
Everyday thinking and dialogue depend on the existence of contradictory themes.
In this matter, there is no disagreement between social representation theory and
rhetorical theory—something that Moscovici has acknowledged (Moscovici, 2000,
pp. 147–8; see also Marková, 2005). However, the general contrast between the
supposed univocality of science and the multivocality of commonsense contains a
number of difficulties.

PSYCHOANALYSIS AND MULTIVOCALITY

Psychoanalysis provides Moscovici’s object of study, not its means of analysis, as

was noted by Daniel Lagache in his preface to Psychoanalysis. For example, Moscovici
does not offer psychoanalytic explanations to account for the loss of libidinal
themes as psychoanalysis moves from theory to representation. This leads to a
curious situation. Moscovici presents psychoanalysis as “first and foremost, a science
or a theory” (p. 57, emphasis in original). He also assumes that social psychology
is a science. But if the theory of social representations and psychoanalytic theory
are both scientific theories, how can they produce very different accounts when
they study the same object?
Of course, the simple answer is that the world of science contains different
theories. Each theory may be consistent within itself and, thus, “constitute a
univocal theoretical structure” (p. 178). These various theories may contradict
each other, filling the world of science with debate and controversy. If the law of
contradiction is the dominant feature of scientific thinking, then scientists cannot
permit the equality of contradictory scientific theories. They will assume that one
theory—their theory—will eventually prove its competitors to be mistaken. This
is how scientists think in practice about their own theories and those of rivals
(Gilbert and Mulkay, 1984). Yet, Moscovici, by theoretically assuming that both
psychoanalysis and social psychology are sciences, rather than rivals in a zero-sum
game, is expressing the multivocality of non-scientific thinking. At the same time,
when explaining phenomena, he acts as if his own version of non-psychoanalytic
social psychology produces univocally superior interpretations. In this regard, his
work multivocally expresses both multi- and univocality.
Matters become more complicated with respect to Moscovici’s representation
of psychoanalytic theory as scientific. If the movement from science to
social representation is the movement from the abstract to the concrete, then
psychoanalytic theory needs to be shown as an abstract theory. Moscovici draws
parallels between psychoanalytic and Newtonian theories (pp. 57ff ). The various

© 2008 The Author

Journal compilation © The Executive Management Committee/Blackwell Publishing Ltd. 2008
 Social Representations and Repression 359
theoretical elements of Newtonian theory are held together by the unifying
principle of gravitation. Moscovici writes that “libido” is to psychoanalysis “what
universal gravitational theory is to the Newtonian system”. Then he adds the
all-important caveat “or at least Freud’s first formulations (whose presence we
note in our society) were based upon that fundamental phenomenon” (p. 57).
The caveat seems both to save and defeat the principle that Moscovici is
proposing. The obvious problem with the principle was noted by Lagache, who
observed that Moscovici was proposing the libidinal model as the prime
psychoanalytic theory (p. xx). However, according to Lagache, this distorted the
situation for there are many different psychoanalytic theories, including some that
explicitly rejected the primacy of the libido. Lagache did not elaborate the point.
A whole peloton of important psychoanalytic theorists criticised Freud for
exaggerating the importance of the libido: Adler, Jung, Horney, Lacan, Reik, Fromm
etc and etc. Instead of accepting psychoanalysis as multivocal, Moscovici appeals
to “Freud’s first formulations” to establish the primacy of libidinal theory (p. 57).
Thus, he enlists history to rescue the law of contradiction. The different versions
of psychoanalytic theory may indeed contradict each other, but the psychoanalytic
theory—the one that has been diffused into wider society—is to be found in
Freud’s early work.
Quite apart from other problems to be discussed later, there is one obvious
difficulty. Moscovici assumes that the Freudian theory—as opposed to other
versions—has found its way into commonsense. Yet, the very distinguishing
feature of that theory—its libidinal principle—is exactly what has failed to pass
into common sense. This means that the social representation of psychoanalysis is
closer to the theories of Adler, Fromm and Horney than it is to those of Freud. One
might ask whether the loss of the libidinal theory in the public representation of
psychoanalysis is actually the result of theory being translated to the world of
concrete life—or whether loss has occurred already within the psychoanalytic world.
There is evidence that early in the development of psychoanalysis the iconic
symbol of “Freud” had become detached from Freud’s original formulations and
attached to scientific offspring. The peloton, cited above, all cycle in Freudian
team-colours. Lacan, despite abandoning some of Freud’s core assumptions and
his style of reasoning, claimed to be returning to Freud. Horney in New Ways in
Psychoanalysis rejected Freud’s emphasis on the libido, but nevertheless located her
own ideas as Freudian, claiming to follow “the imperishable values Freud has
given to psychology and psychiatry” (1939, p. 18). Fromm’s position was similar.
If, according to Moscovici, the passage from scientific theory to social
representation lost the notion of “libido”, then it saw another concept gaining in
prominence—the concept of “a complex”. Its meaning also changed. “Complex”
became drained of its specific theoretical meaning to become more concrete, as
ordinary people talked generally about “inferiority complexes” or “timidity com-
plexes”. In this way, the term “complex” was “materialized” (p. 68). Its importance
expanded to become emblematic of the whole of psychoanalysis. Thus, Moscovici

© 2008 The Author

Journal compilation © The Executive Management Committee/Blackwell Publishing Ltd. 2008
360 Michael Billig
writes that “the entire social representation of psychoanalysis is, as it were,
concentrated in this notion and assimilated to it” (p. 158).
To attribute this change in the meaning of “complex” just to the journey from
science to social representation may be historically too simple. Freud, in his
“History of the Psychoanalytic Movement”, complained about the way that Jung
had popularised the word “complex”. It is worth spelling out Freud’s complaint.
According to Freud, the word “complex” had not produced “a psychological
theory”, nor was it easily assimilated into “psychoanalytic theory”; yet “the word
“complex” . . . has become naturalized, so to speak, in psychoanalytic language”.
It has become, he wrote, a convenient short-hand: “none of the other terms coined
by psychoanalysis for its own needs has achieved such widespread popularity or
been so misapplied to the detriment of the construction of clearer concepts”.
Thus, analysts speak of a “return of a complex” when they should refer to the
“return of the repressed”; or they say “I have a complex against him”, when,
according to Freud, “the only correct expression would have been ‘a resistance
against him’ ” (Freud, 1914/1993, p. 87).
In the context of Moscovici’s Psychoanalysis, the passage is extraordinary. Freud
was writing just before the First World War, almost ten years before he began
using the theoretical terms “id” and “superego”. It was a time when the
psychoanalytic movement was still comparatively small. Yet, Freud, in criticising
how psychoanalysts were using the word “complex”, was using virtually identical
terms to the way that Moscovici would depict the way that the word was being
used in French popular culture following the Second World War. Freud was noting
the word’s popularity (at least among psychoanalysts), its lack of theoretical
specificity and the way that it had become “naturalized” (or, to use, Moscovici’s
preferred term, “objectified”). Freud, unlike Moscovici, was deploring these
linguistic developments, holding Jung and his followers to blame. It is significant that
Freud could complain that this was happening within the world of psychoanalysis
before “psychoanalysis” had become an important element of popular culture.
This suggests that science and popular culture may not be entirely distinct, but
both may be affected by similar linguistic and social processes.

FIRST FORMULATIONS AND MULTIVOCALITY

Moscovici recommended that we should turn to Freud’s first formulations to see

the science of psychoanalysis. Where should we find Freud’s first formulations?
The obvious place is Freud’s first book, Studies on Hysteria, (published in 1895). It
was here that Freud introduced the concept of repression, which he would
describe as “the corner-stone on which the whole structure of psychoanalysis
rests” (1914/1993, p. 73). The notion of the unconscious, which also first
appeared in Studies, depends upon repression, for the unconscious comprises that
which is repressed. This is why some scholars have seen Studies as containing the

© 2008 The Author

Journal compilation © The Executive Management Committee/Blackwell Publishing Ltd. 2008
 Social Representations and Repression 361
essential ideas of psychoanalysis (e.g., Billig, 1999; Grubrich-Simitis, 1997;
Grünbaum, 1998). Freud, in his preface to the second edition, was to be more
ambivalent, notably because, when he wrote Studies, he had not yet developed his
theory of infant sexuality.
Studies is not a univocal work, quite apart from the obvious fact of having two
authors, Freud and Joseph Breuer. Only the first theoretical section was jointly
written by the two. The other sections can be attributed to one or the other. In
another respect—and more importantly so for the present argument—Studies is
multivocal. It contains theoretical sections, describing the mental operations
underlying of hysteria, but the core of the book comprises five case histories—one
written by Breuer and four by Freud.
In many respects, the style, structure and language of the case histories differ
from the theoretical sections. Freud, after outlining the story of the final case-
history, commented on his style of writing. He mentions that he had not always
been a psychotherapist but had trained as a neuropathologist: “It still strikes me
as strange that the case histories I write should read like short stories and that, as
one might say, they lack the serious stamp of science” (1895/1991, p. 231). He
added that “the nature of the subject is evidently responsible for this, rather than
any preference of my own” (p. 231).
Certainly, Freud’s case histories have a wonderful, literary quality. Because of
the richness of their detail and their narrative power, later analysts have returned
to them again and again, claiming to discover new secrets about the patients, about
Freud and about the nature of psychoanalysis. None of this would have been
possible had the case histories resembled the sort of short diagnoses and descrip-
tions of treatment that came to characterise many psychoanalytic case histories
after psychoanalysis became well established (Meehl, 1990). It would not have
been possible for Freud to have summarised the condition of those early patients
in conventional psychoanalytic terminology, for he had not yet invented that
terminology. The accounts had to be concrete and extended, filled with detail.
For this, the style needed to follow literary fiction rather than medical report.
Freud’s early case histories should not be treated as if they belonged to a
pre-scientific stage before psychoanalysis emerged as a proper, univocal, abstract
science. Case histories are integral to psychoanalysis. As Moscovici noted, science
stands in need of verification. Case histories provide the verification for
psychoanalytic theory. This means that psychoanalysis never was separate from
the world of commonsense and social representations. Its object of study was the
world of ordinary people; and its means of study was conversation. Ordinary
people needed to bring their representations into the laboratory, and the analyst
needed to re-present those representations back to them. The consulting-room
was psychoanalysis’s laboratory. There could be no control group, from which
social representations were excluded. Such a control group would have been a
silent consulting room—and that would have constituted a poor scientific test of
the talking cure.

© 2008 The Author

Journal compilation © The Executive Management Committee/Blackwell Publishing Ltd. 2008
362 Michael Billig
The cases histories in Studies took place before Freud’s reputation was made
even in the small Viennese circles from which his early patients came. When those
first patients came to Freud, they had little notion what to expect; nor, in truth,
did their doctor. Word would soon spread about the spectacular successes of the
young nerve doctor. By October 1907, when Ernst Lanzer, the “Rat Man”,
started treatment, things were different. He came to Freud expecting to talk about
his sexual experiences and began doing so in the first session (1909/1991, p. 39).
Freud, in the published case history, records that he had asked the young man
why he thought he should be talking about sex. Ernst answered that although he
had not read any of Freud’s books, a little while earlier “he had been turning over
the pages of one of my books and he had come across the explanation of some
curious verbal associations” (pp. 39–40). These so reminded him of his own
thoughts that he had decided to consult Freud. Fortuitously the notes that Freud
wrote after each session—his “process notes”—have survived for this case. These
reveal that Freud’s published account was downplaying Lanzer’s knowledge of
psychoanalysis. The process notes show Lanzer claiming to have read parts of
Freud’s Psychopathology of Everyday Life (Hawelka, 1974, p. 63). In the second
session, the patient praised Freud for his ideas; Freud noted that “he had read an
extract from my theory of dreams” (Hawelka, 1974, p. 53).
The patients, of whom Freud wrote in Studies, did not spontaneously talk about
their sexual feelings. Freud had to approach such personal matters delicately.
Even then, some were reluctant to talk openly. One of the “patients” was not even
a patient. “Katherina” was the teenage daughter of an inn-keeper whose
establishment Freud visited when climbing mountains during his summer break.
The young girl had served him lunch and then asked him whether he was a
doctor, having, she said, seen his signature in the visitors’ book. She has having
problems with her nerves and her local doctor had not really helped.
Clearly, the young country girl knew nothing of Freud’s work. But the Viennese
patients of Studies could not have glanced at, let alone read, any Freudian books,
as none existed. The theories could not pre-date the cases, for the patients and
their talk were helping their doctor to create the theory. This is why those early
cases are so significant. By the time Lanzer began his treatment, Freud’s ideas were
becoming known in psychoanalytic circles. Freud and Jung had met earlier in the
year that Lanzer began his treatment. Freud was greatly impressed by the young
Swiss doctor (see, for instance, Gay, 1995, pp. 197ff ). Jung had not behaved
entirely differently from Lanzer. Jung had read some of Freud’s works and written
to him, the difference being that Jung contacted Freud as a fellow psychiatrist
rather than as a potential patient, although many followers were to contact Freud
hoping that the great man would analyse them personally. It would only take a few
years before Freud would be accusing Jung of naturalizing the concept of “complex”.
By 1907 some patients like Lanzer may have read some of Freud’s works;
others might have known someone who knew someone who had read a book by
Freud; or they knew someone who had been treated by Freud; and some belonged

© 2008 The Author

Journal compilation © The Executive Management Committee/Blackwell Publishing Ltd. 2008
 Social Representations and Repression 363
to the growing circle of Freud’s followers. All these patients would have some idea
what to expect when they came to the consulting-room and lay down on the couch.
None could be as innocent as Katherina, serving Freud his dinner in the remote
mountain inn. Psychoanalytic ideas were spreading, albeit slower than Freud
would have wished and less extensively than they would do in the years to come.
The diffusion of psychoanalytic ideas did not just run from theory to public
representation, from abstraction to concreteness, or from Freud to the outside
world. Lanzer’s understandings and misunderstandings would feed back into
psychoanalytic theory. When Freud published the case history, he claimed that it
permitted him to develop theoretically the “first observations” on the subject of
obsessional neurosis that he had made (1909/1991, p. 36). He hoped that the
publication would be a “starting-point for the work of other investigators” (p. 38).
As it was, the case-history became a classic in the history of psychoanalysis.
Accordingly, Lanzer’s ways of representing the world were to become a
powerful source of verification for psychoanalysts. His was the classic case
showing that repressed desires lay at the root of “obsessional neurosis”. Lanzer’s
contribution was more than that of an object to be studied, as if he were a
chemical solution in a test-tube or specimen under a microscope. He was an active
participant, not just in providing the details of life, but also in the ways of under-
standing those details. In this respect, he was a contributor to the development of
theory. He supplied the phrase “the omnipotence of thought” to describe his
obsession that his private thoughts could affect the events of the world. Later
Freud would use this phrase in Totem and Taboo to explain the obsessive nature of
religious thinking. Religious believers, like obsessive neurotics, are convinced that
their thoughts can determine whether good or evil events occur. Freud described
this as “omnipotence of thought”, writing that it was not he but his “highly
intelligent” patient, who had “coined the phrase as an explanation of all the
strange and uncanny events by which he, like others afflicted with the same illness,
seemed to be pursued” (Freud, 1913/1990, p. 143).
If psychoanalysis was to become an abstract science, as Moscovici argued, it
needed its verifying data. The patients—their voices, their ways of talking, their
common and not so common ways of understanding—provided the data. But
they did more than that. They were active participants in the conversations that
formed the basis of psychoanalysis. Their words were central to the creation of
psychoanalysis as a science. In this respect, psychoanalysis could never be univocal—it
had to be multi-voiced or multivocal.

ACTION AND SCIENTIFIC LANGUAGE

There is a deeper, rhetorical level at which psychoanalytic theory has tended to

be multivocal. Not only can the voices of scientist (or doctor) and patient be
heard, but the psychoanalytic scientist does not speak (or write) in a single voice.

© 2008 The Author

Journal compilation © The Executive Management Committee/Blackwell Publishing Ltd. 2008
364 Michael Billig
Two different psychoanalytic voices can be broadly distinguished. As Freud noted,
his case histories read like short stories. These stories demand to be written in the
sort of language that Roy Schafer (1976), in an important revision of psychoanalytic
theory, called “action language”. This is principally the language of ordinary life:
it is the language which tells us what people do. By contrast, there is also a
technical language which outwardly parades the stamp of serious science. When
used to describe people’s actions, Schafer calls this a “reified” language. Instead
of attributing acts to people, this language ascribes them to forces, energies and
other biological or psychic entities. Schafer argued that this language has been
disastrous for the development of psychoanalytic theory. It draws attention away
from what people actually do and it assumes that their ways of thinking are
governed by unseen mechanical entities.
Linguists have identified two important features of scientific language. Scientific
writers tend to favour “nominalizations”, or a preference for using nouns, rather
than verbs, to denote processes; and they tend to use passive, rather than active,
sentences (e.g. Goatly, 2007; Halliday and Martin, 1993). As critical discourse
analysts have pointed out, there are ideological dangers in using nominalization
and passivization, for, by using these forms, writers/speakers can avoid describing
how people perform actions (Fowler et al., 1979; Fowler, 1991; Lemke, 1995).
When human or social scientists use this sort of language, their analyses can often
be ambiguous (Billig, in press, a; in press, b; but see Fairclough, in press; van Dijk,
in press). Writers use technical nominals to denote abstract entities, but it can be
unclear whether such entities are believed to exist in a realistic or in a metaphorical
sense. What exactly is a “cognitive representation”? Is it, to use the words of
Vaihinger, a theoretical fiction, which psychologists have invented as a metaphorical
“as-if ” to stimulate insight? Or do human scientists believe such “representations”
actually exist? If so, where and how, do they exist? Researchers, following
conventional experimental paradigms, often forget to detail exactly what it is
about people’s actions—their ways of representing the world—that led observers
to assume the existence of “cognitive representations” in the first place. Instead a
research tradition develops, and a community of researchers takes for granted the
existence of non-observable, fictional entities.
The Studies contains both “action” and “reified” language, as Freud talks about
people and about their presumed unconscious, biological mechanisms of mind.
The former language is not confined to the case histories and the latter to the
theoretical sections. Instead, the two voices are interconnected. As Moscovici
suggested, the first formulations are particularly revealing. So it is interesting to
examine Freud’s first use of “repression” (Verdrängung).
Significantly the first use comes, not as a nominal (repression) but as an active
verb (repressing). The opening section, written jointly with Breuer, discusses the
traumas underlying neuroses: “It was a question of things which the person
wished to forget, and therefore intentionally repressed from his conscious
thought” (1895/1990, p. 61; 1895/1952, p. 89). Here, the authors do not use the

© 2008 The Author

Journal compilation © The Executive Management Committee/Blackwell Publishing Ltd. 2008
 Social Representations and Repression 365
noun “repression” (Verdrängung). The verb indicates an action that a person performs:
the person can push aside thoughts—verdrängte—and does so intentionally—
absichtlich. In the case histories, we can see Elisabeth von R struggling to dispel
from her own mind the shameful thought that she desired her sister’s husband.
Her symptoms are a way of distracting herself and others. Repressing, here, is not
merely a human action; it is an all too human failing.
“Repression”, as the noun derived from the verb verdrängen (to push aside),
makes its appearance later in the case history of Lucy R., which was written
wholly by Freud:

“(A)n idea must be intentionally repressed from consciousness and excluded from associative modifi-
cation. In my view, the intentional repression is also the basis for the conversion, whether total or
partial, of the sum of excitation. The sum of excitation, being cut off from psychical association,
finds its way all the more easily along the wrong path to a somatic innervation” (1895/1990,
pp. 181–2, emphasis in English translation but not in original German text: 1895/1952, p. 174).

In this passage, Freud starts by using “repress” as a verb, again with the adverb
“intentionally” (absichtlich), but this time in the passive tense. Instead of writing
that a person intentionally represses an idea, he writes of an idea being intentionally
repressed (verdrängt), thereby omitting the human agent, who might be doing the
repressing. Then Freud moves from passive verb to nominalization. “Repression”
( Verdrängung) is the subject of the following sentence, which includes two
other abstract nominals, both denoting processes—“conversion” (Konversion) and
“excitation” (Erregungssumme: or excitation-sum). In the next sentence Freud
describes how “the sum of excitation” does something—namely, “finding its way”
along a path to a “somatic innervation” (körperlichen Innervation).
Freud has made a move from action language to reified, nominalized language.
The actor now is not a person but a process—a supposed, but unspecified, process
of excitation. This is certainly not the language of short stories. Despite its
technical quality, this language is actually quite vague—what and how things are
being converted, excited, innervated is unclear. What exactly is this “sum of
excitation”? It is described as being cut off from “psychical association” (psychische
Assoziation), as if it could be associated with psychical, non-material entities, but it
is also described as finding its way to something material or bodily (körperlichen). Is
it a neurological phenomenon? How is it to be identified? How does it go about
“finding its way”? Freud does not say.
The paradox is that ordinary language can be quite specific when describing
human action whereas technical, scientific sounding language is frequently
imprecise, especially when used to “explain” human actions. There is a cost in
moving towards the reified language. “Repression” is posited as a thing that does
hidden bodily tasks. This way of representing psychoanalytic processes draws
attention away from what the person actually has to do in order to accomplish
the task of repressing. In consequence, there is a large gap in Freud’s theorising
(Billig, 1999). He does not specify the skills that the person needs to acquire to be

© 2008 The Author

Journal compilation © The Executive Management Committee/Blackwell Publishing Ltd. 2008
366 Michael Billig
able to repress. In Freud’s developmental theory, it is as if the biological engine of
repression automatically starts up when the child is at the Oedipal stage. Even in
his case history of Little Hans, Freud does not observe how the parents are
teaching the child that ideas should be pushed from his mind, and demonstrating
by their own actions how this might be done (Billig, 1999, chapter five). Much of
this repressing is not consciously intentional, as Freud and Breuer might
have implied in their first formulation, but it is subtly habitual; in this way it is
unselfconsciously learned, practised and transmitted.

REPRESSING AND REPRESENTING

Are there parallels between the linguistic history of the concepts of “repression”
and “social representation”? This is a big question that requires detailed study in
its own right. But a few very brief suggestions are possible. Moscovici’s first
formulations of “social representation” do not begin with an active verb (“social
representing”) and then move to the technical nominal. The first sentence of the
first chapter of Psychoanalysis is a bold statement of existence: “Social representa-
tions are almost tangible entities” (p. 1). The existence of social representations is
not a hypothesis to be tested. As Vaihinger (1935) noted, scientists do not propose
their fictional entities as hypotheses. They knowingly propose them as if they exist,
using metaphor as integral to creativity. Thus, Moscovici claims that social repre-
sentations are almost tangible. He knows that they are not actually tangible, but
he is treating them as if they are actually existing entities. The second sentence of
this first chapter makes social representations the grammatical subject of actions:
“They circulate ceaselessly in our day-to-day world, intersect and crystallize
through a word, a gesture, an encounter” (p. 1).
The metaphor of social representations doing things continues throughout
Psychoanalysis.1 There is a parallel with Freud and “repression”. Both emphasise the
entity, rather than the activity. Certainly at times in Psychoanalysis, Moscovici mentions
the activity of “representing” but overwhelmingly he concentrates on the entities—
the “social representations”. This has continued in later research. It is rare for
a member of the “social representation community” to focus on the activity
of “social representing” rather than on the entity (see, for example, Valsiner, 2003).
As the term “social representation” circulates, its non-metaphorical reality
becomes firmly established, just as “repression” did with psychoanalysts. In this,
there is a historical movement from scientific metaphor to realism. As Moscovici
notes in a phrase that echoes Vaihinger, at least in spirit, “a metaphor is a young
analogy” and when it is mature “it becomes a hypothesis” (2008, p. 356). However,
the hypothesis is rarely tested as such, for the existence of the entities, whose
reality might be hypothesised, is taken for granted. In this regard, objectification
can be observed within the world of science, which is always more than just a
scientific world. The “social representation community” cannot treat the existence

© 2008 The Author

Journal compilation © The Executive Management Committee/Blackwell Publishing Ltd. 2008
 Social Representations and Repression 367
of social representations as a mere hypothesis that has to be tested, any more
than practising psychoanalysts could treat the existence of “repression” or “the
unconscious” as just hypotheses. These particular communities require their
members to hold certain existential beliefs; otherwise the communities would not
exist as communities. Members are expected to employ regularly their special
words of identity in their official discourse. Members of the social representation
community must repeat the phrase “social representation”, as Freudians must
repeat “repression”, “id” and “Oedipal stage”. If such words have become almost
magical, their magic lies in the way that they are held to represent literal truths.
Without the ritual repetition members risk being ostracised by their community.
But that is another story.

Michael Billig
Department of Social Sciences
Loughborough University
Leicestershire LE11 3TU
m.g.billig@lboro.ac.uk

NOTE

1
This is also true of the second part of Psychoanalysis, in which Moscovici examines
systems of communication, rather than social representations per se. Again nouns
predominate over verbs. He discusses “diffusion” rather than people diffusing, “pro-
pagation” rather than propagating etc. When Moscovici claims that “diffusion”, “propagation”
and “propaganda” have “goals” (p. 282), these processes appear as actors (rather than
those who diffuse, propagate or propagandise). The same points, which Billig (in press a;
in press b) addresses to the way that linguists use terms such as “nominalization” and
“passivization”, apply equally when Moscovici writes that “tautologization organizes and
crystallizes the representation” (p. 331). A presumed linguistic process, rather than specific
speakers/writers, seems to be performing the action of organizing. When analysts use this
sort of phrasing, they assume the existence of the entities to which they refer—in this case
“tautologization” and “representation”. Here, we are no nearer to examining closely and
specifying exactly what a speaker/writer has to do in order to be said to be “socially
representing”.

REFERENCES

Billig, M. 1996. Arguing and Thinking. Cambridge: Cambridge University Press.

Billig, M. 1999. Freudian Repression. Cambridge: Cambridge University Press.
Billig, M. 2008. The Hidden Roots of Critical Psychology. London: Sage.
Billig, M. in press, a. The Language of Critical Discourse Analysis: the Case of Nomi-
nalization. Discourse & Society.
Billig, M. in press, b. Nominalizing and De-Nominalizing: A Reply. Discourse & Society.
Davis, P.J. and Hersh, R. 1990. Descartes’s Dream. Brighton: Harvester Press.
Editors. 2003. Editorial May 2003. Papers on Social Representations, 12, 3.1–3.2.

© 2008 The Author

Journal compilation © The Executive Management Committee/Blackwell Publishing Ltd. 2008
368 Michael Billig
Fairclough, N. in press. The Language of Critical Discourse Analysis: Reply to Michael
Billig. Discourse & Society.
Fowler, R. 1991. Language in the News. London: Routledge.
Fowler, R., Hodge, B., Kress, G. and Trew, T. 1979. Language and Social Control. London:
Routledge.
Freud, S. 1909/1991. Notes upon a Case of Obsessional Neurosis. In S. Freud, Case
Histories, volume II. Harmondsworth: Penguin, pp. 39 –128.
Freud, S. 1913/1990. Totem and Taboo. In S. Freud, The Origins of Religion. Harmonds-
worth: Penguin, pp. 49–224.
Freud, S. 1914/1993. On the History of the Psychoanalytic Movement. In S. Freud,
Historical and Expository Works on Psychoanalysis. Harmondsworth: Penguin, pp. 63–128.
Freud, S. and Breuer, J. 1895/1952. Stüdien über Hysterie. In S. Freud, Gesammelte
Werke, Volume 1. Frankfurt am Main: S. Fischer.
Freud, S. and Breuer, J. 1895/1991. Studies on Hysteria. Harmondsworth: Penguin.
Gay, P. 1995. Freud: A Life for Our Time. London: Papermac.
Gilbert, N. and Mulkay, M. 1984. Opening Pandora’s Box: A Sociological Analysis of Scientists’
Discourse . Cambridge: Cambridge University Press.
Goatly, A. 2007. Washing the Brain. Amsterdam/Philadelphia: John Benjamins.
Grubrich-Simitis, I. 1997. Early Freud and Late Freud. London: Routledge.
Grünbaum, A. 1998. A Century of Psychoanalysis: Critical Retrospect and Prospect. In
M.S. Roth (Ed.), Freud: Conflict and Culture. New York: Alfred Knopf.
Halliday, M.A.K. and Martin, J.R. 1993. Writing Science. London: Falmer Press.
Hawelka, E.R. 1974. L’homme Aux Rats: Journal D’une Analyse. Paris: Presses Universitaires
de France.
Horney, K. 1939. New Ways in Psychoanalysis. New York: W.W. Norton.
Jovchelovitch, S. 2006. Knowledge in Context. London: Routledge.
Lagache, D. 2008. Foreword. In S. Moscovici, Psychoanalysis: Its Image and its Public. Cam-
bridge: Polity, pp. xviii–xxi.
Latour, B. and Woolgar, S. 1986. Laboratory Life. Princeton: Princeton University Press.
Lemke, J.L. 1995. Textual Politics. London: Taylor and Francis.
Marková, I. 2005. Dialogicality and Social Representations. Cambridge: Cambridge University
Press.
Marx, K. and Engels, F. 1846/1970. The German Ideology. London: Lawrence and Wishart.
Meehl, P.E. 1990. Selected Philosophical and Methodological Papers. Minneapolis: University of
Minnesota Press.
Moscovici, S. 1984. The Phenomenon of Social Representations. In R. Farr and
S. Moscovici (Eds.), Social Representations. Cambridge: Cambridge University Press, pp. 3–68.
Moscovici, S. 1985. Comment on Potter and Litton. British Journal of Social Psychology, 24,
91–92.
Moscovici, S. 2000 Social Representations: Studies in Social Psychology. Cambridge: Polity.
Moscovici, S. 2008. Psychoanalysis: Its Image and Its Public. Cambridge: Polity.
Mulkay, M. 1991. Sociology of Science. Milton Keynes: Open University Press.
Schafer, R. 1976. A New Language for Psychoanalysis. New Haven: Yale University Press.
Vaihinger, H. 1935. The Philosophy of “As If ”. London: Routledge and Kegan Paul.
Valsiner, J. 2003. Beyond Social Representations: A Theory of Enablement. Papers on
Social Representations, 12, 7.1–7.16.
Van Dijk, T.A. in press. CDA and nominalization: Problem or Pseudo-Problem? Discourse
& Society.
Volklein, C. and Howarth, C. 2005. A Review of Controversies about Social Repre-
sentations Theory: A British Debate. Culture and Psychology, 11 (4), 431– 454.
Woolgar, S. 1988. Science: The Very Idea. London: Tavistock.

© 2008 The Author

Journal compilation © The Executive Management Committee/Blackwell Publishing Ltd. 2008