Science in Context 7, 1 (1994), pp.

83-101

JOSE BRUNNER

"Every Path Will End in Darkness"

or: Why Psychoanalysis Needs
Metapsychology

The Argument

This article focuses on the dialectic of metapsychology and hermeneutics in

psychoanalysis. By combining the causal language of the former with the inten-
tional terminology of the latter, Freud's discourse continuously transgresses the
narrowly conceived boundaries of scientific disciplines and places its stakes in
both the humanities and the natural sciences. The argument is made that attempts
to reduce psychoanalytic theory to either causal explanation or interpretation of
meaning turn it into a closed thought system and rob it of its vitality. It is argued,
moreover, that although Freud understood himself to be a scientist, by eschewing
the dichotomous reductionism characteristic of both his orthodox followers and
critics who tried and still try to turn psychoanalysis into either a natural-science-
like discipline or a hermeneutics, Freud demonstrated that his self-understanding
was far more sophisticated than admitted by either of these two groups. This
argument is supported by a detailed discussion of Freud's epistemological pre-
mises, his conception of science and reality, and especially, the place he allocated
to metapsychology in his interdisciplinary science. It is claimed that metapsy-
chology served Freud as a double-edged sword, both enabling creative and
metaphorical thought about the mind's hidden reality and revealing the necessary
incompleteness of hermeneutics. The article concludes with the claim that psy-
choanalysis needs metapsychology in order to pursue this dual task.

Freud persistently refused to observe phenomena from only one point of view, to
ask only one type of question. As is well known he always regarded individual
action and social processes as "overdetermined." For him they were inevitably the
effect of many causes, carried a variety of often contradictory meanings, expressed
several and sometimes incompatible intentions, and fulfilled multiple functions.
To cope with the complexity of human conduct and its determinants, Freud
developed a pluralistic and interdisciplinary science of the mind, which he orga-
nized around four interrelated but nevertheless distinct perspectives. Freud always
discerned meaning in mental phenomena, taking them to be expressions similar to
those of a language. He revealed hidden dynamic forces — drives and energies —
84 JOSE BRUNNER

which were supposed to be akin to blind, impersonal physical entities and to

occupy and propel ideas. He delineated the contours of psychic structures in and
across which the flow of psychic energy was said to take place. Lastly, his inquiry
led him to look for origins in early childhood — that is, for the history of
observable phenomena as well for the developmental sequence of the forces and
structures he could not approach directly.
However, the four types of inquiry in psychoanalysis form two basic languages.
On the one hand there are references to intention, meaning, and history. These
belong to what is commonly known as interpretive or hermeneutic discourse. On
the other hand there are metapsychological hypotheses on causal connections,
forces, and boundaries of psychic structures. These are characteristic of the
mechanistic discourse that we know from the natural sciences. And though these
latter hypotheses may not refer to observable entities, by their very existence they
also point to the shortcomings of a "pure," clinical hermeneutics and urge psy-
choanalysis to move beyond interpretive procedures to a causal level of
explanation.
I wish to show in this paper that this uncomfortable duality of interpretation
and explanation in psychoanalytic discourse generates a creative tension. Thus my
claim is similar to that of Morris Eagle in his critical evaluation of recent develop-
ments in psychoanalysis:
It is this search for a deeper level of explanation rather than the specific
content that I take to be the significance of Freud's metapsychology. To limit
psychoanalysis to the so-called clinical theory and reject any form of deeper
theoretical account is to declare this central challenge irrelevant and . . . to
accept an inadequate and truncated form of explanation as well as to isolate
psychoanalysis from an enriching body of facts and perspectives. It is one
thing to insist on the legitimacy, even centrality, of explanation by way of
motives and aims, including unconscious motives and aims. It is another
thing to declare such accounts as final. (Eagle 1984, 149)
In this sense, then, I argue that metapsychology is part of what Samuel Weber has
described as Freud's "willingness to negotiate with the unknown and to acknowl-
edge uncertainty not merely as an impediment or defect, but as an integral part of
thinking and writing" (Weber 1987, 73). Thus, I disagree with the subtitle of
Patricia Kitcher's (1992) book on psychoanalysis, which describes Freud's program
as "a complete interdisciplinary science of the mind." For though she is right in
claiming that Freud's science was designed to be interdisciplinary, he was aware
that its fate was to remain incomplete.
In the wake of this argument I criticize various reductionist narratives that
either transform Freud's work into a set of empiricist, operationalist claims or
attempt to rephrase psychoanalytic discourse exclusively in intentionalist catego-
ries — that is, in terms of conscious and unconscious meanings, motives, and
purposes. The effect of such reductionist readings is to remove the tension between
 Why Psychoanalysis Needs Metapsychology 85

interpretation and explanation from Freudian discourse and thus to turn psy-
choanalysis into a self-contained, closed thought system. But as I shall show,
Freud was well aware that such one-dimensional constructions of psychoanalysis
rob it of its vitality. In his view, only religion and philosophy feigned completeness
and certitude; he took it to be an essential characteristic of science to allow of
tensions and contradictions and to admit to its own imperfections. Thus in this
article I develop a narrative of my own, one opposed to that constructed by most
influential figures in psychoanalysis, philosophy, and history of science, which
presents them as engaged in repeated attempts to eliminate the subversive and
creative role that metapsychology plays in Freud's work.

Let us start by joining Freud on a walk. In August 1899, while working on his
magnum opus, The Interpretation of Dreams, Freud had written to Wilhelm
Fliess that it was "planned on the model of an imaginary walk." "At the beginning
the dark forest of authors (who do not see the trees), hopelessly lost on wrong
tracks. Then a concealed pass through which I lead the reader — my specimen
dream with its peculiarities, details, indiscretions, bad jokes — and then suddenly
the high ground and the question: which way do you wish to go now?" (Freud
1985, 365). In the opening sentences of the book's third chapter Freud uses the
same vocabulary, but this time he likens the process of dream interpretation itself
to such an imaginary walk:

When, after passing through a narrow defile, we suddenly emerge up on a

piece of high ground, where the path divides and the finest prospects open up
on every side, we may pause for a moment and consider in which direction
we shall first turn our steps. Such is the case with us, now that we have
surmounted the first interpretation of a dream. We find ourselves in the full
daylight of a sudden discovery. (Freud [1900], SE 4:122)

This passage is directly preceded by a declaration at the conclusion of the book's

second chapter in which Freud states: "When the work of interpretation has been
completed [yollendei], we perceive that a dream is the fulfillment of a wish [Idsst
sichder Traumalseine Wunscherfiillung erkenneri]" (ibid., 4:121). In other words,
at this stage Freud posits that dream interpretation can reach an end or a comple-
tion, and — in the original German — he claims that this endpoint is marked by the
dream's readiness to give in to the interpreter's efforts and reveal itself as wish
fulfillment.
At the beginning of the seventh chapter, some four hundred pages later, Freud
formulates the basic principles underlying the production of dreams and of mental
functioning in general. There he again takes up the metaphor of the journey into
86 JOSEBRUNNER

the unconscious. But he describes as illusionary the sense of completion obtained

by the sudden illumination of interpretation:
It is only after we have disposed of everything that has to do with the work of
interpretation that we can begin to realize the incompleteness of our psy-
chology of dreams.... For it must be clearly understood that the easy and
agreeable portion of our journey lies behind us. Hitherto, unless I am greatly
mistaken, all the paths along which we have travelled have led us towards the
light — towards elucidation [Aufklarung] and fuller understanding [zum
vollen Verstandnis]. But as soon as we endeavor to penetrate more fully into
the mental process involved in dreaming, every path will end in darkness.
(Ibid., 5:511)
This, then, is the scientific journey on which Freud takes us: We leave the dark
wood through a narrow, concealed pass, ascend to the open high ground — only to
progress into darkness. At the end of the book we discover that The Interpretation
of Dreams has been radically incomplete all along. After several hundred pages of
detailed discussion of hermeneutic techniques, symbolism, associations, and cen-
sorship, we find that psychoanalysis cannot be reduced to an interpretive practice.
Rather than writing a conclusion in the seventh and last chapter, Freud displaces
hermeneutics — that is, the interpretation of meaning and its disguises — in order
to insert causal hypotheses. By pointing to a new realm of causal speculation that
lies beyond hermeneutics, he radically undermines hopes for hermeneutic
completeness.
In the seventh chapter Freud's book is no longer concerned with the hidden
meaning that the dreamer expresses in pictures and symbols. Instead Freud tries to
establish the origins of the dream's signs — that is, he seeks to explain what causes
them. For this purpose he refers to a different category of entities: to nonexperien-
tial mechanisms and systems, currents of psychic energy and mental agencies. As
his comment indicates, he is aware that the referents of these formulations lie in the
dark and that hypotheses concerning their nature and functioning must remain
speculative.

II

In Freud's parlance the term "metapsychology" denotes the ensemble of such

speculative causal postulates. Metapsychological formulations serve to depict
processes in an intrinsically inaccessible realm of the mind, which belongs neither
to the domain of subjective experience nor to such physical entities as the brain or
the nervous system. Thus metapsychological descriptions refer to forces, energies,
and structures in a nonexperiential mental world, of which one can gain knowledge
only indirectly by inference from clinical phenomena. Hence Freud's text presents
metapsychology as exterior to the clinical practice of interpretation, which is
 Why Psychoanalysis Needs Metapsychology 87

supposed to form the basis of all psychoanalytic endeavors. As Freud states in his
Introductory Lectures, when one makes the transition from manifest phenomena
such as symptoms and slips to metapsychological depictions of psychic processes
and structures, "the phenomena that are perceived must yield to trends that are
only hypothetical" (Freud [ 1916-17], SE 15:67). In his metapsychological conjec-
tures Freud attempts to represent the elusive world of the mind from three
perspectives. As he explains: "When we have succeeded in describing a psychical
process in its dynamic, topographical and economic aspects, we should speak of it
as a metapsychological presentation [Darstellung]" (Freud [1915], SE 14:181). In
his explanation of the dynamic aspect of metapsychology, Freud emphasizes that
he seeks "not merely to describe and clarify phenomena, but to understand them as
signs of an interplay of forces in the mind, as a manifestation of purposeful
intentions working concurrently or in mutual opposition" (Freud [1916-17], SE
15:67). To this conflictual view of the mind, the economic point of view adds
estimates of the magnitude or amount of psychic energy involved in the interplay
(Freud [1915], SE 14:181). Finally, from a topographical angle, Freud postulates
psychic systems or agencies — such as the unconscious, preconscious, or con-
sciousness — in which, or between which, mental processes are supposed to take
place (ibid., 173, 175).
Thus although Freud's metapsychological formulations remain on purely psy-
chological ground and are not supposed to point to material substrates, they
portray the mind as a spatial entity. However, Freud repeatedly cautions his
readers not to take these quasi-physical descriptions literally. In fact, already in the
early Studies on Hysteria Freud's co-author, Josef Breuer, warns of the danger of
mistaking metaphors of the mind for descriptions of the psyche's reality:
It is only too easy to fall into a habit of thought which assumes that every
substantive has a substance behind it — which gradually comes to regard
"consciousness" as standing for the actual thing; and when we have become
accustomed to make use metaphorically of spatial relations, as in the term
"sub-consciousness", we find, as time goes on that we have actually formed
an idea which has lost its metaphorical nature and which we can manipulate
easily as though it was real. Our mythology is then complete.
All our thinking tends to be accompanied and aided by spatial ideas, and
we talk in spatial metaphors.... We almost inevitably form pictures... of a
building with its dark underground cellars. . . . If however, we constantly
bear in mind that all such spatial relations are metaphorical . . . we may
nevertheless speak of a consciousness and a subconsciousness. But only on
this condition. (Freud [1895], SE 2:227-28)
A few years later, in The Interpretation of Dreams, Freud both defends his
analogical mode of thought and advises his readers not to reify these tropes. Of
course he uses a metaphor to do so. As he puts it, "We are justified ... in giving free
rein to our speculations so long as we retain the coolness of our judgment and do
 JOSE BRUNNER

not mistake the scaffolding for the building" (Freud [1900], SE 5:536). In later
writings, such as in Beyond the Pleasure Principle, Freud explains that topogra-
phical, dynamic, or economic descriptions of mental processes are not inferred
from empirical observations. Rather, they are a priori categories, a "figurative
language" without which one "could not have become aware" of the mental
processes (Freud [1920], SE 18:60). In the beginning of the essay Freud admits that

the indefiniteness of all our discussions on what we describe as metapsychol-

ogy is of course due to the fact that we know nothing of the nature of the
excitatory process that takes place in the elements of the psychical system,
and that we do not feel justified in any hypothesis on the subject. We are
consequently operating with a large unknown factor [mit einem grossen x]
which we are obliged to carry over into every new formula. (Ibid., 30-31)

Similarly, in "Analysis Terminable and Interminable," he refers to metapsychology

as a "Witch," declaring that "without metapsychological speculation and theorizing
— I had almost said 'phantasying' — we shall not get another step forward.
Unfortunately, here as elsewhere what our Witch reveals is neither very clear nor
very detailed" (Freud [1937], SE 23:225).
Following Freud, I thus view metapsychological concepts both as welcome
supplements to psychoanalytic theorizing, which undermine pretensions at closure,
and as necessary constituents of imaginative thought about the psyche. As Morris
Eagle has rightly pointed out, "The very idea of a purely clinical theory untainted
by any trace of metapsychology is illusory. For example, the very notion of
unconscious wishes and aims, so central to the clinical theory of psychoanalysis,
inevitably entails metapsychological assumptions and considerations" (Eagle 1984,
149). I agree with Eagle that the problems of Freud's metapsychology justify
neither its wholesale denunciation nor its exclusion from psychoanalysis, and that
there is no possibility of a psychoanalysis without implicit and explicit assumptions
about causalities, quantities and hidden dynamics in the mind (For further defenses
of metapsychology, see Meissner 1981; Modell 1981.)
For instance, even though one may acknowledge that Freud's energy concept is
vague and inconsistent and does not conform to natural-science criteria — as
Yehuda Elkana has put it: "Physically it is total nonsense" (Elkana 1985, 229) —
one can still go on to examine its logic more closely. Elkana, for one, has made the
point that Freud uses the notion of psychic energy less in a scientific manner than
as a loose metaphor, as an intuitively accessible image that is part of our culture
and that we invoke "colloquially in talking about our limited mental energies or
about conserving our psychic forces" (ibid.). Indeed, through the cracks in the
mechanistic facade of Freud's energy concepts there appears an underlying eco-
nomic logic that reveals Freud's use of imagery to be a statement on psychic
resources of power and their scarcity (see Fromm 1973, 49). Thus on closer
inspection Freud's references to energy turn out to be related to what he calls "the
 Why Psychoanalysis Needs Metapsychology 89

economic principle of saving expenditure" rather than to a physicalist frame of

thought (Freud [1911], SE 12:222).'

Ill

What conceptions of science and reality could allow Freud to construct unverifi-
able, imaginative metaphors as a legitimate part of his scientific project? The
answer to this question lies in Freud's intellectual background. As Erna Leski, the
historian of the Viennese medical school, suggests, doctors of Freud's era formed
their image of science and human beings not only on the basis of their practical
training in medicine but also through their education in the Kantian spirit of
German humanism (Leski 1965, 314; see also Anderson 1962, 195). The German
and Austrian intellectual tradition of the time was so imbued with Kantian
heritage that the school of Helmholtz — to which Freud's teachers belonged —
developed a type of phenomenalism that, although antithetical to some of Kant's
premises, counted as scientific translation of his philosophy (Galaty 1974; Turner
1977; Leary 1980). The Helmholtz school justified the reduction of life to
chemical-physical mechanics not by reference to the nature of the objects under
scrutiny but by the translation of an epistemological argument into a sensualist
one. According to its presuppositions, a "mechanical analysis of nature" was
entailed by the structure of the human sensory apparatus, which did not allow
human beings to conceive of change independent from mechanical movements of
matter.
Freud alludes to such a quasi-Kantian phenomenalism by stating: "Just as Kant
warned us not to overlook the fact that our perceptions are subjectively conditioned
and must not be regarded as identical with what is perceived though unknowable,
so psychoanalysis warns us not to equate perception by means of consciousness
with the unconscious processes which are their object. Like the physical, the
psychical is not necessarily what it appears to be" (Freud [1915], SE 14:171). This
is not the only comparison Freud makes between the impossibility of direct access
to the mind's internal world and the problems encountered in the investigation of
the external reality. In fact, the influence of quasi-Kantian premises can be seen
throughout his writings. It is evident, for example, in The Interpretation of
Dreams, where he states that the unconscious "in its innermost nature is as much
unknown to us as the reality of the external world, and . . . as incompletely
presented by the data of consciousness as is the external world by the communica-

1
Elsewhere I have elaborated in detail on the advantages of such a commonsensical, metaphorical
reading of Freud's metapsychology and concluded that Freud's metapsychoiogical postulates depict
the mind as a conflictual realm structured in analogy to the political world. In my view this political
construction of the mind constitutes the core of psychoanalytic theorizing (Brunner n.d.). Here,
however, I am not concerned with this aspect of Freud's metapsychoiogical metaphors and limit
myself to a discussion of their role in relation to hermeneutics.
90 JOSEBRUNNER

tions of our sense organs" (Freud [1900], SE 5:613). Almost four decades later,
quasi-Kantian premises continue to underlie a passage in his Outline of Psychoa-
nalysis, in which he compares hypotheses of psychoanalysis to those of physics and
argues:
In our science as in the others the problem is the same: behind the attributes
(qualities) of the object under examination which are presented directly to
our perception, we have to discover something else which is more independ-
ent of the particular receptive capacity of our organs and which approximates
more closely to what may be supposed to be the real state of affairs. We have
no hope of reaching the latter itself, since it is evident that everything new
that we have inferred must nevertheless be translated back into the language
of our perceptions, from which it is simply impossible to free ourselves. But
herein lies the very nature and limitation of our science. . . . Reality will
always remain "unknowable." (Freud [1940a], SE 23:196)
In this spirit Freud states that all that science can achieve are "approximations to
certainty." However, he asserts that such postulates nevertheless enable "construc-
tive work in spite of the absence of final confirmation" (Freud [1916-17], SE
15:51). Again we note that for Freud uncertainty and openness are typical of
scientific procedures in general, and that in this respect he sees no difference
between psychoanalysis and any of the other sciences. In his words, "The processes
with which it is concerned are in themselvesjust as unknowable as those dealt with
by other sciences, by chemistry or physics, for example" (Freud [1940a], SE
23:158). Thus he claims that despite their vagueness, new hypotheses and concepts
in psychoanalysis
are not to be despised as evidence of embarrassment on our part but deserve
on the contrary to be appreciated as an enrichment of science. They can lay
claim to the same value as approximations that belongs to the corresponding
intellectual scaffolding found in other natural sciences, and we look forward
to their being modified, corrected, and more precisely determined as further
experience is accumulated and sifted. So too it be entirely in accordance with
our expectations if the basic concepts and principles of the new science
(instinct, nervous energy, etc.) remain for a considerable time no less inde-
terminate than those of older sciences (force, mass, attraction, etc.). (Ibid.,
158-59)
Let me sum up. The epistemological premises of Freud's phenomenalism legitimize
the vague, metaphorical abstractions typical of metapsychology as a feature
common to all sciences, including physics. According to Freud, it is inevitable for
all disciplines to have recourse to the construction of indeterminate basic concepts
in order to represent an underlying reality of inaccessible processes, structures,
and mechanisms. Thus his self-understanding as a scientist was far more creative,
critical, and complex than most of his supporters and detractors allow for.
 Why Psychoanalysis Needs Metapsychology 91

IV

However, in a field in which only little is left of Freud's adventurous spirit,

metapsychology is continuously denied its role as a vehicle of reflection and
transcendence. One is almost tempted to say — albeit with quite a bit of hyperbole
— that the history of psychoanalytic theorizing since Freud's death is the history of
the struggle to divest metapsychology of its subversive and creative role.
The first step was taken in the 1950s, when the American psychoanalytic school
of "ego psychology" aimed to turn Freud's metapsychological conjectures into
tame, empirically verifiable statements in order to systematize psychoanalytic
theory and thus improve its scientific status. Heinz Hartmann and David Rapa-
port, the two most influential figures of this approach, maintained Freud's metap-
sychological formulations. However, they tried to adapt them to standards of
logical empiricism, which in the 1940s and 1950s not only dominated the natural
sciences, but also swayed the human sciences (Hartmann 1958, 1959; Rapaport
and Gill 1959; Rapaport 1960).
The narrative they constructed and tried to live up to was the positivist one,
which tells of scientific progress from philosophical speculation to the formulation
of operational and verifiable statements. But as Joel Kovel and Martin Klein have
shown in their historical accounts of the "antimetapsychology movement" in the
Unites States, Rapaport's and Hartmann's efforts were counterproductive. Rather
than entrenching the scientific reputation of psychoanalysis, the problematic
nature of their undertaking made it glaringly obvious that metapsychology could
not live up to empiricist standards (Kovel 1988; Klein 1989). Leading psychoana-
lysts of the next generation, such as Merton Gill, Robert Holt, George Klein, and
Roy Schafer — who were Hartmann's and Rapaport's most prominent students —
reacted to the failed scientism of the 1950s by adopting a different approach. The
narrative of an increasingly scientific psychoanalysis, in the positivist sense of the
word, gave way to a new line of thought.
Rather than scientizing metapsychology, they tried to get rid of it. Since the
1970s they have developed the now widely accepted argument that psychoanalysis
can be split into two independent theories: "good," viable clinical theory on the
one hand and "bad" — that is, superfluous — metapsychology on the other (Gill
1976; Holt 1976, 1981, 1985; Klein 1973; Schafer 1976). In this narrative Freud's
metapsychological formulations have been widely denounced by eminent psy-
choanalysts and philosophers as a cumbersome or even pernicious theoretical
ballast that Freud carried over into psychoanalysis from his medical education.
Roy Schafer is probably the contemporary American analyst who has under-
taken the most sustained and sophisticated effort to reformulate psychoanalysis in
a new language so as to provide "a clinically useful and systematic alternative to
metapsychology"(Schafer 1976, ix). The keyword here is "systematic."The subtitle
that Rapaport (1960) gave to what is probably his most influential book also
describes it as "a systematizing attempt," which indicates an attempt at totalization
92 JOS£ BRUNNER

and closure. It seems appropriate to invoke here Freud's criticism of such system-
atization among his followers; for Freud took thinking that was enclosed in
apparently complete systems to be necessarily opposed to scientific thought.
Science, he wrote in the last of his New Introductory Lectures, "is not all-
comprehensive, it is too incomplete and makes no claim to being self-contained
and to the construction of systems" (Freud [1933], SE 22:181-82). Science's
evident imperfection was for Freud one of its crucial characteristics and one that
differentiated it from philosophy and religion, whose illusionary stability and
coherence aimed to soothe their followers and lull them into a misleading sense of
security (ibid., 160-61). Criticizing their paralyzing influence on their believers,
Freud referred to comprehensive thought systems as Weltanschauungen. For
Freud, a Weltanschauung was "an intellectual construction which solves all the
problems of our existence uniformly on the basis of one overriding hypothesis,
which, accordingly, leaves no question unanswered and in which everything that
interests us finds its fixed place" (ibid., 158). And although Freud also referred to
science as a Weltanschauung, he pointed out that contrary to its philosophical and
religious counterparts it did not really deserve this title, since it did not pretend to
be able to answer all questions.

These provisos and stipulations have to be kept in mind when one considers
Freud's famous declarations that "psycho-analysis has a special right to speak for
the scientific Weltanschauung" and that it "is part of science and can adhere to the
scientific Weltanschauung" (Freud [1933], SE 22:158, 181; see also [1940a], SE
23:158; [1940b], SE 23:282). For by themselves such utterances do not explicate
what Freud meant by the terms "science," "scientific," and Weltanschauung.
Nevertheless they have been — and still are — widely quoted out of context and
turned into part and parcel of what one might call the "official" narrative of
psychoanalysis. Propagated by loyal followers and biographers such as Ernest
Jones (1953-57), this narrative places origins and method of psychoanalysis within
a natural-science type of discourse. Frank Sulloway (1979) has catalogued twenty-
six different stories — he calls them "myths" — that together form a narrative of
Freud as The Scientific Hero who started his task as a shunned outsider, a lonely
warrior in the cause of truth, fighting against an evil medical establishment that
rejected his ideas. As Sulloway presents it, the official psychoanalytic narrative
tells of a long and complicated journey full of unexpected dangers and ordeals. It
reports on trials, temptations, and wrong paths taken, temptresses (in the early
seduction theory), a secret helper (Wilhelm Fliess), a daring and superhuman
self-transformation (Freud's self-analysis), loneliness and rejection, an early group
of devoted followers, conquests — and finally, success, victory, and fame (ibid.,
chap. 13).
 Why Psychoanalysis Needs Metapsychology 93

The self-conception of Freud's loyal followers as members of a movement

struggling for scientific achievement and recognition in the natural-science sense
of these terms affected their dissemination of Freud's texts and theories. Such
effects can be discerned, for instance, in James Strachey's Standard Edition of the
Complete Psychological Works ofSigmund Freud (SE), which until recently has
been taken by psychoanalysts and lay readers in the Anglo-Saxon world as an
authentic and faithful rendering of Freud's texts. Indeed, practically all English
commentaries on Freud, including the present article, cite the authoritative
Standard Edition. However, in order to enhance the scientific appearance of
Freud's work by assimilating it to the natural sciences, Strachey's translation got
rid of all instances in which Freud linked psychoanalysis to the Geisteswissen-
schaften (humanities) or characterized it as a Seelenkunde (study of the soul). By
rendering these German terms uniformly and wrongly as "mental sciences," Stra-
chey obscured the significance of crucial statements in which Freud declared his
aim to establish a new science through a synthesis of the natural sciences and the
humanities. Thus the Standard Edition distorts the foundation declaration of
1910, in which the purpose of the International Psychoanalytic Association was
declared to be "to foster and to further the science of psycho-analysis founded by
Freud, both as pure psychology and in its application to medicine and the
Geisteswissenschaften ['humanities,' mistranslated as 'mental sciences']" (Freud
[1914], SE 14:44). Some years later Freud argued that the training most suitable
for analysts was not available in the traditional university curriculum, since it
would have to "include geisteswissenschaftlichen Stoff[le\ements from the human-
ities,' mistranslated as 'elements from the mental sciences'], elements . . . from
psychology, the history of civilization and sociology, as well as from anatomy,
biology and the study of evolution" (Freud [ 1927], SE 20:252). In the same passage
Freud maintained that as long as there were no independent psychoanalytic
institutes in which such comprehensive training could be made possible, a "preli-
minary education in medicine" was recommended. He warned, however, that "it
should not be forgotten . . . that this is not the whole of psycho-analysis, and that
for its other aspects we can never do without the cooperation of people who have
had preliminary education in the Geisteswissenschaften ['humanities,' again mis-
translated as 'mental sciences']" (ibid., 257). Darius Ornston's comment on the
tendentiousness of Strachey's translation is pertinent here. As he points out,
Strachey wanted not so much to place Freud in a specific scientific context or
tradition as to make him speak "like a Victorian Englishman would think a
scientist should sound like"(Ornston 1982, 411; for the contemporary debate on
Strachey's translation see Bettelheim 1984, Brandt 1980; Ornston 1985, 1992;
Timms and Segal 1988).
One widespread response to the official narrative has been to accept the image of
Freud as an aspiring natural scientist of the mind but to debunk him as one who
failed in his enterprise. In this counternarrative — contributions to which vary
widely in terms of their vehemence, scholarship, and sophistication — Freud is
94 JOSE BRUNNER

given the role of The Scientific Fumbler. Thus he is depicted as attempting to

create a new scientific discipline in the natural-science mode but unable to live up
to acceptable standards of theory construction and corroboration. Most bluntly,
this argument has been put forward in books devoted to the demolition of
psychoanalysis, whose titles speak for themselves, such as E. M. Thornton's The
Freudian Fallacy (1986) and Hans Eysenck's Decline and Fall of the Freudian
Empire (1986). Probably the most scholarly attempt to turn Freud into an aspiring
— but ultimately misguided — natural scientist of the psyche is Sulloway's Freud:
Biologist of the Mind, where Freud appears as an heir to Darwin and a "crypto-
biologist," employing "a highly hypothetico-deductive methodology" (Sulloway
1979,421). Since it aims to transpose psychoanalysis from psychology to biology,
Sulloway's book hardly refers to Freud's clinical practice. Thus even though the
bulk of Sulloway's book is devoted to the early years of Freud's work, such seminal
concepts as the "talking-cure" are hardly touched upon and have found no place in
the detailed 45-page index.
The contribution to this genre of literature that has been given the most serious
attention in psychoanalytic and philosophical circles is Adolf Griinbaum's The
Foundations of Psychoanalysis. Griinbaum argues against readings of Freud's
work that present it primarily as part of a hermeneutic tradition concerned with
textual interpretation. Instead, he portrays Freud as a somewhat narrow-minded,
falsifiable and falsified empiricist, who wanted psychoanalysis to be exclusively
part of the natural sciences (Griinbaum 1984, 4-5). To support this claim, Griin-
baum quotes Freud as saying for instance that "psychoanalysis is a part of the
mental science of psychology. . . . Psychology, too, is a natural science. What else
can it be?" (ibid., 2). However, citations such as this one reveal the problematic
nature of Griinbaum's approach. For in the original German Freud said nothing
that would have allowed Griinbaum to cite this short statement in support of a
claim which aims to remove any trace of hermeneutics from Freud's work. In
German, Freud pronounced psychoanalysis to be both "part of psychology's
inquiry into the soul" (ein Stuck der Seelenkunde der Psychologie) and of natural
science (Freud [1940b], SE 23:282). Thus Griinbaum may possibly be seen as one
of the victims of Strachey's translation. Moreover, as John Forrester has demon-
strated, the quoted passage also acquires a meaning opposed to the one attributed
to it by Griinbaum, when one continues reading on to the next sentence. For
directly after saying that psychology is a natural science, Freud qualifies his
statement by adding: "But its case is different" (ibid.; Forrester 1986, 671).
As we see, to turn Freud into a straightforward empiricist one has to recast some
elements of his texts and remain silent about others. For as I have shown, Freud's
work systematically transcends the boundaries of empiricism. For instance, in his
Introductory Lectures Freud states that his insights into the unconscious meaning
of dream symbols had many sources and that they were gained "from fairy tales
and myths, from buffoonery and jokes, from folklore (that is from knowledge
about popular manner and customs, sayings and songs) and from poetic and
 Why Psychoanalysis Needs Metapsychotogy 95

colloquial linguistic usage" (Freud [1916-17], SE 15:159). Obviously, reliance on

evidence from these fields brings psychoanalysis closer to textual and cultural
disciplines such as philology or literature. Moreover, as is well known, Freud's
writings include not only lectures, case studies, clinical guidelines and theoretical
treatises on mental illness, dreams, sexuality, and mental processes in general, but
also a confusing mixture of literary analysis, biographical comments on writers,
and quasi-literary texts. He wrote a book on the structure of jokes, an essay
analyzing the process of creative writing, a "historical novel" on Moses;2 an
interpretation of a fictional dream in Jensen's Gradiva and of a scene from The
Merchant of Venice; reflections on Michelangelo's statue of Moses and on Dos-
toevski's character; an analysis of autobiographical passages by Leonardo da
Vinci and Goethe; a discussion of a seventeenth-century manuscript on demonol-
ogy; a full-length study of the memoirs of a mad German judge. Such publications
can hardly be counted as contributions to medical or psychological discourse in
the traditional sense. Thus even Strachey's scientized reconstruction of Freud's
work cannot conceal the fact that unlike his orthodox successors Freud continu-
ously undermined established conventions of scientific discourse in every possible
way, spread the scope of psychoanalysis over all human endeavors, and explicitly
staked claims not only in psychology but also in philology, philosophy, sociology,
education, aesthetics and cultural history (Freud [ 1913], SE, 13:176-90). In other
words, Freud aimed to establish an interdisciplinary science (see Kitcher 1992).
Aware of Freud's transgressions into the humanities, there are commentators
who have spun Freud into a different narrative, one that is opposed to scientizing
tendencies. Jurgen Habermas, for instance, agrees that Freud wanted to create a
natural-science type of project but claims this to be a "scientistic self-
misunderstanding" to which Freud succumbed. Thus in Knowledge and Human
Interest he places his discussion of Freud between an interpretation of Kant and
Fichte on the one hand and his reading of Nietzsche on the other hand (Habermas
1978). However, rather than simply debunking the scientific reputation of psy-
choanalysis, Habermas' aim is to construct a narrative in which Freud himself
appears as The Great Debunker, on a par with Nietzsche and Marx. Similarly, in
Freud and Philosophy: An Essay on Interpretation, Paul Ricoeur places Freud
alongside Marx and Nietzsche as one of three masters who dominate what he calls
the "school of suspicion" in hermeneutics (Ricoeur 1970, 32). Moreover, both
Ricoeur and Habermas consider Freud's work as being directed not only at
unmasking false consciousness and deciphering the symbolic expressions of its
distortions but also at liberating the individual from its oppressive power (ibid.,
35; Wolfenstein 1990).
Without necessarily sharing the perspective from which psychoanalysis appears
as an emancipatory praxis, other narratives also place Freud primarily in the
history of hermeneutics and belittle Freud's debt to the natural sciences in compar-
2
Freud's title of the first draft of Moses and Monotheism was The Man Moses, a Historical Novel.
96 JOSE BRUNNER

ison to that which he owed to disciplines concerned with the study of language.
John Forrester, for instance claims that "it was the field of philological sciences
that acted as a source of inspiration for Freud's and psychoanalysis' preoccupation
with language" (Forrester 1980, 167-68). Forrester seems to concur with Michel
Foucault's claim that "Freud more than anyone else brought the knowledge of
man closer to its philological and linguistic model" (Foucault 1970, 361). Jacques
Lacan, too, conceives psychoanalysis primarily in linguistic or hermeneutic terms.
In Lacan's words: "If psychoanalysis is to become instituted as the science of the
unconscious, one must set out with the notion that the unconscious is structured
like a language" (Lacan 1977, 160; Lacan 1978, 280).
Obviously, as a hermeneutics psychoanalysis has to be assessed according to
different criteria from the ones that natural science brings into play. Thus such
issues as narrative coherence and "fit," self-consistency and comprehensiveness,
are brought into the foreground by these commentators. Paul Ricoeur, for in-
stance, has formulated the truth criteria of psychoanalysis as those of a narrative
by saying that its aim is "to reorganize facts in a meaningful whole which constitutes
a single and continuous history" (Ricoeur 1977, 861; see also Spence 1982).
Ricoeur holds this position even though he describes Freud's attempted union of
hermeneutics and causal explanation as constituting not only the "central difficulty
in psychoanalytic epistemology" but also as its raison d'etre (Ricoeur 1970, 65).
Throughout his discussion of Freud's writings Ricoeur refers to them as forming a
"mixed discourse that falls outside the motive-cause alternative" (ibid., 363; see
also 394-95). He even seems to reach the conclusion that "we can neither be
satisfied with the Freudian metapsychology, nor find another starting point to
rectify and enrich the theoretical model" (Ricoeur 1977, 857; see also 852-56). But
finally he argues that in terms of the procedures which can be used to establish
their validity, psychoanalytic interpretations are located within the hermeneutic
realm of meaning alone and are subject "to the same kind of questions as the
validity of a historical or exegetical interpretation" (Ricoeur 1970, 374).
In his discussion of the cause-intention merger in psychoanalysis, Jiirgen Ha-
bermas declares Freud's science to be "the only tangible example of a science
incorporating methodical self-reflection," because it "joins hermeneutics with
operations that genuinely seemed to be reserved to the natural sciences" (Habermas
1978, 214). However, Habermas criticizes Freud for failing to recognize psychoa-
nalysis as the self-reflective depth hermeneutics which he, Habermas, proclaims it
to be, and for undergirding it with misleading scientistic models of energy distribu-
tion. In order to develop psychoanalysis in what he takes to be its proper direction,
Habermas, too, suggests replacing causal aspects in Freud's discourse by a herme-
neutics, "which explicates the conditions of the possibility of psychoanalytic
knowledge" (ibid., 254).
British ordinary-language philosophers, and psychoanalysts influenced by them,
agree as a rule as regards this point with the comments made by Ricoeur and
Habermas. Charles Rycroft articulates this position succinctly by stating: "If
 Why Psychoanalysis Needs Metapsychology 97

dreams have meaning and can be interpreted they must be creations of a person or
an agent who endows them with meaning, while, if they are phenomena with
causes they must be explicable in terms of prior events without reference to an
agent. One cannot really have it both ways and the attempt to do so leads only to
confusion" (Rycroft 1981, 4). However, for more than a decade some of the
criticism voiced against Freud by British ordinary-language philosophers has lost
its edge, since the purpose-cause distinction has finally collapsed even in these
circles (Davidson 1980, 3-19; see also Davidson 1982). But as the continuing
controversy around Griinbaum's book shows, the interpretation-of-meaning ver-
sus causal-explanation debate still flourishes and continuously avoids engaging
the challenge inherent in Freud's project: to develop a vision of the mind that
eschews this dichotomy and instead creates a synthesis of explanation and
understanding.
What is common to commentators on both sides of the science-hermeneutics
divide is that they seek to restrict Freud's interdisciplinary work to one discursive
dimension, which they make dominant as representing the "true" Freud. Moreover,
essentially the same strategy of reading is pursued by those who try to rehabilitate
Freud, to free him from a stigma — whatever it may be — and by those who try to
discredit psychoanalysis and unmask its fallacies and errors. They all posit incom-
patibility where Freud refuses the very possibility of separation into two conceptual
worlds, and they set up a dichotomy of two realms that Freud continuously
intertwined. The first is a hermeneutic one referring to the language, reasons, and
purposes or intentions of a subject, whose understanding demands interpretation
and decoding. The other realm deals with the natural-scientific explanation of
causal links and processes, for which mechanistic hypotheses and empirical cor-
roboration are appropriate.
In opposition to these one-dimensional readings of Freud I contend that at-
tempts to "discipline" psychoanalysis by reducing it to one of these two dimensions
rob it of its vitality. For, in my view, the uneasy coexistence of interpretation and
explanation in psychoanalytic discourse gives rise to a creative tension that is
destroyed by relegating Freud to one or other side of the interpretation-explanation
divide. I have thus attempted to criticize such narratives and in the process of this
criticism to construct an alternative narrative on Freud's work that focuses on the
dialectic of metapsychology and interpretation in psychoanalysis. Above all, I
have examined the place that Freud allocated to metapsychology in his interdisci-
plinary science. Metapsychology, I have claimed, served him as a double-edged
sword, both enabling creative and metaphorical thought about the mind's elusive
reality .and revealing the necessary incompleteness of hermeneutics. In order to
pursue this critical, reflexive task, as well as to think creatively about the hidden
world of the psyche, psychoanalysis still needs metapsychology. However, this
does not mean that Freud's metapsychological concepts and categories are sacred
and beyond criticism. On the contrary, it implies that they should provoke further
thinking about the mind's hidden dynamics, forces, and structures, a thinking
98 JOSEBRUNNER

which continues to intertwine the interpretation of meaning with causal explana-

tion and is thus both critical and creative.

Acknowledgments

For their helpful comments, which contributed much to the preparation of this
paper, I would like to thank Eva Illouz, Leah Rosen, Sasha Waitman, Eugene V.
Wolfenstein, and Yael Janette Zupnik.

References

Anderson, O. 1962. Studies in the Prehistory of Psychoanalysis. Norstedts, Swed-

en: Bokvorlaget.
Bettelheim, B. 1984. Freud and Man's Soul. New York: Vintage.
Brandt, L. W. 1980. "Some Notes on English Freudian Terminology," Journal of
the American Psychoanalytic Association 28:461-75.
Brunner, J. n.d. Freud and the Politics of Psychoanalysis. Oxford: Basil Blackwell.
Forthcoming.
Davidson, D. 1980. "Action, Reasons and Causes," in his Essays on Actions and
Events. Oxford: Clarendon Press.
. 1982. "Paradoxes of Irrationality." In Philosophical Essays on Freud, edited
by R. Wollheim and J. Hopkins. Cambridge: Cambridge University Press.
Eagle, M. N. 1984. Recent Developments in Psychoanalysis: A Critical Evaluation.
Cambridge, Mass.: Harvard University Press.
Elkana, Y. 1985. "Borrowing the Concept of Energy." In Comparative Social
Dynamics: Essays in Honor ofS. N. Eisenstadt, edited by E. Cohen, M. Lissak,
and U. Almagor. Boulder, Colo.: Westview Press.
Eysenck, H. 1986. Decline and Fall of the Freudian Empire. Harmondsworth:
Penguin.
Forrester, J. 1980. Language and the Origins of Psychoanalysis. London:
Macmillan.
. 1986. "Hardly." Isis 77.
Foucault, M. 1970. The Order of Things: An Archeology of the Human Sciences.
New York: Vintage.
Freud, S. 1953-74. Standard Edition of the Complete Psychological Works of
Sigmund Freud, 24 vols. Edited by James Strachey. London: Hogarth Press.
Hereafter cited as SE.
. With J. Breuer. [1895] . Studies on Hysteria. SE, 2.
. [1900] . The Interpretation of Dreams, SE, 4:1-338, 5:339-621.
. [1911] . "Formulations on the Two Principles of Mental Functioning." SE,
12:218-26.
 Why Psychoanalysis Needs Metapsychology 99

. [1913]. "The Claims of Psycho-Analysis to Scientific Interest." SE, 13:165-90.

. [1914]. "On the History of the Psycho-Analytic Movement." SE, 14:7-66.
. [1915] . "The Unconscious." SE, 14:166-204.
. [1916-17] . Introductory Lectures on Psycho-Analysis. SE, 15:15-239,
16:243-463.
. [1920] . Beyond the Pleasure Principle. SE, 18:7-64.
. [1925]. An Autobiographical Study. SE, 20:3-70.
. [1927] . "Postscript to The Question of Lay Analysis." SE, 20:251-58.
. [1933] . New Introductory Lectures on Psycho-Analysis. SE, 22:5-182.
. [1937]. "Analysis Terminable and Interminable." 23:216-53.
. [1939] . Moses and Monotheism. SE, 23:7-137.
. [1940a] . An Outline of Psycho-Analysis. SE, 23:144-207.
. [1940b]. "Some Elementary Lessons in Psycho-Analysis." SE, 23:281-86.
. 1985. The Complete Letters of Sigmund Freud to Wilhelm Fliess, 1887-1904.
Edited by J. M. Masson. Cambridge, Mass.: Harvard University Press.
Fromm, E. 1973. "Freud's Model of Man and Its Social Determinants." In his The
Crisis of Psychoanalysis. Harmondsworth: Penguin.
Galaty, D. H. 1974. "The Philosophical Basis of Mid-Nineteenth-Century Reduc-
tionism." Journal of the History of Medicine 29:295-316.
Gill, M. M. 1976. "Metapsychology Is Not Psychology." In Psychology versus
Metapsychology, edited by M. M. Gill and P. S. Holzman. New York: Interna-
tional Universities Press.
Griinbaum, A. 1984. The Foundations of Psychoanalysis: A Philosophical Cri-
tique. Berkeley: University of California Press.
Habermas, J. 1978. Knowledge and Human Interest. London: Heinemann.
Hartmann, H. 1958. "Comments on the Psychoanalytic Theory of the Ego." In
Essays in Ego Psychology. New York: International Universities Press.
. 1959. "Psychoanalysis as a Scientific Theory." In Psychoanalysis, Scientific
Method, and Philosophy, edited by S. Hook. New York: New York University
Press.
Holt, R. R. 1976. "A Review of Freud's Biological Assumptions and Their
Influence on His Theories." In Psychoanalysis and Current Biological Thought,
edited by N. S. Greenfield and W. C. Lewis. Madison: University of Wisconsin
Press.
. 1981. "The Death and Transfiguration of Metapsychology." International
Review of Psychoanalysis 8:129-43.
. 1985. "The Current Status of Psychoanalytic Theory." Psychoanalytic Psy-
chology 2:289-315.
Jones, E. 1953-57. The Life and Work of Sigmund Freud, 3 vols. New York: Basic
Books.
Kitcher, P. 1992. Freud's Dream: A Complete Interdisciplinary Science of the
Mind. Cambridge, Mass.: MIT Press.
100 JOS£ BRUNNER

Klein, G. S. 1973. "Is Psychoanalysis Relevant?" In Psychoanalysis and Contem-

porary Science, vol. 2, edited by B. Rubinstein. New York: Macmillan.
Klein, M. H. 1989. "Throwing Out the Baby with the Bathwater: A Historical
Analysis of the Antimetapsychology Movement." Psychoanalysis and Con-
temporary Thought 12:565-98.
Kovel, J. 1988. "Things and Words: Metapsychology and the Historical Point of
View." In his The Radical Spirit: Essays on Psychoanalysis and Society.
London: Free Association Books.
Lacan, J. 1977. "The Agency of the Letter in the Unconscious or Reason since
Freud." In Ecrits: A Selection. New York: W.W. Norton.
. 1978. The Four Fundamental Concepts of Psychoanalysis. New York: W.W.
Norton.
Leary, D. E. 1980. "German Idealism and the Development of Psychology in the
Nineteenth Century." Journal of the History of Philosophy 18:299-313.
Leski, E. 1965. Die Wiener Medizinische Schule im 19. Jahrhundert. Graz-Koln:
Hermann Bohlaus.
Masson, J. M. 1984. Freud: The Assault on Truth: Freud's Suppression of the
Seduction Theory. London: Faber.
Meissner, W. W. 1981. "Metapsychology: Who Needs It?" Journal oj the American
Psychoanalytic Association 29:921-38.
Modell,A. H. 1981. "Does Metapsychology Still Exist?" International Journal of
Psycho-Analysis 62:391-402.
Ornston, D. 1982. "Strachey's Influence: A Preliminary Report." International
Journal of Psycho-Analysis 63:409-26.
. 1985. "Freud's Conception Is Different from Strachey's." Journal of the
American Psychoanalytic Association 33:379-412.
, ed. 1992. Translating Freud. New Haven: Yale University Press.
Rapaport, D. 1960. The Structure of Psychoanalytic Theory: A Systematizing
Attempt. New York: International Universities Press.
Rapaport, D., and Gill, M. 1959. "The Points of View and Assumptions of
Metapsychology." International Journal of Psycho-Analysis 40:153-62.
Ricoeur, P. 1970. Freud and Philosophy: An Essay on Interpretation. New Haven:
Yale University Press.
. 1977. "The Question of Proof in Freud's Psychoanalytic Writings." Journal
of the American Psychoanalytic Association 25:835-71.
Rycroft, C. 1981. The Innocence of Dreams. Oxford: Oxford University Press.
Schafer, R. 1976. A New Language for Psychoanalysis. New Haven, Conn.: Yale
University Press
Spence, D. P. 1982. Narrative Truth and Historical Truth: Meaning and Interpre-
tation in Psychoanalysis. New York: Norton.
Sulloway, F. J. 1979. Freud, Biologist of the Mind: Beyond the Psychoanalytic
Legend. London: Burnett/Andre Deutsch.
 Why Psychoanalysis Needs Metapsychology 101

Thornton, E. M. 1986. The Freudian Fallacy: Freud and Cocaine. London:

Paladin.
Timms, E., and N. Segal, eds. 1988. Freud in Exile: Psychoanalysis and Its
Vicissitudes. New Haven, Conn.: Yale University Press.
Turner, R. S. 1977. "Hermann von Helmholtzand the Empiricist Vision. "Journal
of the History of the Behavioral Sciences 13:48-58.
Weber, S. 1987. "The Blindness of the Seeing Eye: Psychoanalysis, Hermeneutics,
Entstellung." In his Institution and Interpretation. Minneapolis: University of
Minnesota Press.
Wolfenstein, V. E. 1990. "A Man Knows Not Where to Have It:Habermas,
Grunbaum and the Epistemological Status of Psychoanalysis." International
Review of Psycho-Analysis 17:23-45.

Department of Political Science

Tel Aviv University