WHAT ARE DEFENSE

: RICHARD
S. SIEGAL,PH.D.
MECHANISMS?

T h e comparative historical study of the germ theory and

the vitamin theory leads us to realize that there are at least
2 different kinds of analysis involved in scientific advance.
T h e first kind we may speak of is the analysis by simple
subdivision. T h e type question may be given in symbolic
form: “Is all fruit, f r u i t - o r are there apples and oranges?”
Once the question is asked, success in finding an answer is
almost assured. TVhen one suspects diversity, he usually finds
it. Thus the skier discovers many kinds of snow and the
physician many kinds of fever. . . . T h e second type of
analysis is far more difficult, for it involves changes in the
categories of thought. TVe may call it categorical atialysis.
T h e type question takes this symbolic form: “Is it an apple
or an orange that I’m dealing with-or is it perhaps the
singing of a bird?” So stated, it sounds ridiculous: but
inability to ask such an odd-sounding question has repeat-
edly delayed the progress of science. Consider “heat,” for
example. From the time of the ancient Greeks down to, and
including, the work of Robert Boyle, the facts connected
with heat were terribly confused because “heat” was as-
signed to the wrong category-that of the substantives. Be-
ing a substance, it should have weight, of course; convinced

Richard S. Siegal died on April 3. 1967 at the age of thirty-nine. This paper
was perhaps a draft short of what he might have wished. It was developed from
an earlier collaboration with Dr. Gerald Ehrenreich whose contributions Dr. Siegal
would have gratefully acknowledged in print as he did informally.
From the work of the Psychotherapy Research Project of T h e Menninger
Foundation. This investigation is currently supported by Public Health Service
Research Grant MH 8308 from the National Institute of Mental Health, and was
previously supported by the Foundations’ Fund for Research in Psychiatry and the
Ford Foundation.

785
Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015
586 RICHARD S. SIEGAL

of this, a British physician, George Fordyce, found that heat

did indeed have weight. T h e first experiments of Count
Rumford seemed to confirm this belief. But Rumford was
convinced that the wrong category of thought was being
employed in calorimetric studies, and so he went to a good
deal of trouble to look for experimental errors, which he
found and corrected, thus arriving at the correct conclu-
sion that heat, like the singing of a bird, is an activity, a
process-and not a substance or object, like apples or
oranges. Its category has rather more to do with verbs than
with nouns [4].

hlany of the implications of biologist Garret Hardin’s state-

ments are amply demonstrated in the history of the psychoanalytic
concepts of defense and defense mechanisms and in the cur-
rent clinical and theoretical use of these concepts. It is the intcn-
tion of this contribution ‘to remind the sophisticated, and make
explicit for the novice, that defense mechanisms, which we some-
times call “structures,” are concepts, man-made abstractions-as
is “heat.” Further, they are concepts which have several possible
“real” referents in nature. T h e concept defense mechanism may
denote: (i) mental content (feelings, thoughts, etc.); (ii) mental
outcome or aim, or purposes (a certain fate of an instinct deriva-
tive, a certain state of mind in which one feeling has replaced
another, etc.); (iii) mental process (how the defensive outcome
and/or content comes about). It is our contention that the con-
cept defense mechanism is most fruitfully used to refer to
processes, activities of the human being. Defense mechanisms,
from this point of view, have more in common with such human
activities as walking or running than with such human “struc-
tures” as muscles or glands. They have, in short, “rather more to
do with verbs than with nouns” (4).
Let u s take seriously the notion that “defense mechanisms
are processess”1 and attempt to explore some of the consequences
stemming from subtly bypassing this bit of knowledge. Some.well-
known confusions in this area of psychoanalytic theorizing, for
example, the lack of a satisfactory basis on which to classify de-
1 W c use this shorthand to signify that the referent of the abstraction “defense
mechanism” is, for our present purposes, the intrapsychic procesws which precede
and bring about the defensive outcome and content.

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

 WHAT ARE DEFENSE MECHANISMS? 787

fense mechanisms, may in part be due to this. Reification (i.e.,

regarding a n abstraction or mental construction as a thing) may
have hindered the forward movement of clinical and theoretical
understanding of defensive functioning by allowing us to evade
the task of examining and specifying the mental processes or
actiiities which are defense mechanisms. If these processes were
specified, a sound basis for classifying defense mechanisms could
perhaps be found and our overall understanding of defensive
functioning would be furthered.
T h e distinction between process and outcome is perhaps
worth pursuing for a moment. (Since process and outcome are
terms used commonly in psychotherapy research, let it be clear
that our use oE these terms here refers only to mental or intra-
psychic-not treatment-processes and outcome.) Mental processes
produce movements, changes, transformations in one or another
aspect of the substance, the content, of mental life-that is,
changes in thoughts, .feelings, impulses to act or not to act, percep-
tions of inner or outer stimuli, images, beliefs, memories and so
on. Mental processes are, of course, not directly “visible” or per-
ceivable. They are only inferable. This is not because of their
mental quality-whether they are conscious or unconscious, re-
pressed or not. Rather, in the sense we have in mind, no process
is visible. One can only see difference between what was and
what is, changes in position, or quality or intensity, whether con-
tinuous and smooth or discrete and sporadic. In mental life, one
can see only contents, not processes. Processes must be inferred.
Merton Gill, in the introduction to Topography and Systems in
Psychoanalytic Theory (3) writes, “It is not easy to define a. mental
‘process’; nor is it entirely justifiable to use it interchangeably
with mental ‘content’ in discussions of access to consciousness,
since, logically speaking, one can become conscious only of a
content which is the outcome of the working of the mind, not of
the working itself. Although Freud often discusses mental $roc-
ems reaching or not reaching consciousness, I have attempted to
describe only mental contents as doing so. It must also be pointed
out that, while a mental content is presumably a structure rather
than a process, it seems likely that the only mental structures
which can become conscious are contents.”

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

788 RICHARD S. SIEGAL

Thus any mental process can be “seen” only by way of

inference. JVe see certain changes in content-hanges which re-
sult from or are outcomes of the operation of certain processes.
JVe then infer, that is, construct or reconstruct, the mental
processes which have led to this particular outcome. On the basis
of our perception of‘the mental contents and our inferred recon-
struction of the mental processes which have brought them about,
we postulate an aim or purpose adhering to the processes. Current
imprecise use of the term “defense mechanism” refers equivocally
either to the mental content, the process, or the purpose. Any of
these may be implied in specifying a defense mechanism. This
indiscriminate and unsystematic .reference to varying aspects and
attributes of defensive functioning maintains logical chaos in
regard to the classification and understanding of defense mecha-
nisms since allusion to any mechanism may denote either process,
content, purpose, or some combination of these.
It is perhaps instructive to examine how defense mechanisms
are inferred clinically, in contrast to how they are inferred on
the basis of psychological test evidence, and to consider the corre-
sponding implicit notions of the nature of defensive functioning
held by clinicians who work on the basis of interview material
and those of clinicians who work on the basis of psychological
test data. We start with a clinical vignette.
A female patient who had been in analysis for some time
came for her analytic hour carrying herself with a conspicuously
haughty air, and soon after lying down on the couch expressed
considerable anger toward the analyst. This behavior followed an
hour in which the analyst retrospectively could see no ostensible
reason for the patient to feel angry. H e became aware then that
the previous appearance of this constellation of feelings and be-
haviors was regularly preceded by the arousal of positive feelings
toward the analyst. Apparently these positive feelings provoked
anxiety in this patient. Instead of expressing or experiencing the
positive feelings or even the feeling of anxiety, the patient felt
and acted angry. Thus the content, anger and haughty behavior,
could be recognized as serving primarily a defensive purpose
rather than being simply a more or less appropriate expression
of anger in reaction to some situation or experience. Insofar as

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

 WHAT ARE DEFENSE MECHANISMS? 789

it prevented the patient from saying freely what came to her mind,
including the ideas connected with the positive feelings, ideas
which would be available were it not for the presence of this
anger, this behavior can be considered a form of resistance. Inso-
far as it prevented her from experiencing and/or expressing the
waded-off ideas and feelings it can be considered defensive.
Which defense mechanisms, however, are here being invoked?
Clearly, such a bit of behavior has other. meanings in the
analytic situation. One could infer, for example, a “testing” of
the relationship involved here and probably many other determin-
ing factors. T h e technical use made of the analyst’s recognition of
the defensive function of this feeling of anger will obviously de-
pend on the whole analytic context. It is quite likely he will feel
no need to attempt to infer the particular defense mechanisms
involved in this clinical incident, and certainly he will think
little about the actual processes involved except perhaps to note
and keep in mind that this woman defends herself against unac-
ceptable or threatening love feelings by experiencing the reverse
thoughts or feelings.
In the early days of psychoanalytic writing, some writers felt
satisfied if they could understand the unconscious dynamic mean-
ing of a thought, a bit of.behavior, or a slip of the tongue. Simi-
larly, with the popularization and spread, outside of psycho-
analysis, of defense mechanism concepts, some of us now permit
ourselves to bask in the glow of achievement secured from spotting
a possible defensive purpose of some bit of mental life. Because
it seems clinically adequate, we are satisfied to proceed on the
basis of a somewhat vague conception of defense mechanisms as
simply the defensive use of specific mental contents.
We could understand such a bit of behavior as reported above
as necessitating that we infer the functioning of several underly-
ing defense mechanisms, principally denial and reversal. One
reconstruction of what occurred intrapsychically might go some-
thing like this: the patient experienced thoughts and feelings of
a positive nature toward the analyst. For example, “I felt under-
stood and accepted in the last hour; in fact, I felt loved and I
loved you in return.” These thoughts and feelings may have
been associated with an inner threat, perhaps, for example, feel-

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

790 RICHARD S. SIEGAL

ing vulnerable to disappointment or hurt (or some other threat

to be understood on the basis of her past life experiences), and
thus anxiety was experienced. T h e developing anxiety signal led
to a series of denials of the ideas and feelings, for example, “No,
I don’t love you,” and possibly further, “I didn’t feel loved by
‘you, you were not understanding and accepting.” These denials
were not successful in preventing the further development of
anxiety and thus, we infer, had to be followed by the defense
mechanism reversal, a mechanism characteristic for this particu-
lar patient, which the analysis had revealed as taking different
forms from time to time, such as, “In fact you are mean to me,”
“Really you are just trying to seduce me and trap me,” I am
angry at you and hate you.”
T h e clinician working with psychological test responses may
also infer the underlying functioning of defense mechanisms on
the basis of the content of the test responses, the ideas and the
feelings the patient reports. If he does, the inference process is
substantially the same as the one reported above, differing only
in that the tester knows in a more explicit normative way what
ideas and feelings generally occur to people in response to certain
test stimuli or questions (though he by no means assumes that
any departure from the norm represents the functioning of
defense mechanisms).
Much of the time, however, the psychologist (though, of
course, this kind of inference is made by all clinicians) infers the
presence of defense mechanisms in a different way-not on the
basis of the thoughts, feelings, memories and so on that the pi-
tient reports, but rather on the basis of some particular constella-
tion of formal characteristics of an individual’s modes of thinking
or feeling or perceiving or remembering. Examples are legion.
H e may note that the patient, in answering a question, tries each
time to consider all the possible alternative meanings of the
question, all the possible alternative implications of each of the
possible answers, all of the possible alternative logical routes to
arrive at an answer, and so on. This attempt to be exhaustive, to
leave nothing out, to make no mistakes. may lead the psychologist
to infer that “a reaction formation against the inner impulse to
be careless, messy, and quick is present.” Or, another example, he

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

 FVHAT ARE DEFEXSE hlECIIANISSIS? 79 1

may find in a set of tests, “a general inhibition or constriction of

ideational activity (especially of that kind which is reflective and
serves adaptive reality purposes).” He may also find that the think-
ing of the individual whose tests he views tends to be “naive, ego
centered, unreflective, affect laden, and clichC ridden” (8). On this
bas’is he may conclude that the individual utilizes repression gen-
erally and pervasively, “in order to forestall the possibility that a
remotely associated derivative may provide a path for the con-
scious experiencing and discharge in behavior of the impulse.”
He may then talk about the “rigid repressive barrier” in this
individual.
TVith this kind of inference he puts his finger not so much
upon the immediate operation of a defense mechanism, but
rather says, “this is the kind of thinking which occurs as a result
of the long time use of repression,” or, “this person has lots of
repression.” T h e .language involved, “reaction formation is pres-
ent,” “rigid repressive barriers,” etc., betrays and encourages an
underlying substantive conception of defense mechanisms as
something a person has.
So, depending partly on the evidential basis upon which
one infers the presence of defensive functioning, one’s concep-
tion of the nature of defense mechanisms may emphasize either
their content aspect, as if mental contents are defense mechanisms,
or an implicit notion of defense mechanisms as things, substances
within the person, rather than mental activities of the person.
A definition of defense mechanisms which avoids these in-
correct emphases and captures the clarifications we have offered
above between process and outcome, form and content, construct
and observable, has arisen out of the work of T h e Menninger
Foundation Psychotherapy Research Project.* It is as follows:

A defense mechanism is a construct that denotes a mode of

functioning of the mind. It describes how behaviors, affects,
and ideas serve to inhibit, avert, or modulate unwanted im-
pulse discharges. Defenses, in contrast to defense mechanisms,
2 This definition was written by Robert S. IVallerstein, X D . , in association with
a committee of the Psychotherapy Research Project including Ann Appelbaum.
hI.D., Leonard Horwitz, Ph.D., Otto Kernberg, M.D., Gardner Murphy, Ph.D., Irwin
Rosen, Ph.D.. Richard Siegal, Ph.D., and Harold Voth, M.D.

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

792 RICHARD S. SIEGAL

are behaviors, affects, or ideas which serve defensive purposes.

Their functioning is explained in terms of the operation of
the defense mechanisms. Defenses range from -discrete at-
tributes explicable by reference to the simple operation of a
single defense mechanism to complex behavioral and char-
acterologic constellations that are likewise specific, recurrent,
and serve defensive purposes. These more complex configura-
tions are variously called the defensive operations, defensive
patterns, maneuvers, etc. They are made up of various com-
binations and sequences of behaviors, affects, arrd ideas, the
operations of which are explicable by reference to a variety
of “classical” defense mechanisms, admixed with other ego
activities.

I n a further delineation of this definition, IVallerstein (12)

notes that defense mechanisms, as theoretical abstractions, cannot
of course be conscious, but that the contents of the defenses can
be and usually are unconscious in their defensive working and
can be rendered conscious by psychoanalytic work.
T h e failure to distinguish defenses from defense mechanisms,
the tendency, that is, to confuse mental coiitetifs with mental
processes, subtly stands in the way of achieving further clarity in
our understanding of defensive functioning. T o find a sound
basis for an overall understanding and systematic classification of
defense mechanisms, it would seem necessary to penetrate further
into the nature of the processes which m e defense mechanisms
rather than the contents or purposes.
Since Anna Freud’s (2) classic work in which she listed nine
defense mechanisms which had by then been established in psy-
choanalytic literature (regression, repression, reaction formation,
isolation, undoing, projection, introjection, turning against the
self, reversal, and “a tenth, which pertains rather to the study of
the normal than to that of neurosis: sublimation, or displacement
of instinctual aims”) we have become very adept at identifying
and spotting the defensive intent of many simple and complex
behaviors. TVe have become so adept, as a matter of fact, that
Bibring et al. (1j can list thirty-nine “defenses” divided into “basic
(first order)” and “complex (second order)” categories. Anna
Freud (2) indicated, as do Bibring et al. and many other writers,
that in listing mechanisms of defense “we are including under a

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

 WHAT ARE DEFENSE BlECIIANIS>IS? 593

single heading a number of heterogeneous phenomena.” It has

not been possible to secure unanimity or anything approaching
it in psychoanalytic circles as regards the nature, number, purpose,
and mode of operation-in short, the structure and functioning
of de,fense mechanisms.
Even about repression, that most venerated and most central
weapon in the defensive arsenal, there have been divergent views
depending to some extent on whether one refers to repression as
an outcome or purpose, on the one hand, or a mechanism or
process, on the other.
I n his earliest writings Freud used the concept of repression
to cover all defensive activities. Later, the concept denoted one
among several mechanisms of defense. In discussing “The Place
of Forgetting in Pvlemory Functioning,” Schlesinger (9) implies
the necessity of distinguishing process and content from purpose.
He describes the .“usual paradigm” for repression as “an impulse
derivative threatening to emerge into consciousness” and trigger-
ing signal anxiety. “The opposition of countercathexis or the
withdrawal of attention cathexis, together with the ‘pull from be-
low’ exerted by the mass of the repressed, is then suficient to
interrupt the threatened intrusion.” Schlesinger points out that
the experience (for which we might substitute the term “content”)
associated with this would be a “mental block, an obvious hiatus
or disturbance of the continuity of consciousness”-a point which
was applied to clinical psychological testing by Siegal and Ehren-
reich (10) and which has been systematically studied by Luborsky
(7). “\Vhile blocking is not a rare phenomenon, neither is it as
common as we consider the Occurrence of repression to be. . . . In
any well-run psyche, the balance of repression and expression is
handled smoothly. . . rather than encountering repression accord-
ing to this ‘brute force’ [7] model, we are likely to find it accom-
plished by far more subtle shifts in the stream of thought, barely
perceptible changes of subject or changes of emphases, a shift to a
more cumbersome syntax, a failure to allude to an obvious con-
nection, a shift from a concrete and specific level to a general and
abstract level and so forth. I mention these commonplaces to
bolster the point that repression is not only a specifiable mech-
anism of defense, but can be thought of also as a defensive aim

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

794 RICIIARD S. SIEGAL

or purpose which can be accomplished in different ways involving

a greater or lesser economy of effort and greater or lesser coordi-
nation between the defensive and synthetic functions’ of the ego.
In the examples I just mentioned the purpose of defense has made
use of the capacities of the memory apparatus including the at-
tributes of its logical-conceptual organization and the reliance on
schema formation” (9).
Divergence and confusion in our conceptions of defense
mechanisms are perhaps surprising because despite lack of agree-
ment on the classification or structure and functioning of defense
mechanisms, there has been agreement that the understanding of
defense mechanisms in psychoanalytic theory and in psychoana-
lytic clinical practice is of central importance. Robert IYaelder in
“The Structure of Paranoid Ideas” stated that clarification in this
is particularly important because the “defense mechanism is
largely responsible for the form of pathology and the old prob-
lem of the choice of neurosis can be in part reduced to the prob-
lem of choice of defense mechanism” (1 1).
Classifications of the defense mechanisms have been predi-
cated on different bases. Defense mechanisms have been classified
according to (i) what they do, their outcome or purpose; (ii) when
and where they develop, their genetic origin. IYe should like to
propose the desirability of viewing defense mechanisms according
to what they are, implying classification on the basis of the
processes which comprise, which are the mechanisms. A tentative
attempt at such an approach suggests, as will be illustrated below,
that there are, after all, only a relatively limited number of types
of process which underlie defense mechanisms.
An example of classification according to the outcome or aim
of defense mechanisms may be found in IVaelder’s, “The Struc-
ture of Paranoid Ideas” (11). IVaelder pointed out that defense
mechanisms are of differing complexity and he made a plea for the
development of “an alphabet of defence mechanisms, a catalogtle
of elementary responses.” After commenting that defense mecha-
nisms are described in the literature on many levels, Waelder says
that some “merely indicate what happens to the instinctual repre-
sentation, and to the access to the motor system. Others imply a
special form of countercathexis (e.g., reaction formation). In

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

 W H A T ARE DEFEKSE AlECliANlS1IS? 795

others, like regression and sublimation, changes in the aim or

the object of the instinct are implied.” IVaelder goes on to say,
“It may be that the elementary mechanisms of dealing with in-
stinctual drives can be arranged in a sequence, according to the
greater or lesser degree of chauge they impose upon the instinct.
At the head of the list is, of course, successful repression, which,
for all we know, has put an end to manifestations of the instinct.
It is followed by mechanisms which succeed in changing the aim
or object of the instinct such as, among others, regression or dis-
placement. Then there follows unsuccessful repression i n which
the unconscious cathexis seems to remain unchanged but precon-
scious cathexis has been withdrawn. Then there is isolation in
which not even unawareness has been fully achieved. Finally, in
d e n i a l - o r its most common specimen projection-the ego has not
even tried to fight the instinct.”
JVhile this is clearly a valuable suggestion, one which, in fact,
provides the original- basis out of which grows this attempt to elu-
cidate the relevant processes, a problem seems to arise from at-
tempting classification on the basis of the degree of alteration
brought about in the nature of the discharge of instincts. T h e
problem is to encompass in this conceptualization varying degrees
of “success” of defensive functioning. For example, JVaelder in
the just-quoted paragraph, although this is clearly not his inten-
tion, 5eem5 to treat unsuccessful and successful repression as if
they were two different mechanisms and thus tends to include
in the description of a defense mechanism some of its variable
characteristics‘of operation. It is as if one were to agree to call
a person who attempts to prevent the entrance of intruders into
a bank a guard only if his protective efforts are successful and to
call him something else if his efforts should fail. T h e dificulty
here, though perhaps more apparent than real, serves to point to
the difficulties in attempting to encompass a defensive purpose
and its outcome simultaneously in the classification.
Viewed somewhat differently, it is possible to conceptualize
JVaelder’s scheme as a classification solely by defensive purposes
or aims. That is, the purpose “to c h n g e the aim or object of the
instinct,” as opposed to the defensive aim “that unconscious
cathexis remain unchanged while pre-conscious cathexis is with-

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

796 RICHARD S. SIEGAL

drawn," without reference to the degree of success in bringing

about these outcomes, might exemplify a basis for categorization.
This would reveal, of course, that the double classih'catory basis
is indeed more apparent than real. It might also suggest that
microscopic enough examination of the nature of defensive aims
or purposes leads in the direction of (though it stops short of)
elucidation of the processes themselves. Classification of this
kind, that is, on the basis of what the successful defense mecha-
nism does without specific inclusion of how it does it, fails to
distinguish those defense mechanisms which lead to a visible, posi-
tive, present outcome-in the form of a behavior, a thought, or
an affect-from those which lead rather to a negative outcome-
the absence of something. For example, no distinction is made
between a defense mechanism like reaction in which the coun-
tercathexis is visible through the intensity with which an idea or
affect exactly the reverse of the initial unacceptable impulse de-
rivative is held and asserted and, on the other hand, a defense
like isolation, in which the outcome of the underlying processes
can be seen primarily through the absence of something, a felt
connection between, say, a thought and an affect, which should
be there.
IVhile IVaelder clearly points to one functional aspect of
defensive processes which needs to and can be fruitfully pur-
sued, that is, the effect of the defensive process on the instinct
derivative, the attempt to classify defense mechanisms on this
basis seems to run into some difficulties, T h e root of these diffi-
culties, from one point of view, may be the use of what is essen-
tially a double classificatory basis-the effort to encompass purpose
and outcome in a single description and along a single continuum.
A more serious problem is the lack of attention to the process
aspects of defense mechanisms.
To Waelder, however, must go the credit for emphasizing
that ego psychology has the task of clarifying our understanding
of the structure and functioning of defense mechanisms. His excel-
lent analysis of the structure OE paranoid ideas was an important
step in this direction and led him to suggest the development of
"an alphabet of defence mechanisms, a catalogue of elementary
responses" (1 1). His attempt to distinguish between elementary,

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

 WHAT ARE DEFESSE MECHANISMS? 797

basic, unitary, irreducible processes and more complex ones not

only is an attempt with which we are in wholehearted agreement:
in fact, the principal purpose of this contribution is to clear
away some of the barriers which prevent us from arriving at just
a classification.
A second classificatory dimension which has been used in
attempts to bring some logical order into our understanding of
defenses and defense mechanisms involves a developmental or
genetic approach. Attempts have been made to consider defense
mechanisms on a genetic basis in terms of the order of their ap-
pearance in the development of ego functioning. T h e earliest
appearing mechanisms often are considered to be the most ele-
mentary or “primitive” ones and specific mechanisms are linked
to specific phases of psychosexual development. Anna Freud (2)
for example, said, “Such processes as regression, reversal, or
turning round upon the self are probably . . . as old as the in-
stincts themselves, o r a t least as old as the conflict between instinc-
tual impulses and any hindrance which they may encounter on
their way to gratification.” She adds, “l\re should not be sur-
prised to find that these are the very earliest defense mechanisms
employed by the ego.” Yet it seems unsatisfactory to equate early
origin with later simplicity. For example, reaction formation, a
complex defense mechanism as we hope to demonstrate below,
would presumably have its origin in a period of development
preceding the phallic phase with which repression, a more ele-
mentary mechanism, has been most clearly linked.
Aside from the difficulty of securing consensus upon the
order of appearance of defense mechanisms, and apart from our
as yet incomplete knowledge of the individual consistencies and
variations in the development of ego functioning, one runs into
serious practical and conceptual problems in trying to equate
some of the early ego tendencies and activities with later more
fully developed ego processes. T h e problem of relating so-called
“precursors to defense,” processes which have their origin in a
time when ego development is diffuse, undifferentiated and in
flux, to functioning defense mechanisms in the adult, is a difficult
and unsolved one.
It is perhaps somewhat analogous to determining whether

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

798 RICHARD S. SIEGAL

or not one owns the same old knife after he gets a new handle
for it and then replaces the old blade. T h e adult’s ego within
which and with which defending occurs is quite different from
his ego when he was an infant. Thus, the process of defending
must also, of necessity, be different (granting, of course, some
degree of similarity between infantile and adult egos and de-
fensive processes). T h e content defended with, and against which
defenses are arrayed, is also different. JVhat then remains un-
changed enough to make it reasonable to postulate the “precur-
sory” relationship? Clearly, since what is defended against, what
does the defending, and thus how it is done, are considerably
(though not completely) changed, it must be the outcome or
purpose, the state aimed at (whether it be unawareness, inability
to act, or a neutral affect state) which is least changed. T h e “old
knife” with new blade and new handle is “old” only in some
conceptual sense-that is, it belongs to the same organism and
fills the same purpose. Thus a precursor to a defense mechanism
implies a conceptual similarity in purpose between two con-
structs rather than concrete similarities in processes or contents.
T h e genetic point of view cannot be sufficient by itself to
help us understand and adequately conceptualize the phenomena
of adult defensive functioning since this approach too encounters
the empirical problem that it is rarely clear which processes com-
prise given defense mechanisms. Nor is there reason to assume
that all processes used in the service of defense develop out of a
matrix of conflict. Defense mechanisms do not, of course, spring
into existence full blown. Following Hartmann (5), it is clear
that some ego functions develop within the conflict-free sphere
of the ego and may not thus be implicated in conflict or defense
at the time of their development. They may be enlisted only
later in the service of defense. IVe here touch upon the crucial
question of the relationship between defense and adaptation
which is in the forefront of a good deal of psychoanalytic think-
ing these days. IVhile it is of great interest to classify defense
mechanisms and defensive functioning along some continuum
of contribution to adaptation, a prior necessity is the conceptual
separation of processes of defense from defensive outcomes or
purposes.

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

 WHAT ARE DEFENSE ~IECHANISJIS? 799

This leads us to propose the necessity of viewing and classi-

fying defense mechanisms on the basis of what they are-that is,
on the basis solely of the ego processes involved without regard
to their origin or outcome. Only this way can IVaelder’s ini-
portant suggestion that we develop “a catalogue of elementary
respo’nses” be implemented.
An attempt to differentiate defense mechanisms according
to .their complexity has been made by Bibring et al. (I), but
while the intent of these authors is sound, their execution mani-
fests certain difficulties. Acknowledging that in writing a “Glos-
sary of Defenses” their intent is n’ot primarily a classificatory one,
they go on to state, “It appears that there is’a continuum OE de-
fensive measures making up the defensive organization of the
ego. T h e extremes can be readily distinguished from each other,
but there is an indeterminate middle range which defies exact
specification regarding those defensive functions which justify
explicit specification as defense mechanisms, and those more
complex measures made up of various combinations and se-
quence of defense mechanisms and admixtures of other ego func-
tions. However they are so closely related to those relatively
irreducible defense mechanisms as to justify inclusion in a tabula-
tion of defenses” (1). T h e authors then list and describe twenty-
four defenses which they categorize as “basic (first order)” and
fifteen defenses which they categorize as “complex (second
order).”
Their attempt to differentiate complex from irreducible de-
fense mechanisms fails on two counts: (i) failure to distinguish
sufficiently between process, content, and outcome or purpose;
(ii) failure to clarify suficiently the meaning of “irreducible.”
To illustrate these points let us consider some examples.
Bibring et al. include the defense mechanism reaction formation
in their list of “basic, first order, relatively irreducible” mecha-
nisms. They define and describe reaction formation as “the man-
agement of unacceptable impulses by permitting the expression
of the impulse in an exactly antithetical form; in effect the es-
pression of the unacceptable impulse in the negative. Reaction
formation may be a temporarily invoked defense mechanism,
but on the basis of a persistent instinctual conflict, i t may become

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

800 RICHARD S. SIECAL

embedded in the developing ego structure in the form of a

character trait on a relatively permanent level” (1). Reaction
formation is inferred when some aspects of observable or re-
portable behavior (including thoughts, words, actions, or atti-
tudes) are considered to be the reverse of unconscious impulse
derivatives. This means it is necessary to infer the presence of the
processes we conceptualize as reaction formation in order to ac-
count for the appearance of behavior which is the reverse of in-
stinct derivatives believed to be operative in an individual.
Typically, reaction formation is inferred when a person behaves
overtly, for example, in a very kindly or passively compliant
fashion while the observer has reason to feel there exist covert,
unconscious impulses to behave in an actively hostile or directly
defiant fashion.
Close inspection will reveal that it is necessary to infer the
operation of several interrelated processes to account for thC out-
come which causes one to infer reaction formation to have been
operative. This is another way of saying, of course, that reaction
formation does not seem to us to be unitary, elementary or irre-
ducible from a process point of view. Since one outcome of re-
action formation is that an instinct derivative is kept unconscious,
it’can be said that repression is one process operative in reaction
formation. By repression here we mean the barring from con-
sciousness of the ideational representative of an unacceptable im-
pulse and more or less closely related ideational representatives.
( T o be quite technically correct it would seem to us, though we
are aware that there are other points of view, that one can speak
only of repressed ideas not of repressed affects. If affects are defined
as a certain class of discharge processes-as they have traditionally
been in psychoanalytic theory-it does not make sense to speak of
barring discharge processes from consciousness through repression.
In a kind of clinical shorthand, we do speak of repressed or un-
conscious affect in order to convey the idea that certain specific
potential channels of impulse discharge have become associated
with certain ideas or classes of ideas which may be unconscious.
These specific channels of impulse discharge are utilized or are
triggered off when the specific associated ideas arise-unless some
processes interfere. Thus, repressed or unconscious affects exist

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

 WHAT ARE DEFENSE MECIIANISMS? 80 1

only as potentials, like the potential energy in a coil spring, in

contrast to repressed ideas which exist in the form of latent con-
tents, analogous to latent images on undeveloped photographic
film.) Thus, to say that repression is operative in reaction forma-
tion points to a process involving the withdrawal and withholding
of attention or consciousness from certain ideas. In other words,
reaction formation includes a process of not paying attention to,
in the loose sense, or of turning one’s consciousness away from, in
a still looser sense, certain ideas.
Secondly, when reaction formation has taken place, not only
is an idea or a group of ideas kept from consciousness, but the
affects, the feelings, and emotions associated with the particular
unconscious impulse are displaced from their original object onto
a different object. This means that a feeling, within the self, is
experienced as referring to, being directed toward, a person (e. g.,
the boss) different from the one originally and unconsciously in-
tended (e.g., father). Since reaction formation is often persistent,
in fact, characterological, displacement from the original object
(say, from father to all authority figures) is a characteristic, if not
invariable, component of reaction formation. Such displacement
implicates (i.e., is the outcome of or necessitates that we infer)
another process or set of processes which can be distinguished
from those involving attention since their outcome is seen not in
unawareness of certain thoughts but in shifts of emotional direc-
tion. T h e nature of processes which bring about such displace-
ments of affect is not yet fully clear.
Thirdly, in order to infer the operation of the group of
processes we designate reaction formation it is necessary that the
original urge be represented in consciousness only as its opposite.
(“I wish only to submit myself to the authority of my father or
my boss”-as a conscious thought-affect experience, while the un-
thought-of, unfelt aspect would be the wish to destroy father.) And
this transformation is still of a different kind, in that the opera-
tion of the process is still obscurc, though it can be seen as in-
volving a change in the relation of the self to the wish which we
designate reversal.
It is this latter aspect of the complex of processes making u p
reaction formation that is most readily visible as an’ outcome

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

802 RICIIARD S. SIECAL

which makes it superficially plausible to describe reaction forma-

tion as “basic or irreducible.” T h e purpose or outcome of reaction
formation is the experience of an unacceptable wish only as its
opposite. This “opposite,” this reversed impulse derivative, is, in
our terminology, a defense, a mental content, brought about by
the operation of a defense mechanism which is reducible to a
group of processes operating together either simultaneously in
time or in a certain more or less invariant and organized sequence.
(Processes which operate together but in a more variable, loosely
organized sequence may perhaps result in defenses which are com-
plex and uniquely characteristic of given individuals-particular
character traits or styles, for example, like a certain person’s ironic
sense of humor or view of the world.)
While there is general recognition that, as Hoffer (6) puts it,
.
“defensive processes . . are not only patterned and organized as’
defence mechanisms, but also come to interact with each other in
the course of mental development,” not much has been done to
go beyond this acknowledgment to investigate the clinical and
theoretical ramifications arising from an attempt seriously to de-
scribe the particular combinations of processes in defense mecha-
nisms. It becomes possible to think in more detail about such
combinations when we distinguish mental activities which take
place at a certain time and are of a certain duration. This raises,
as a speculative aside, the possibility of exploring defensive activi-
ties, in general, with respect to the sequential relationships among
the interrelated mental processes involved.
Reaction formation, then, seems to us not a defense mecha-
nism which can be understood as single, unitary or irreducible.
Rather it is a constellation of processes which operate together-
processes which include: (i) repression of an idea from conscious-
ness; (ii) displacement of an associated affect to a different object;
(iii) reversal of the repressed wish (or idea-affect complex). I t is
clear, however, that when one infers reaction formation fiom
tests or clinical interviews, one cannot separately identify repres-
sion, displacement, and reversal. T h e mechanism functionally
speaking is unitary: Conceptually, however, it can be analyzed into
component processes.
T h e same kind of conceptual analysis can be carried through

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

 WHAT ARE DEFENSE MECHANISMS? 803

with other defense mechanisms seen by Bibring et al. as relatively

“irreducible,” for example, acting out, which is described as serv-
ing “as a resistance against conscious recognition of an impulse.
T h e unconscious fantasy, involving objects, is lived out impul-
sively in behavior.”
A simple hypothetical instance might be that a woman pa-
tient has a sexual affair with a man in which she acts out a re-
pressed sexual wish toward her analyst, which unconscious wish is
a displaced version of wishes originally directed toward her father
many years before. On the face of it several different processes are
specified here:

(1) repression of the ideational presentations of the sexual

wish (“I want analyst and I want father”);
(2) displacement of affect from one object to another (sexual
feeling experienced towards the men, not the analyst or
the father):
(3) carrying the displaced affect and repressed idea (in a
form, incidentally, which may have been altered by proc-
esses such as symbolization and fantasy formation) into
action (possibly as sexual intercourse, possibly in some
other form) via some discharge through the motor ap-
paratus or “opening the sluices to motility” (here a “mys-
terious leap” from intrapsychic representation to the
somatic motor apparatus has occurred).

So, even in the simplest paradigm of acting out, several

processes are implicated, only one of which, “opening the sluices
to motility,” stands for or is taken for the whole concept.
Perhaps these two examples will suffice to demonstrate the
point that the Bibring et al. approach can classify certain de-
fense mechanisms as unitary or irreducible only because the
aspect of the designated defense mechanisms to which they refer
is a certain outcome. If one refers instead to processes, the de-
fense mechanisms seem Ear from irreducible.
W e use the Bibring “Glossary of Defenses,” which is a very
useful compendium, only as the vehicle to make the point that
conceptualizing a defense mechanism on the basis of its outcome

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

804 RICHARD S. SIECAL

(for example, reaction formation on the basis of a visible attitude

the exact reverse of the inferred unconscious one, or acting out
on the basis of visible actions corresponding to the inferred un-
conscious wishes) covers a kind of pars pro toto thinking, or
reification, which inhibits further understanding of defense
mechanisms.
Perhaps this can also serve to call attention to the question
of the meaning of “elementary,” “unitary” or “irreducible.” If
defense mechanisms really are constructs (which implies that they
exist in the minds of the beholders, while only contents which
serve defensive purposes or eventuate in defensive outcomes exist
in the minds of the persons doing the defending), then it follows
that the reduction referred to is conceptual reduction, moving to
a less abstract, a more concrete, level of discourse.
T o move from the level of designating defense mechanisms
as the outcomes of certain processes, down to considering the
processes themselves, seems to be one step in such a reduction.
For example, to note that defense mechanisms are combinations
of processes which transform thoughts, grant and withhold con-
sciousness from thoughts and affects, and lead to or prevent
motor or somatic behavior, is to take a step in reduction. T o
specify, for a defense mechanism, which processes it comprises
(whether remembering, attending, fantasying, thinking, com-
paring) and in which sequence they occur, is also reducing. Re-
ducing still further, at some future date it might well be possible
to be able to understand transformations in thoughts, feelings,
and movements, and the associated transformations in “mental
energy” in more physiological, chemical, and physical terms. TVe
may some day be talking about chemistry, electricity, and neuro-
physiology when we are trying to describe functions and struc-
tures underlying behavior and those special processes we call de-
fense mechanisms. There is no implication, however, that this
would necessarily be desirable. T h e desirable level of discourse is,
of course, determined by the purpose at hand. For our present pur-
poses, the search €or a sound basis for the further understanding
and classification of defense mechanisms, we must stay within the
realm of psychology, on the level of mental events. Even upon
this level it seems desirable to reduce conceptually some defense

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

 WHAT ARE DEFENSE MECHANISMS? 805

mechanisms thought to be unitary to the combined working of

several more elementary intrapsychic processes. This is a job of
un-reifying defense mechanisms and involves attention to minute
details of defensive processes. Such reduction suggests there may
be only a relatively limited number of kinds or classes of proces-
ses, *which, combined in various sequences, are defense mecha-
nisms.
Our process descriptions of reaction formation and acting
out seem, for example, to have implicated essentizlly three dis-
tinguishable classes of intrapsychic processes:
(1) those which grant and withhold consciousness to one or
another aspect of the mental representatives of intra-
psychic conflict or defense;
(2) those which relate the self to various aspects of the con-
flict representatives;
(3) those which open and close access to motor and other
somatic systems.
T h e first type of process, of course, is repression in all its
various guises. T h e point has frequently been made that repres-
sion is present in all defense mechanisms. This can be taken to
refer to the fact that in all defending, something, whether it be a
warded-off idea, a connection between idea and affect, or some-
times, merely the fact that one is defending one’s self, is kept
from consciousness.
T h e second type of process places the self in spatial or affec-
tive relation to inner representatives of the conflict or wish-
counterwish situation. This can be done in a number-of ways-
for example, by placing a wish, a feeling, a thought, inside or
outside one’s self, putting one’s self behind an impulse or other-
wise shifting the position or attitude of one’s self so that an im-
pulse is experienced only as a fear or in a reversed way. This
type of process is most apparent in defense mechanisms such as
projection, introjection, reversal, etc.
T h e third group of processes seems to involve the relation
between intrapsychic representatives of wishes and the somatic
apparatus. I n acting out, for example, there are mysterious proc-
esses which control “the sluices to motility.” In other defense

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

806 RICHARD S. SIECAL

mechanisms, there seem to be processes which “control the slui-

ces” to other aspects of the somatic apparatus, for example, the
circulatory or gastrointestinal systems, etc.
One may ask: to what extent are these three classes of
processes generally applicable in describing defensive function-
ing? TVhat other classes of processes need to be distinguished to
enable us completely to describe defensive functioning? Can one
understand defense mechanisms to be relatively highly organized
and relatively invariant sequences of a certain limited few
classes of processes?
Our intention has been simply to clarify and remind our-
selves and our colleagues that defense mechanisms are con-
structs, man-made abstractions, which may be taken to refer
either to mental contents, purposes, or outcomes, or to certain
processes. T h e tendency to refer, when speaking of defense
mechanisms, indiscriminately to any or all of these aspects of
defensive functioning has led to considerable confusion and
unclarity. TYe have tried to demonstrate that it is necessary to rid
ourselves of the archaic tendency to take a concept as a thing or
a substance. Only if we are able to do this can we deepen our
systematic understanding of this aspect of intrapsychic function-
ing. To describe defense mechanisms as composed of certain
processes and to specify the nature and sequential organization of
these processes, these mental activities, as we ha’ve tried to do, per-
haps helps to open the way toward a sounder basis on which to
understand and classify defense mechanisms. Surely our under-
standing will be enhanced when we are able to describe defense
mechanisms simply as ways of relating one’s self, one’s conscious-
ness, and one’s body to inner stimuli.

Summary
T h e archaic tendency to take a concept as a thing or a substance
leads to confusion and unclarity in the theory of defense mecha-
nisms, and inhibits the fullest understanding and classification
of them. Defense mechanisms are constructs, man-made abstrac-
tions, which may be taken to refer either to mental contents,
purposes, outcomes, or to certain processes. I n this paper the

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015

 WHAT ARE DEFENSE MECHANISMS? 807

author attempts to specify the nature and sequential organization

of these processes as he describes defense mechanisms as ways of
relating one’s self, one’s consciousness, and one’s body to inner
stimuli.

BIBLIOGRAPHY

1. Bibring, G. L. e t al. A study of the psychological processes in pregnancy and of

the earliest mother-child relationship. T h e Psychoanalytic Study of the Child,
16:9-72. New York: International Universities Press, 1961.
2. Freud, A. T h e Ego and the Mechanisms of Delense (1936). New York: Intern+
tional Universities Press, rev. ed.. 1966.
Gill, M . M . Topography and Systems in Psychoanalytic Thcoty [Psychologicnl
Zsssites, BIonogr. 101. New York: International Universities Press, 1963.
Hardin, G. The threat of clarity. Amer. J. Psychiat., 114:392-396, 1957.
Hartmann, H. Ego Psychology and the Problem of Adaptation (1939). New
York: International Universities Press, 1958.
Honer, \V. Defensive process and defensive organization. Int. J . Psycho-Anal.,
85~194-198,1954.
7. Luborsky, L. A psychoanalytic rescarch on momentary forgetting during free
association. Bull. Phila. Assn. Psychoanal., 14:119-157, 1964.
8. Schafer. R. Psychoanalytic Interpretation in Rorschach Testing. New York
Grune S: Stratton, 1954.
9. Schlesinger. H. J. The place of forgetting in memory functioning. Abstr. in
Panel: Memory and rcprcssion, rep. W. G . Niederland. This Journal, 13:619-635,
1965.
10. Siegal, R. & Ehrenreich. C. Inferring repression from psychological tests. Bull.
hfenninger Clin., 2682-91, 1962.
11. Waelder. R. The structure of paranoid ideas. Znt. J . Psycho-Anal., 12:167-177,
1951.
12. \Vallerstein. R. Panel report: Development and metapsychology of the defense
organization of the ego. This Jotrrnal, 15:130-149, 1967.

Submitted September 1, 1967 by Stephen A . Appclbatrm, Ph.D.

Box 829
Topeka, Kansas 66601

Downloaded from apa.sagepub.com at OAKLAND UNIV on June 14, 2015