DESPERATELY
SEEKING ATTACHMENT:
A PSYCHOANALYTIC
REFRAMING OF THE
HARSH SUPEREGO
Elizabeth F. Howell, Ph .D.
Elizabeth F . Howell, Ph .D., is Adjunct Assistant Professor of
Psychology at New York University in New York, New York.
For reprints write Elizabeth F . Howell, Ph .D., 111 Hicks St .,
Apt ., 5P, Brooklyn, NY 11201.
ABSTRACT
This article presents a new model of superego organization that is
dissociation-based . Clinicians often work intuitively with the harsh
superego as if it were a dissociated internal persecutory fragment or
self-state, even though theory has not provided a clear conceptual
basis for this kind of approach . Despite differing theoretical under-
standings of the nature and origins of the harsh superego, there
appears to be consensus that the harsh superego is best approached
therapeutically with the aim of softening its hold.
Although the term superego is commonly used, its meanings
are not always clear . Superego is understood both as a source of psy-
chopathology and as a moral agency . Understood as psychopathol-
ogy, the meaning of superego is problematic, as the theoretical model
of superego frequently differs radically from the clinically observed
phenomena . Understood as morality, the meaning of superego is also
unclear in that it can be relativistic and subject to corruption
(Kohlberg, 1971 ; Sagan, 1988) . In addition, the interrelationships
of these two meanings can be confusing.
This article begins by describing some of the problems in the
superego construct . It then recasts this construct in terms of attach-
ment theoryand dissociation . It is proposed that in many cases harsh
superego may be understood more usefully in terms of dissociation
than in terms of the structural model and the Oedipus construct.
There may be reasons in the history of psychoanalytic theory that have
obscured this possibility. The role of attachment in superego and moral
development (Schore, 1997 ; Lewis, 1981, 1983 ; Wilson, 1983) is
emphasized. It is proposed that it is the attachment, not the rule-fol-
lowing in itself, that is moral . The implications of this model for
working more effectively with self-punitive and self-critical behavior
are addressed.
PROBLEMS WITH THE SUPEREGO CONSTRUCT
Regarding the psychoanalytic literature 's "troublesome "
treatment of the superego, Helen Lewis, a noted scholar on
the topic of shame and guilt, stated, "The superego, although
formally acknowledged to be a theoretical construct, was nev-
ertheless treated as an established fact or explanatory sys-
tem " (Lewis, 1990, p . 239) . This was also true of the Oedipus
complex.
The superego construct involves identification and
introjection in response to the oedipal conflict . The child
attempts to solve the conflict presented by incestuous and
murderous oedipal wishes and consequent fear of punish-
ment (counter-aggression, such as castration or loss of love
from the parent) by identifying with the parents . The author-
ity of the parents is introjected into the ego and forms the
kernel of the superego . Actual or anticipated parental pun-
ishment for transgressions is internalized as guilt and shame
which become prime motivators of moral behavior . The
child 's superego takes the place of the parents ' superegos
in the delivery of proscriptions and prescriptions ; in this way,
the normative values of the culture are preserved across gen-
erations.
However, the superego can become overly harsh and hos-
tile, causing excessive guilt and/or self-punishment in
response to conflict over wishes or deeds of transgression,
generating neurosis and psychopathology . This notion, that
the same agency which motivates morality also causes psy-
chopathology, is highly problematic . This hypothetical con-
nection between morality and neurosis has profound impli-
cations, including both the potential legitimization of abuse
of self and others as inherently moral, and the potential triv-
ialization of morality itself.
THE PATHOLOGICAL HARSH SUPEREGO
The pathological harsh superego, rather than always
holding the individual to higher moral principles, can cause
a person to do terrible, even immoral things . It can make
people sabotage their own success and behave punitively
toward those they love . The functioning of the harsh
superego can be seen as compellingly similar to that of a dis-
sociated, aggressive, internal persecutor self-state . In describ-
ing the dissociative aspects of (moral) masochism in earlier
work, I (Howell, 1996) suggested that the self-punishment
in masochism originates in a self-fragment that has become
organized around protecting the individual from further trau-
ma . The self-torture of masochism involves the action of one
part of the self against another, as dissociated self-states . This
230
 
DISSOCIATION, Vol. X. No, a, December 1991
HOWELL
self-organization is similar to that of a severe superego.
The dissociated fragment of self, the self-critical
piece, may be the mediating mechanism of self-
criticism in both masochism and some types of
depression ....In fact, Cameron and Rychlak
(1985) note that the depressive's superego
behaves "as if the two parts were two persons ."
(p . 301) Perhaps the all-too-vague notion of
"aggression turned against the self" may in some
cases best he described by this psychic organiza-
tion, invol v . (Howell, 1996, p . 437)
How does such a self-organization come about? I pro-
pose that it originates in a trauma-laden attachment . Bowlby
(1969) has presented evidence and theory that the human
infant is hard-wired for attachment in the service of survival.
Proximity to an attachment figure serves survival by provid-
ing the infant protection against predators . Therefore, the
child whose parent or caretaker is the predator faces a dilem-
ma . How does the child handle the risk of separation when
the parent or attachment figure is overly punitive or is preda-
tory? To maintain attachment in the service of survival, the
child may employ a dissociative solution which aims to pre-
vent any behavior that could provoke attack or abandonment
from the attachment figure.
Unless the child is otherwise threatened, attack or
frightening separation will normally elicit anger (Bowlby,
1973) . Under conditions of severe threat, the child's ability
to stay attached may depend upon the dissociative com-
partmentalization of anger and aggression . The intolerable
rage at being abandoned and/or abused may be dissociat-
ed, developing into a protector/persecutor self-state which
holds the aggression and vigilantly monitors the child 's behav-
ior . Now the child perceives his or her own angry behavior,
rather than that of the abuser, as the threat (Beahrs, 1983;
Blizard, 1997b ; Blizard & Bluhm, 1994 ; Goodman & Peters,
1995 ; Howell, 1997) . The harsh persecutory aspect of this
self-state arises from the facts that 1) it holds the aggression
and 2) the execution of cruelty can be self-reinforcing . In
many ways this functions like the harsh superego.
What is the nature of the harsh superego? Is there value
in thinking of it as a dissociated structure? Although superego
may be thought of as an abstract source of ideals, prescrip-
tions, and proscriptions, when harsh - it has often been
observed to be more or less personified - as if it were a dis-
sociated self state . One patient has named it her "Silencer ."
Others speak of "the Critic" or "the Censor" . Watkins and
Watkins (1997) refer to a patient with a "super-ego entity"
ego-state called "the Old One" which "demanded that he con-
stantly study and do everything `rational' - no foolishness ."
(p . 88)
CASE HISTORY
The following is a case description of person who is
plagued by a very harsh superego . Clearly a highly moral,
thoughtful, sensitive, considerate, intricately conscientious,
and high-functioning person, John is also extremely judg-
mental of himself . He projects this judgmental quality onto
significant others . In a way that feels inescapable to him, he
feels that he must meet his own and others' standards and
demands perfectly . As a result, he works himself like a dog,
frequently feels exhausted and resentful, and sometimes feels
suicidal.
John is married with two children, both boys ; he works
hard as the business manager of a new car dealership . He is
a very conscientious son, husband, father, and employee . It
seems to him that everybody wants more than "their share "
of him, and he over-functions to provide it . Recently, con-
scious resentment and rage has erupted along with intense
suicidal ideation and impulsivity.
As a child, John was terrified by the verbal and physical
abuse he received and witnessed from his father . He saw his
one sibling, an older brother, being chased around the house
and beaten by his father with an aluminum baseball bat.
While enacted in the name of discipline, his father's abuse
was inconsistent and usually little more than an outlet for
his amorphous rage . The father's only addiction was to this
cruelty . In John's family of origin, it was his father 's will that
children were supposed to have big ears and no mouths . As
a result, the injuries and humiliations that he suffered in the
family were, as Freud described his hysterics' sufferings, "suf=
fered in silence, " (Breuer & Freud, 1955, p . 8), and the long-
ings and rage that he felt were "strangulated " (Breuer &
Freud, 1955, p . 17) . As a child, John learned to stay out of
his father's way as much as possible and developed a highly
inhibited but eager-to-please style of interaction . Not sur-
prisingly, by the time he reached adolescence, he had begun
elaborating suicidal fantasies . Despite a few conflict-ridden
failures as an adult, John has been able to please almost every-
body with a claim on him, more or less, because of a com-
bination of extraordinary intelligence and high self-disci-
pline, until the occurrence of recent triggering events.
The clinical presentation is of two co-existent, self .
states . The normal, usual, conscious and conscientious self
would not dream of hurting anyone . This self has very high
and firmly articulated moral standards, as well as intense con-
scious self-restraint and guilt about causing harm to others.
The other self-state has only recently become spontaneous-
ly expressive of intense, violent, destructive rage, which is at
times amorphously homicidal, sometimes Kamakazi-like,
sometimes simply suicidal . While the first self-state is highly
solicitous of all the significant, resented or hated others, the
second self-state says things like "When I feel like this, noth-
ing else and no one else really matters to me . . . I'll really
show them by killing myself," with a smile on his face . On
231
DISSOCIATION . Vol. X. No. 4, December 1997
DE PE ` TELY EEKIN T i r y
the one hand, this person is terrified of displeasing his par-
ents, boss, or wife ; on the other hand, he has such a head of
fury and triumphant rage that he is willing to hurt them in
the most powerful way he can . Clearly, the self-state that is
joyfully murderous does not understand that killing the con-
scientious John would also be the end of him as well . Clearly
the conscientious John is being held hostage in the service
of the "mad" John that furiously wants to get back at and
"show "others what he needs . Both self-states have their own
delimited experience of the situation : the first . feels helpless,
and the second feels omnipotent.
It took highly potent stressors, as well as the advantage
of already being in therapy, to allow the spontaneous emer-
gence of the "mad" John, but clearly he has been in exis-
tence and quite active all the time . From a perhaps carica-
tured classical framework, one would work with John to soften
his harsh superego, by working with the transference of the
patient, as he usually presents himself, tracing his history,
etc . His compulsivity and his masochism would probably be
seen as neurotic . In the view of this author and according to
the proposed formulation, splitting along good/bad lines is
not his predominant problem . However, pathological dis-
sociation, involving the sequestering of self-states from each
other, is a major problem . This distinction will be discussed
later .
The conscientious but cruel self-state, which in our cur-
rent lexicon of concepts would be called "superego," is also
not part of a unified self . The important thing to notice is
that this harsh, pathological "superego "state has a different
experience from John's usual self . It smiles when it talks about
doing incredible violence, violence that the ordinary con-
scious self would find unthinkable.
Because of this violent rage it has been necessary for treat-
ment to be multifaceted, involving safety checks, cognitive
components, sometimes medication, as well as transference
analysis . But the most important difference is in the under-
lying theoretical focus . Instead, of trying to soften, a harsh,
by implication unwanted and "bad" superego, the therapist
can welcome this harsh, dissociated "mad " self-state, along
with its own very important message which it needs to com-
municate.
CASTRATION ANXIETY, CHILD ABUSE, AND
DISSOCIATION
Is this "superego" ofJohn's, replete as it is with high moral
standards, more usefully thought of as a protector/persecutor
self-state than in terms of the oedipal theory and the struc-
tural model of id, ego, and superego? Is there a way in which
it can be understood as both, or as levels of degree? In Freud's
theory, the impetus for the boy 's identification with his father
is fear of castration . Castration anxiety develops into moral
anxiety . As Cameron and Rychlak (1985) note, "The male
conscience (I believe that to be right and wrong which father
believes) is therefore born of fear . `Conform or be castrat-
ed' is the civilizing rule" ( p . 71) . Indeed it is this fear of cas-
tration, according to Freud, that makes the boy 's superego
superior to the girl's . Freud's famously sexist statement that
"for women the level of what is ethically normal is different
from what it is in men ...[and that} ...they show less sense of
justice than men" (1925/1961c, p . 25) demonstrates his
emphasis upon the relationship between castration anxiety
and superego development . Because the boy is more high-
ly motivated to defend against his oedipal longings, his
superego is stronger, and moral strength is implicitly equat-
ed with superego strictness.
In Freud's case histories castration threats were common,
as a reflection of how severely children were punished in nine-
teenth century Europe (Miller, 1983) . Today even threat-
ening castration would be considered child abuse . Depending
upon the particular circumstances, such a threat of violent
dismemberment and deprivation of masculinity could be ter-
rifying and traumatic, potentiating dissociation rather than
healthy superego development . Consciousness of such dan-
gerous feelings might be completely incompatible with
attachment and psychic survival ; and, if dissociated, the inten-
sity of these feelings would account for the forcefulness of
the protective prohibition against their expression . In this
way harsh superego development (minus the input of the
id) may resemble dissociation . How, then, do we understand
the outcome of these dangerous and intolerable negative feel-
ings experienced by the young child toward the punitive
and/or abusive caretaker? Is the result repression, as Freud
described, or is it dissociation?
REPRESSION OR DISSOCIATION?
Although this question is important for an understand-
ing of the harsh superego, answering it is not such an easy
task . Repression is usually considered to be a higher level
defense than is dissociation . It is often described in a visual
metaphor, as involving a "horizontal split" between conscious
and unconscious (Kernberg, 1974) . Thus, an event that was
once conscious can be "forgotten " in a way that continuity
of identity is preserved . Davies and Frawley note the active
mastery aspect of repression that "creates a context for sig-
nal anxiety or symptom formation that maintains disavowed
mental contents from awareness " (1994, p . 65) . In contrast,
dissociation may be described as "vertical " splitting . DSM-IV
(American Psychiatric association, 1994) describes it as a "dis-
ruption in the usually integrated functions of consciousness,
memory, identity, or perception of the environment" (p.
477) . While dissociation is generally considered to be a
response to being overwhelmed by trauma, Young (1988)
demonstrates the complexity and utility of dissociative pro-
cesses, highlighting how "all that switches is not split " (p . 33),
and pointing out that many of those with dissociative disor-
ders are very high functioning individuals who are not
plagued by a structural split between contradictory ego-states.
He states that "The unique feature of dissociation is a pro-
232
 
DISSOCIATION, Vol, X, Nu . 9, December 1997
HOWELL
tective and inhibitory capacity of the ego to maintain con-
flict-laden material in dissociated states" (p . 35) . He points
to the prevalence of fantasy elaborations in the service of
mastery as well as their compensatory function in dissocia-
tion . As he sees it, dissociation may actually serve unity in
that it reduces over-stimulation and flooding . Likewise,
Putnam (1988, 1997) notes the protective function of dis-
crete, dissociative states of consciousness for the traumatized
child . He views mental states as "core components of con-
sciousness, behavior, and personality" (1997, p . 151) . His "dis-
crete behavioral states" (DBS) model (1997) proposes that
"a young child's behavior is organized as a set of discrete
behavioral states"(p . 20) . In the course of development, these
states become more interconnected . Trauma interferes with
the integration of states, increasing the number of discrete,
altered states . Recall the case ofJohn, in which splitting along
good/had lines, was not his primary problem . His primary
problem was an extremely harsh superego, or as this author
would conceptualize it, a dissociated self state that contained
aggression, and which almost gleefully goaded him and tor-
tured him.
FREUD : DISSOCIATION VERSUS REPRESSION?
Davies (1996) points out that Freud first used the term
"repression" to describe "psychical incompatibility and split-
ting, "which had "more in keeping with Janet's descriptions
of `traumatic dissociation ' than with his own later under-
standing of repression as a defensive manifestation within
the (early) topographical model of unconscious, precon-
scious and conscious, and the (later) structural model of id,
ego, and superego" (p . 556-557) . Erdelyi (1990) notes that
Freud never did distinguish between repression and disso-
ciation, but treated them as the same . Davies and Frawley
(1994), Davies (1996), and Bromberg (1996a) have noted
that an important consequence of Freud's abandonment of
the seduction theory was that theoretical emphasis upon dis-
sociation was replaced by repression . While amnesia and dis-
sociation were the initial key concepts proposed by Breuer
and Freud in Studies on Hysteria (1893/1955), Bromberg
(1996a) tells us that "After Studies on Hysteria, Freud was, for
the most part, openly contemptuous about the possible use-
fulness of theorizing about dissociation, hypnoid states, or
alterations in consciousness ..." (p .60).
In 1896, Freud presented his paper The Aetiology of
Hysteria, in which he articulated his view that psychoneuroses
were the result of childhood sexual trauma . Gay (1988)
reports that Freud's immediate response to the "traumatic"
chilly reception that he received to this paper was, "and this,
after one has shown them the solution of a thousands-year-
old problem, the source of the Nile? " (p . 93) . In late 1897,
Freud abandoned his seduction theory . In the intervening
time, his father had died and his self-analysis had revealed
dreams, memories, and fantasies that led him to his oedipal
theory . In September, 1897, he stated in a letter to his friend,
Fliess, "I no longer believe in my Neurotica " (quoted in Gay,
1988, p . 94) . Gay tells us that by October, 1897, "Everything
now fell into place . He recognized that his remembered
`infatuation with the mother and jealousy of the father' was
more than a private idiosyncrasy . Rather, he told Fliess, the
oedipal relationship of the child to its parents was `a gener-
al event in early childhood "' (p . 100) . In October, Freud also
wrote to Fliess that he felt that he was about to discover the
origins of morality (Cameron & Rychlak, 1985, p . 74).
Clearly, Freud's view of the child's dilemma changed sig-
nificantly by the fall of 1897 . The nature and complexity of
all the reasons for the change is a subject of much contro-
versy, including, among many matters, conjectures about
Freud's complicated feelings about his father's death, about
his mother, his need to feel important, his patients ' respons-
es to his interventions, as well as the official theoretical rea-
sons that were given (Freyd, 1996 ; Gay, 1988 ; Pines, 1987;
Ellenberger, 1970 ; Salyard, 1988 ; Kupersmid, 1993 ; Tabin,
1993 ; Davies & Frawley, 1994 ; Ulman & Brothers, 1988).
It would not be unreasonable to wonder if the above men-
tioned matters might have affected Freud's formulations
about the Oedipus complex and, consequently of the
superego construct . Interestingly, Freud focused only on
selected aspects of the Oedipus myth, leaving out themes of
patriarchy, parenthood, trust, fatherhood, and, notably,
infanticide . (Ross, 1982 ; Devereux,1953 ; Fromm, 1980 ; Pines,
1989 ; Betcher & Pollack, 1993) . In contrast to Freud's ren-
dition, the myth of Oedipus is framed in the context of his
father 's pederastic behavior . Prior to his ascension to the
throne of Thebes, Laius, Oedipus's father, had abducted and
raped the teenage son of the king Polybus, who was the ruler
of a neighboring kingdom . For this act, Polyb'us cast a curse
upon Laius such that his son would murder him and marry
his own mother . To avoid this curse, Laius left his infant
Oedipus (meaning "swollen foot") exposed with a stake
pierced through his ankles, to die . Oedipus was rescued and
brought up as the son of a neighboring king . Having heard
of the curse as applied to himself from the Delphic oracle
(Ross, 1982), Oedipus left home in order to avoid his pro-
phesied fate, but on the way got into an altercation with and
slew another traveler, who unbeknownst to him happened
to be his real father, Laius . Since Oedipus was only acci-
dentally patricidal while his father 's attempt to murder him
was deliberate, Freud 's interpretation blames the victim and
exculpates the perpetrator . (Ross, 1982 ; Pines, 1989, Betcher
& Pollack, 1994).
In contrast to the abandoned seduction theory, the new
model viewed the child as the guilty one . As Stolorow and
Atwood (1979) comment about Freud's views of infantile
development, "the sources of evil ...were located in the child
himself, in his own sexual and aggressive impulses, which
emerge according to an innate, biologically predetermined
sequence in relative independence of environmental influ-
233
DISSOCIATION. Vol. X, Ito . 4, December 1997
DESPERATELY SEEKING ATTACHMENT
ences" (p . 63) . As Fairbairn (1952) described the child's self-
blaming assumption of badness, "it is better to be a sinner
in a world ruled by God than to live in a world ruled by the
Devil" (pp . 66-67).
Perhaps fragments of the old trauma/dissociation the-
ory survived in an embedded way in the superego construct.
A crucial distinction may hinge on whether the structure
called "superego" arises from the child's need to control
uncivilized impulses, which would be moral ; or from the
child's attempt to deal with traumatic impingement from an
uncivilized world, which would have more to do with self-
preservation . Wishes, drives, and impulses are not necessarily
problematic in themselves in this model : it is the harsh pun-
ishment for them that is problematic . The harsh superego,
arising from the threat of castration linked to the oedipal
conflict, seems more about avoiding traumatic impingement.
As such it may better describe how the child becomes self-
punishing than how the child achieves mature morality.
Uncivilized wishes can be forgotten (repressed) in an over-
all sense of the continuity of personal history (Ogden, 1986;
Bromberg, 1996b), whereas overwhelming terror seems
more likely to lead to a segmentation of self-experience (dis-
sociation).
IDENTIFICATION, ATTACHMENT, AND THE
DEVELOPMENT OF THE SUPEREGO
One could describe Freud's Mourning and Melancholia
(1917/1957) as an early object relations formulation of
attachment to an abandoning and/or abusing caretaker . In
this essay some of Freud's later ideas about the role of iden-
tification in superego formation are adumbrated . Freud notes
that in contrast to normal mourning, in melancholia, the
person berates him/herself with a kind of shameless merci-
lessness . However, the castigation is usually not for the per-
son 's own qualities, but unconsciously for those of an inti-
mate other, where there has been an attachment loss "owing
to a real injury or disappointment concerned with the loved
person" (p . 249) . The melancholic has dealt with this loss
by establishing an "identification of the ego with the aban-
doned object ."
Thus the shadow of the object fell upon the ego,
so that the latter could henceforth be criticized
by the special mental faculty like an object, like
the forsaken object . In this way ....the conflict
between the ego and the loved person is trans-
formed into a cleavage between the criticizing fac-
ulty of the ego and the ego as altered by the iden-
tification . (Freud, 1917/1957, p . 249)
As a result, "in spite of the conflict with the loved per-
son, the love-relationship need not be given up " (p . 249).
As Freud later put it,
These melancholias ...show us the ego divided, fal-
len apart into two pieces, one of which rages
against the second . This second piece is the one
which has been altered by introjection and which
contains the lost object . But the piece which
behaves so cruelly is not unknown to us either . It
comprises the conscience, a critical agency with-
in the ego, which even in normal times takes up
a critical attitude toward the ego, though never
so relentlessly and so unjustifiably ...some such
agency develops in our ego which may cut itself
off from the rest of the ego and come into con-
flict with it . We have called it the "ego ideal "....
(Freud, 1921/1955, p .52)
Thus, the "critical agency " (superego, ego-ideal) which
has been differentiated out of the ego, rages against the ego,
which itself has been modified on account of identification.
And, this has occurred in order to preserve attachment . In
The Ego and the Id (1925/1961a) Freud develops these ideas
about identification further . The child replaces the inces-
tuous and parenticidal feelings that arose in response to the
Oedipus complex with identification:
The broad general outcome of the sexual phase
dominated by the Oedipus complex may, there-
fore, be taken to be the forming of a precipitate
in the ego, consisting of these two (the parents)
identifications in some wav united with each
other . This modification of the ego retains its spe-
cial position ; it confronts the other contents of
the ego as an ego ideal or super-ego.
(Freud, 1923/1961a, p . 34)
"The ego ideal is therefore the heir to the Oedipus com-
plex, and this is also the expression of the most powerful
impulses ...of the id . By setting up this ego ideal, the ego has
mastered the Oedipus complex and at the same time placed
itself in subjection to the id ." (Freud, 1923/1961a, p . 54 .)
The hostility of the superego is directly related to a person 's
need to control aggression : "the more a man controls his
aggressiveness to the exterior the more severe - that is aggres-
sive - he becomes in his ego ideal . . . It is like a displace-
ment, a turning round upon his own ego ." (Freud, 1923/
1961a, p .54)
THE CONSTRUCT OF THE OEDIPUS COMPLEX
In Freud's view, superego was heir to the Oedipus com-
plex . In analyzing the origins of superego, then, one might
ask, then, whether oedipal desires need to be repressed, dis-
solved, demolished, or smashed, for psychic and moral
health, or is it enough for them to be benignly acknowledged?
Although most of the psychoanalytic literature empha-
234
 
DISSOCIATION, 1'ol, X, No . 4, December 1997
HOWELL
sizes the role of aggression in superego formation (Brenner,
1982), some of the writings on the development of healthy,
mature superego notably emphasize love and idealization.
Schafer (1960) addresses the importance of the "loving
superego" derived from loving, positive identifications of the
preoedipal period . Brenner (1982) emphasizes that superego
formation involves an alliance with, willing submission to,
and desire to emulate the beloved parent(s) . While acknow-
ledging the importance of such defensive maneuvers such
as identification with the aggressor, inhibition of competi-
tiveness, reaction formation, and the like, Brenner states that
the "common thread" that runs through them is "the belief
that it is of vital importance to be approved of and loved by
one's parents" (p . 125) . Observing that excessive guilt does
not signify a healthy superego, Schafer, (1960) and Lewis
(1981, 1983) note that Freud 's understanding of normal guilt
was drawn from depressive, obsessive, and melancholic mod-
els and was not representative of higher levels of ego and
superego functioning . Further, Schafer, (1968) notes that
as they mature, superego identifications become more
autonomous and integrated . Thus, while identifications,
including hostile ones, may have formed the kernel of the
superego, the end product may be a moral code that has
become largely autonomous.
In consonance with these ideas, Davies and Frawley
(1994) object to the idea that:
it is primarily superego injunction and fear of cas-
tration or loss of love that brings about the
destruction of the complex . Although this may
certainly explain the fate of events in families
where jealous impingements and boundary trans-
gressions have been the hallmark of the Oedipus
situation, it is unlikely to represent the normal
course of events in families where the young
child's emergent sexuality has been welcomed
and enjoyed   Although superego may become
heir to an Oedipus complex marked by jealousy,
guilt, impulsivity, and boundary transgression, a
more positive experience and result lead from
...healthy identifications with two parents com-
fortably at peace with their own sexualities.
(Davies & Frawley, 1994, pp .231-232)
They then ask whether "Freud 's theories (might} be subject
here, too, to the particular coloration given them by his early
work with adult survivors?" (p . 232) . Similarly, Kohut (1984)
states that castration anxiety, although not infrequent, is a
pathological symptom of a disorder of the self . He states, "a
boy who is exposed to the responses of psychologically healthy
parents does not experience a significant degree of castra-
tion anxiety during the oedipal phase ...[and further] ... the
healthy child of healthy parents enters the oedipal phase joy-
fully" (Kohut, 1984, p .14) .
Sagan (1988) objects to Freud's statements that in boys,
the Oedipus complex is "literally smashed to pieces " by the
threat of castration (p . 73) : "Thus is our morality, mankind's
higher nature, born in the environment of the penitentiary.
Freud does not tell us what happens to the Oedipus com-
plex in households where the parents never threaten cas-
tration . Such a situation is clearly unimaginable for him"
(Sagan, 1988, p . 75) . In accordance with Davies and Frawley,
Sagan suggests that Freud has given us "an accurate and pen-
etrating description of the pathological ending of the
Oedipus complex" (1988, p .78).
Noting the importance for the child's psychological
health to not be overwhelmed by threats of counter-aggres-
sion for oedipal desires, Sagan states:
The greater the actual or imagined threat of cas-
tration, the less the child will be inclined to risk
- imaginatively- the accomplishment of oedipal
goals, the more the child will retreat from the
Oedipus complex, the less, the superego will be
internalized and the more it will speak in an exter-
nal, authoritarian voice . The less the actual or
imagined threat . of castration, the more the child
will be willing to risk the accomplishment of oedi-
pal inclinations, the more the superego will be
incorporated and speak with an internal voice.
(Sagan, 1988, p . 81)
Perhaps Sagan presages the notion that when parents are
severely punitive, the superego becomes dissociated, and
seems external to the self, rather than becoming integrated
into the self.
HARSH, PATHOLOGICAL, AND CORRUPT
SUPEREGO
There are further problems with the notion of the need
for the impetus of castration fears and the resulting "exter-
nal" superego as a developmental necessity . Does this mean
that the relatively unterrified, securely attached child will
develop no moral code? Does the often assumed positive cor-
respondence between superego strictness and moral strength
(discussed in Lewis, 1983, p . 187) mean that acquisition of
a moral code is synonymous with self-punishment? Research
on moral development suggests otherwise . In his literature
review Kohlberg (1963) concluded that punitive discipline
was an antecedent condition of delinquency . In her discus-
sion of the literature, Lewis (1983) states "Other studies have
shown that the fathers of men with strong moral standards
have not been strict with their sons ; on the contrary, men
with weak moral standards have had very punishing fathers "
(p . 187) . As noted earlier, in contrast to the problems of harsh
superego, a healthy, more mature superego is well-integrat-
ed with and often indistinguishable from the ego . As they
235
DISSOCIATION, Vol . X. No. 4, December 1997
I
mature, superego identifications become more autonomous
and integrated so that the end product may be a moral code
that has become largely autonomous (Schafer, 1968).
Superego, conceptualized as a static, unmodified in tro-
jection of parental superegos will not necessarily grow and
change . It will be relativistic (Sagan, 1988 ; Kohlberg,19 71) ,
reflecting the familial and local mores, which have been inter-
nalized as normative standards . Because a variety of condi-
tions will differ for cultures, so will their "rules" (Wilson.
1993) . This cultural relativism, which holds that moral stan-
dards vary with cultures, is often confused with ethical rela-
tivism which adds to cultural relativism the view that these
differences are rationally irreconcilable, that each culture 's
view of "right" is right (Kohlberg, 1971) . The fact that infan-
ticide or clitorectomy might be considered moral by the cul-
tures that practice them becomes interpreted by ethical rel-
ativism as an injunction not to judge . Because the theory of
moral internalization via the Oedipus complex invol v
nalization of parental and cultural mores, it is (ethically) rel-
ativistic : morality is defined according to these rules and
mores, rather than being understood as based on universal
guiding principles or ethics which have been formulated by
the morally maturing individual over time . As Sagan (1988)
observes, some of the worst atrocities in history, such as the
Holocaust and the Inquisition, have been committed in the
name of superego and/or of morality . In such cases the
superego has become "corrupt" : the executors of the hor-
rors fully believed that they were doing "the right thing ." Of
course, corruption or maturity in superego function does
not have to be across the board : people can have moral laps-
es in accordance with their own experiences in culture and
families.
A harsh, pathological, even "corrupt "superego can com-
mand conformity to the standard mores of the culture . It
can become a kind of internal slave-driver, forcing confor-
mity to society's rules, in this way perpetuating relativistic
morality . Obviously, this is a far cry from mature morality.
Perhaps it is this relativistic, corrupt, often harsh morality
that makes which more sense via the dissociation model.
Indeed, the perforrnance of atrocities maybe considered to
require dissociation (Lifton & Markusen,1990) . Certainly,
dissociated experience is less likely to be integrated into a
more autonomous moral code as is characteristic of mature
superego . It is more likely to remain like a child 's superego,
which may view the avoidance of punishment as morality
(Brenner, 1982) . Remaining in such a state, it is more vul-
nerable to be corrupted . Perhaps it is the harsh, severe, patho-
logical superego that "speaks in an external voice ."
The question then arises as to whether "superego " is too
inclusive a category, that is, whether or not the harsh, patho-
logical superego belongs in the same category as mature
superego . If the harsh superego refers to a dissociated self=
state, then its integration with other aspects of self will be
more difficult, thereby inhibiting the development of an
autonomous moral code .
DISSOCIATION, HARSH SUPEREGO, AND MORALITY
Recall the earlier observation of similarity of the harsh
superego to a dissociated protector /persecutor self-state that
has the function of safeguarding attachment . While the harsh
pathological superego stands as morality because it is based
on attachment (it is more than an internal signaling device
for "realistic" social behavior based on adherence to rein-
forcements, punishments, anticipated dangers, etc .), it is not
mature morality . However, it is a child 's morality that was
stunted in its growth, and that, as a consequence of dissoci-
ation, may have become and/or remained corrupt . It does
represent internalized rules and values, but because this self-
state is dissociated, these values are relatively unavailable to
improvement, growth, and integration by conscious intro-
spection.
Currently, the accumulating psychological and psy-
chobiological evidence supports the preoedipal, even pre-
verbal, emergence of superego functions . According to
Schore (1997), "Moral development thus begins in prever-
bal periods of infancy, earlier than generally thought . In fact,
the practicing phase, from the end of the first through the
middle of the second year, is a critical period in the early
development of the superego " (p . 352) . The most viable
model of the superego now seems to be that of an affect-reg-
ulator . Shame and guilt can both be considered to be
superego affects (Lewis, 1981, 1983 ; Schore, 1997) . However,
shame, which emerges earlier in the child's development,
has adaptive, psychobiological aspects which regulate attach-
ment . Shame is involved in both the deactivation and reac-
tivation of attachment and in the switches of psychobiolog-
ical states (Schore, 1997).
Wilson (1983) poses an evolutionary hypothesis that it
is not the "rules" themselves, but a biological predisposition
to attachment that is adaptively moral . More specifically,
Lewis (1983) posited that shame can effectively modulate a
child's behavior- so as to produce the desired cultural mores.
Ideally, the experience of shame itself should be within tol-
erable bounds . Prosocial and empathic behavior depends
upon tolerated, experienced shame (Schore, 1997):
Like the mature superego which has its roots in attach-
ment, the harsh superego (understood as a dissociated self-
state which originated to protect attachment) has a moral
basis . According to Lewis, "Human beings are social by bio-
logical nature and ...shame and guilt are `givens ' whose func-
tion it is to maintain the basic affectional bonds . When these
bonds are threatened, shame and guilt work overtime to pre-
serve them through the formation of primary-process neu-
rotic symptoms at the expense of the self' (1983, p . 227).
Further, "Morality is the affective-cognitive outcome of
attachment . Threatened attachment, which first evokes
protest aimed at the caretaker - 'other,' is then trans-
formed, mainly by identification into states of shame and guilt
that aim at maintaining the attachment" (1983, p . 173).
Understood through this attachment model, mature moral-
236
 
DISSOCLITION . No
HOWELL
ity would also be an outgrowth of attachment, but it does
not have to weaken if the harsh superego softens . Indeed,
in this model, the softening of the harsh superego, or the
lessening of the dissociation that it involves, would foster the
development of a mature, strong, and integrated morality
as opposed to a harsh and dissociated morality.
IDENTIFICATION WITH THE AGGRESSOR
Most of the psychoanalytic literature has given aggres-
sion the primary role in superego formation during the oedi-
pal period (Brenner, 1982) . Much of this hinges on the con-
cept of identification with the parental superego(s), in
essence, something like identification with the aggressor
(Cameron & Rychlak, 1985) . Anna Freud (1966) described
identification with the aggressor as a pre-stage in superego
development in which the child identifies with the power of
the aggressor as a way of avoiding being overwhelmed by ter-
ror . Identification with the aggressor, then, is a pivotal con-
cept used both in the psychoanalytic literature (describing
superego formation) and in the dissociation literature
(describing the creation of a dissociated identity state which
holds the aggression and performs aggressive functions).
Blizard (1997a) integrates psychoanalytic object relations and
dissociation concepts, describing how this internalization
enables the severely traumatized child to disown the repre-
sentation of self as powerless and weak, while at the same
time maintaining attachment to the abusive caretaker.
How does "the object" become "part of" the individu-
al? How does the object's behavior come to be replicated by
the individual? Such questions are especially relevant to
understanding hostile introjects, internal abuser alters, as well
as the development of a mature superego . Pathological, harsh
superego may be understood to involve identification with
the aggressor in the context of attachment . This identifica-
tion protects attachment to the abusive caretaker (Blizard,
1997b ; Blizard & Bluhm, 1994 ; Goodman & Peters, 1995;
Howell, 1996 ; MacGregor, 1996) . Because this dissociated
structure which acts as superego feels external to the self, it
sets the stage for various disorders . In DID it is personified.
In melancholia and paranoia the protector/persecutor/
superego feels external and is projected . In OCD the com-
pulsive harsh demands seem to come from "out there " (Ross,
1989) . In masochism and often in depression the dissociat-
ed protector/persecutor mediates self-criticism . (Howell,
1996a) . This formulation is similar to Freud's in Mourning
and Melancholia (1917/1957), with the difference that this
formulation is explicitly based on the concept of dissocia-
tion.
TREATMENT
If the harsh superego is understood as a dissociation-
based structure, then dissociation needs to be addressed in
its treatment . Comments which are intended to soften the
harsh superego, such as, "You are being hard on yourself "
or "You have turned your aggression inward" may inaccu-
rately address self-experience as a unity . The very fact that
self experience is not a unity is often an important thing that
the patient needs to notice . Comments such as the above
examples are actually addressing the wrong self-state . The
normally conscious, abused self-state is likely to regard such
statements as either non-sensical or as accusatory, since it
experiences little or no power to do anything about the sit-
uation . It is the rageful part that needs to he addressed and
that needs to be listening.
This dissociated structure, this pathologically harsh
superego, arose out of the need to preserve attachment . As
such, it keeps alive the person's hope for and illusion of
attachment to the original parent figures and others who
could be imagined to fulfill that function . As a depository
for isolated affect, it also defends against shame and terror.
However, the perils of childhood being past, this personal-
ity organization is maladaptive because the harsh superego
maintains unrealistically high standards in anticipation of
punishment from others who may not be as punitive as the
parents.
The patient's recognition of the attachment-related ori-
gins of the internal persecutory structure can foster re-asso-
ciation of self and superego . This recognition may also trig-
ger separation anxiety and depression (Goodman & Peters,
1995 ; Blizard,1997b ; Blizard & Bluhm, 1994 ; Watkins &
Watkins,1997) . The persecutory, cruel, even antisocial aspects
of the harsh superego can be recognized as aspects of the
self, both needed in the past and painful to possess in the
present . While attachment to the abuser may be initially nec-
essary and adaptive, it is probably undeserved . The disso-
ciative process needs to be reversed (Watkins & Watkins,
1988), and the dissociative split needs to he healed so that
the harsh superego can be appreciated for its protective, truly
moral function, but also exposed as it is - a cruel and imma-
ture container of otherwise intolerable affect . Part of the
patient 's problem may be that rigid, conventional morality
may be replicated in a dissociated structure, which then
begets and encourages self-harshness with an internalized
"pat on the back ." Interpretation which focuses on the hid-
den effort to maintain traumatic attachment may be crucial.
SUMMARY
This article recasts the harsh superego in terms of attach-
ment and dissociation . It is the author 's contention that clin-
icians frequently work with the harsh superego this way intu-
itively, even though they may not conceptualize it as such.
This new model of the harsh superego requires changes in
language, in hypothetical constructs, and in explicit moral
values . Encasing the destructiveness and pain of dissociated
abuse inside the concept of morality trivializes both .
237
DISSOCIATION, Vol, X, No .1 . December 1997
DESPERATELY SEEKMN
  o
1
This article focuses on the psychoanalytic approach to
the harsh superego through the lens of dissociation . As we
look through this lens, the contents of the picture of the harsh
superego may be much the same as what classical psycho-
analytic theory has shown us, but the way we perceive them
is different . a73
REFERENCES
American Psychiatric Association . (1994) . Diagnostic and statistical
manual of mental disorders(4th ed .) . Washington, DC : Author.
Beahrs,J . (1982) . Unity and multiplicity. New York : Brunner/hazel.
Betc her, W ., & Pollack, W . (1993) In a timeoffallen heroesNew York:
Guilford Press.
Blizard, R . (1997) . The origins of dissociative identity disorder from
an object relations theory and attachment theory perspective.
DISSOCIATION, 10, 224-230.
Blizard, R . (1997) . Therapeutic alliance with abuser alters in dis-
sociative identity disorder : The paradox of attachment to the abus-
er . DISSOCIATION, 247-255.
Blizard, R ., & Bluhm, A . (1994) . Attachment to the abuser:
Integrating object-relations and trauma theories in treatment of
abuse survivors . Psychotherapy, 31,
Bowlby, J . (1969) . Attachment and loss . I, Attachment New York:
Basic Books.
Bowlby, J . (1973) . Attachment and loss. Vol. II, Separation. New York:
Basic Books.
Bowlby, J . (1984) . Psychoanalysis as a natural science, Psychoanalytic
Psychology, 1, 7-22.
Bowlby, J . (1988) . A secure base New York : Basic Books.
Brenner, C . (1982) . The mind in conflict : International
Universities Press.
Breuer, J ., & Freud, S . (1955) . Studieson hysteria. InJ . Strachey (Ed.
and Trans .), The standard edition of the complete psychological
works of Sigmund Freud (Vol . 2) . London : Hogarth Press . (Original
work published in 1893-1895)
Bromberg, P . (1996a) . Hysteria, dissociation, and cure : Emmy von
N ., revisited Psychoanalytic Dialogues, 6,55-71.
Bromberg, P . (1996b) . Standing in the spaces : The multiplicity of
self and the psychoanalytic relationship . Contemporary Psychoanalysis,
32, 509-536.
Cameron, N ., & Rychlak, J .F. (1985) . Personality development and
psychopathology . Boston : Houghton Mifflin
Company.
Davies,J .M. (1996) . Linking the "pre-analytic" with the post-classi-
cal : Integration, dissociation, and the multiplicity of unconscious
process . Contemporary Psychoanalysis, 32,553-576 .
Davies, J .M., & Frawley, M .G . (1994) . Treating the adult survivor of
childhood sexual abuse New York : Basic Books.
Devereux, G . (1953) . Why Oedipus killed Laius - a note on the
complementary Oedipus complex in Greek drama . International
journal of Psychoanalysis, 34,132
Ellenberger, II . (1970) . The discovers of the unconscious. New York:
Basic Books.
Erdelyi, M . (1990) . Repression, reconstruction, and defense:
History and integration of the psychoanalytic and experimental
frameworks . In J .L. Singer (Ed .), Repression and dissociation(pp . 1-
31) . Chicago : University of Chicago Press.
Fairbairn, W .R. (1952) . Psychoanalytic studies o/Ihepersonality. Boston:
Routeleclge and Kegan Paul,
Freud, A . (1966) . The ego and the mechanisms of defenc.se(Rev . ed) . New
York : International Universities Press . (Original work published in
1936)
Freud, S . (1955) . Group psychology and the analysis of the ego . In
J . Strachey (Ed . and Trans .), The standard edition of the complete psy-
chological works of Sigmund Freud(Vol . 18, pp . 69-143) . London:
Hogarth Press . (Original work published in 1921)
Freud, S . (1957) . Mourning and melancholia . In J . Strachey (Ed.
and Trans .), The standard edition of the complete psychologicalworks of
Sigmund Freud (Vol . 14, pp . 243-258) . London : Hogarth Press.
(Original work published in 1917)
Freud, S . (1961a) . The ego and the id . In J . Strachey (Ed . and
Trans .) , The standard edition of the complete psychological works of Sigmund
Freud (Vol . 19, pp . 12-66) . London : Hogarth Press . (Original work
published in 1923)
Freud, S . (1961b) . The dissolution of the Oedipus complex . InJ.
Strachey (Ed . and Trans .), The standard edition of the complete psy-
chologicalworks of Sigmund Freud (Vol . 19, pp . 173-182) . London:
Hogarth Press . (Original work published in 1924) .
Freud, S . (1961c) . Some psychical consequences of the anatomi-
cal distinction between the sexes . In J . Strachey (Ed . and Trans .),
The standard edition of the complete psychological worksSigmund Freud
(Vol . 19, pp . 248-260) . London : Hogarth Press . (Original work pub-
lished in 1925)
Freud, S . (1962) . The aetiology of hysteria . 1n J. Strachey (Ed . and
Trans .), The standard edition of the complete psychological worksof Sigmund
Freud (Vol . 3,pp . 191-221) . London : Hogarth Press . (Original work
published in 1896)
Freyd, J . (1996) . Betrayal trauma . Cambridge, Massachusetts:
Harvard University Press.
Fromm, E . (1980) . Greatness and limitationsof Freud's thought. New
York : New American Library.
Gay, P . (1988) . Freud: A life for our time. New York : Norton.
Goodman, L ., & Peters, J . (1995) . Persecutory alters and ego states:
Protectors, friends, and allies . DISSOCIATION, 8(2), 91-99.
238
 
DISSOCIATION, Vol, X,1No. 4, December 1997
HOWELL
Grusec, J .,& Goodnow,J . (1994) . Impact of parental discipline meth-
ods on the child's internalization of values : A reconceptualization
of current points of view . Developmental Psychology, 30,
Greenberg, J ., & Mitchell, S . (1983) . Object relations in psychoanalyt-
ic theoryCambridge : Harvard University Press.
Hoffman, M . (1970) . Moral development . In P . Mussen (Ed .),
Carmichael's manuelof child psychology : Wiley.
Howell, E . (1996) . Dissociation in masochism and psychopathic
sadism .Contemporary Psychoanalysis, 32,
Howell, E . (1996, November) . Masochism: A bridge to the other side
abuse. Paper presented at the annual conference of the Interna-
tional Society for the Study of Dissociation, San Francisco.
Howell, E . (1997) . Masochism : A bridge to the other side of abuse.
DISSOCIATION, 10(4), 241-246.
Kohlberg, L . (1963) . Moral development and identification . In H.
Stevenson (Ed .) Child psychology
for the Study of Education : University of Chicago Press.
Kohlberg, L . (1971) . From is to ought : How to commit the natu-
ralistic fallacy and get away with it in the study of moral develop-
ment . In T . Mischel (Ed .) , Cognitive development and epistemology
York : Academic Press.
Kupersmid, J . (1993) . Freud's rationale for abandoning the seduc-
tion theory . Psychoanalytic Psychology, 10,275-290.
Kohut, H . (1984) . How does analysis cure? Chicago : University of
Chicago Press.
Lewis, H . (1971) .Shame and guilt in neurosis : International
Universities Press.
Lewis, H . (1981) . Freud and modern psychology (Vol : The emotional
basis of mental illnessNew York : Plenum Press.
Lewis, H . (1983) . Freud and modern psychology (Vol : The emotional
basis of human behavior New York : Plenum Press.
Lewis, H . (1990) . Shame, repression, field dependence, and psy-
chopathology . In J . Singer, (Ed .), Repression and dissociation,233-
258 . Chicago : University of Chicago Press.
Lifton, R ., & Markusen, E . (1990) . The genocidal mentality
Basic Books.
MacGregor, W . (1996) . Multiple personality disorder : Etiology,
treatment, and treatment techniques from a psychoanalytic per-
spective . Psychoanalytic Psychology, 13,
Ogden, T . (1986) . Matrix of the mindNorthvale, NJ : Jason Aronson.
Pines, M . (1989) . On history and psychoanalysis . Psychoanalytic
Psychology, 6,
Putnam, F . (1988) . The switch process in multiple personality dis-
order and other state-change disorders .DISSOCIATION, 1(1) , 24-32.
Putnam, F. (1989) . Diagnosis and treatmentof multiple personality dis-
order. New York : Guilford .
Putnam, F . (1997) . Dissociation in children and adolescents
Guilford.
Ross, C . (1989) . Multiple personality disorderNew York : John Wiley
and Sons.
Ross,J . (1982) . Oedipus revisited - Laius and the "Laius complex ."
Psychoanalytic Study of the Child, 37,
Sagan, E . (1988) . Freud, women and morality: The psychology of good
and evil. New York : Basic Books.
Salyard, A . (1988) . Freud as Pegasus yoked to the plough . Pycho-
analytic Psychology, 5,
Schafer, R . (1960) . The loving and beloved superego in Freud's
structural theory . Psychoanalytic Study of the Child, 15,
Schafer, R . (1968) .Aspects of internalization : International
Universities Press.
Schore, A . (1997) . Affect regulation and the origin of the self
Lawrence Erlbaum Publishers.
Spiegel, D . (1990) . Trauma, dissociation, and hypnosis . In R .P. Kluft
(Ed .), Incest-related syndromes of adult psychopathology (pp247-261).
Washington, DC : American Psychiatric Press.
Stolorow, R .D., & Atwood, G .E. (1979) .Faces in a cloud: Subjectivity
in personality theoryNew York : Jason Aronson.
Tabin, J . (1993) . Freud's shift from the seduction theory : Some over-
looked reality factors . Psychoanalytic Psychology 291-298.
Ulman, R ., & Brothers, D . (1988) . The shattered self A psychoanalyt-
ic study of trauma.. Hillsdale, NJ : Analytic Press.
Watkins, J .G., & Watkins, H .H . (1997) . Ego states
py. New York : Norton.
Watkins, J .C., & Watkins, H .H . (1988) . The management of malev-
olent ego states in multiple personality disorder .DISSOCIATION, 1(1),
67-72.
Wilson, J .Q . (1993) . The moral senseNew York : The Free Press
Young, W .C. (1988) . Psychodynamics and dissociation . DISSOCIA-
TION, 1, 33-38 .
239
DISSOCIATION, Vol . 4