THREE STRANDS OF NARCISSISM

Vann Spruiell, M.D.

This paper was published in The PSYCHOANALYTIC QUARTERLY, 1975, Vol. 44: 577-595. It is an effort to account for the development of the transformations of narcissism – seen here as constituted by three ‘strands’ of development. They begin apparently separately, but eventually as functions become amalgamated if maturation and development go well within the mother-infant field, and later, through the subsequent stages of individuation. Then they are hard to separate and distinguish. But they may remain relatively separate if development has been blighted, as it has been in so-called ‘borderline patients’ or in so-called ‘narcissistic characters’. I stand by the ideas of twenty-three years ago, but have made minor textual alterations designed to increase clarity.

The paper is related to an earlier paper (vs84b), not published before, called, "Thinking Blind," although it was presented on various occasions, e.g., as the Fourth Annual Maurice Friend Lecture, in New York, 1984. Unlike the present paper, it is presented almost entirely in phenomenal, experiential, subjective terms – in ‘experience near’ terms rather than in the more abstract forms used in the present paper. "Thinking Blind" is more accessible to the reader than ‘Three Strands of Narcissism.’ However, in my view, neither the more abstract paper nor the more experiential paper is sufficient by itself.

The paper is also related to subsequent ones, especially "Narcissistic Transformations in Adolescence" (vs75b), "Alterations of the Ego Ideal in Girls in Mid-adolescence" (vs79d), "Freud’s Conceptions of Idealization" (vs79b), and "Screens, Splits, Frames and Keys: The Analysis of an Omnipotent Man" (vs91c).

INTRODUCTION

The prediction that the clinical and theoretical understanding of narcissism would open new territories for psychoanalysis is finally beginning to come true – more than eighty-five years later. In the last few decades, there have been efforts to unravel the theory and to reach some clear ideas of just what is intended by the word, ‘narcissism’, as a noun, and ‘narcissistic’ as an adjective. But even now the topic continues to be bedeviled by ambiguities and unnecessarily new metapsychological constructs.

In this paper, I temporarily back away from highly abstract metapsychological theories pertaining to economics and postulated intrapsychic structures, although I believe that conceptions of structures are necessary,¹ I consider the development of narcissism more from the point of view of the drive regulating apparatuses – the ego and the superego. Clinical experience and particularly observational studies of children make possible a genetic view of narcissism as changing and transforming sets of human motivations, as well as taking into account views of the types of individuals appropriately labeled ‘narcissistic personalities’. From such a perspective, we may be able to visualize a psychoanalytic situation, and the techniques within that situation, which can take into account apparently incompatible observations – for example those of Kohut and his followers, and Kernberg and his followers² – and others too. It should be made clear, however, that rather than seek some sort of compromise of differing views, my purpose is to suggest a way to utilize theories and data already available to us.

As a beginning, we should recognize that the term, ‘narcissism’ has not been used as a single entity, a singular category. Rather, in the literature it is used to label a variety of developing phenomena encompassed by a broad and varying theoretical formulations. These have never met with universal acceptance: for example, Freud’s construction regarding libido (in which aggression has never been satisfactorily included) has not been endorsed by all analysts. Libido, according to Freud’s construction, is variably distributed between the ego and its objects. He believed that out of narcissistic libido there gradually develops first homosexual and then heterosexual object libido. An adult-type distribution is not acquired, according to him, before adolescence. At any time there is a possibility of a reversed flow – a withdrawal of investments (‘cathexes’) from objects and their reinvestment in the self. Secondary narcissism, Freud thought, can also arise by way of identification with objects (that remain invested by libido). Certain identifications also play a role in structuring the superego which functions in part as secondary narcissistic structure.

Beneath this useful ‘umbrella’ formulation – one which, however, may be difficult if not impossible to apply to specific clinical events³, narcissism was described by Freud (1914) in the following forms: a) as a phase; b) as a perversion; c) as a major part of the regulation of feelings of well-being – methods for maintaining an inner state of orderly functioning, balance, and integration signaled by apparent feelings of well-being in the young child and by evidences of what we call self-esteem in later life; d) as an aspect of self-love; e) as a type of object choice; f) as omnipotence; and finally, g) in terms of more individuated forms, as secondary structures – in particular the ego ideal aspects of the superego.

Obviously, some of these refer to one logical category, some to others. A phase is one sort of thing, self-love is something else, and the ego ideal is something still different. Yet all are in one way or another related to each other.

Three of these forms – self-love, the regulation of self-esteem, and omnipotence – should be treated as separate but related variables. Perhaps they can be thought of as independent ‘developmental lines’ (cf., A Freud, 1963), which represent coherent themes of ego motivations, and which may become integrated and mingled in the healthy individual, but which may be kept apart pathologically, or individually stunted, or compenstorily intensified, swollen, and overused.

In the section to follow, these three motivational sets of developing narcissism (which I call ‘strands’ in spite of the potential misunderstanding that reifications are implied) will be separated artificially for heuristic purposes. The period before self and object representations are reliably differentiated (roughly, the first sixteen to eighteen months) is considered first. In the second section the evolution of object relations extending into the remaining separation-individuation proce