FREUD, THEORIST OF COMPLEXITY
Raymond E. Fancher
When I was invited to participate in this symposium, my first impulse was to decline. In recent years my primary research interests have shifted away from Freud and psychoanalysis, and I did not think I could find anything new or worthwhile to say. Just before conveying my official regrets, however, I happened to be reading Mitchell Waldrops Complexity (1992), an engaging popularized account of so-called "complexity theory." The subject was new and interesting to me, and further, some of the things Waldrup said triggered associations to Freud. When I proposed to discuss some of these for the symposium, the moderator agreed to let me give it a try. So here I am.
Some of my associations were triggered by apparent similarities between the early work of Freud, and that of some of the pioneers of complexity theory - similarities both in the way they theorized and in the general content of their conclusions. Thus Waldrup describes the youthful efforts of Stuart Kauffman, who as a student in the 1960s conceptualized the body's regulatory genes as complex networks of interconnected on-off switches. Kauffman became preoccupied by the question of what would be the typical behaviors of systems with relatively "sparse" versus relatively "dense" interconnections - that is, whether the on or off state of each particular switch was determined by the states of just one or two of its neighbors, or by large numbers of them. Lacking a computer to model the unlimited number of possibilities, Kauffman let his formal studies drop and obsessively filled his notebooks instead with "little diagrams of his regulatory genes hooked up in [various] wiring diagrams,¼ trying to understand how they turned each other on and off" (Waldrup, 1992, p.109).
Waldrup also described the early experiments by John Conway and his colleagues with the so-called "Game of Life." On a grid of squares capable of being colored black or white, they investigated the patterns and sequences that emerged when each square's color was determined by the colors of its eight immediate neighbors, under differing sets of rules. For example, each square might become or remain black only when six of its immediate neighbors were black, or four, or any other arbitrary set of rules. Starting out without computers like Kauffman, these investigators too had to conduct their earliest experiments "by hand," in an obsession-like process that came to have an almost hypnotic fascination.
In both of these cases similar findings emerged - findings that were first suggested by the "hand" experiments and later confirmed by much more exhaustive computer modelling. When the rules governing the sequence of patterns were either very simple or very complex - as when Kauffman's interconnections were either very sparse or very dense, or where the squares' colors in the Game of Life were determined by relatively many or relatively few of their neighbors, then the overall patterns tended quickly either to "lock up" into rigidly fixed entities, or to fluctuate chaotically. With rules of intermediate complexity, however, some fascinating and initially unpredictable consequences emerged. In the Game of Life, when each square became or remained black only if exactly two or three of its neighbors was black - then intriguing patterns of black and white groupings would develop, including regular forms that would "glide" indefinitely across the grid, "eating" individual black squares that got in their way. Parallel developments occurred in Kauffman's gene networks when each unit was interconnected with just two or three others.
Waldrup's accounts of these developments reminded me of Freud during what has always been my favorite phase of his career, the period in the mid-1890s when he attempted to integrate his neurological background with his newer observations about hysteria and dynamic psychology, in his speculative and long-unpublished Project for a Scientific Psychology. In this first systematic attempt to construct a comprehensive model of the mind, Freud calculated the consequences of a purely mechanical exchange of varying amounts of excitation or "cathexis" within a network of interconnected neurons representing specific ideas or memories, which he called the "Psi system."
Like the young complexity theorists, Freud became obsessed with the workings of his network. He wrote to his friend Fliess: "I am in a bad way; I am so deep in the [project] that it quite consumes me.... I have never been so intensely preoccupied by anything.... The hours of the night from eleven to two have been occupied with imaginings, transpositions, and guesses." Freud's spirits rose or fell depending on how the model was working. At one point he moaned, "after an excess of mental torment, I just apathetically tell myself that it does not hang together yet and perhaps never will." At another moment, however, "Everything fell into place, the cogs meshed, the thing really seemed to be a machine which in a moment would run of itself.... I can naturally hardly contain myself with delight" (Freud, 1954, pp. 118, 120, 126, 129). Here is a clear picture of a theorist fully immersed in, and at times consumed by, his theorizing. And with his diagrams of interconnected neurons, and his calculations of the consequences of differing patterns of simple rules for their interaction, his manner of working strikes me as quite similar to that of the youthful complexity theorists. (I might add parenthetically here, that I have no sense of Freud during this period as trying to fudge his data or to mislead his audience. His intense intellectual involvement in his task strikes me as completely genuine.)
Some of Freud's conclusions also resembled the complexity theorists'. Today I lack the time to describe these in detail, but I will try to illustrate with the example of his concept of the ego as presented in the Project. A neurological version of this concept was already familiar to Freud from the work of his teacher Theodore Meynert (1833-1892). Meynert taught that individual cells in the brain's cortex represent specific memories and ideas, and become interconnected with each other via "association fibres" whenever two or more of them become simultaneously excited. Subsequently, cells can exchange portions of their excitation with each other, in the presumed neurological underpinning of the association of ideas by contiguity. Meynert referred to the totality of each person's acquired cortical connections as the "ego" (German Ich). Meynert further taught that egos can vary in their overall "strength" or organizing power. An adult's ego is more developed and thus stronger than a child's, for example, but in the state of sleep the associations of even an adult ego become temporarily weakened. And in some conditions of insanity, the weakening may be more or less permanent. In all of these conditions, the consequence of a weak ego is randomness and disorganization of the thought processes - the infant's "big buzzing blooming confusion" (as characterized by William James), or the bizarre illogicalities of dreams and hallucinations. In essence, Meynert's ego was a neurological representation of a simple associationistic psychology, such as that espoused by James Mill (see Amacher, 1964).
But Freud had learned from his other teacher Franz Brentano that simple associationistic psychologies could not easily account for the motivated nature of thought - for example, for the fact that associations to ideas about food go in very different directions depending on whether one is feeling nauseated or hungry (Fancher, 1977). And further, Freud had learned at first hand from his budding clinical practice, about the power of dis-associated ideas and unconscious motivation. Accordingly, he attempted in the Project to expand upon Meynert's model of the ego, making it a genuinely dynamic agency. He did so by adding a deceptively simple condition to the system.
First, he accepted Meynert's basic premise that specific ideas are represented by specific neurons in a cortical network - his Psi System. At birth, every possible pair of these neurons is separated by a "contact barrier" that impedes the transmission of excitation, or what Freud called "cathexis," from one to the other. But after two neurons have been simultaneously cathected, the resistance at their contact barrier becomes permanently somewhat reduced, thus facilitating their association in the future. So far, this is simply a restatement of Meynert's model, using slightly different language. Freud's "dynamic" addition was the further provision that whenever a particular idea-representing neuron is in a state of partial cathexis, there follows an additional but temporary reduction in the resistance at all of its contact barriers. In psychological terms, partially cathected or aroused ideas become temporarily "primed" to become actively involved in all ongoing trains of association, regardless of prior experience. When one is very hungry, most of one's thoughts and associations will tend to run in the direction of food.
Freud defined his ego as the totality of the Psi System at any given moment, comprising both its permanent, experientially acquired associations, and its ever-changing overall pattern and distribution of cathexes. Thus when the overall level of cathexes in the ego is relatively low, the overall direction of thought tends to be dominated by prior associations, and the system operates mechanically in basically the same way as Meynert's. When the overall excitation is very high, then relatively many of the Psi System's neurons become excited close to their capacity, and have a great tendency to "discharge" their excitation into the musculature in a chaotic pattern, creating much disorganized activity. When the levels of cathexes in the ego are moderate, however, all sorts of interesting things can happen. If they are relatively evenly distributed, for example, then relatively many ideas become primed and ready for involvement in the thought process - a prerequisite for adaptive, rational thought. But if the cathexes are unevenly distributed and concentrated in small and isolated clusters of neurons, then discharges due to overload will be especially frequent within just those subsystems, and will lead to phenomena such as hysterical conversion symptoms, behavioral "slips," or dreams with a strong wish-fulfilling content. Indeed, Freud deduced a general manner of what he called "primary process" functioning, that would be characteristic of a moderately but unevenly cathected ego, which contrasted with the "secondary process" of an ego with more even distributions of cathexes. There is strong circumstantial evidence that Freud's "secret of dreams" - the wish fulfillment hypothesis - originally occurred to him as a logical deduction from this model (Fancher, 1971). So for Freud as for the complexity theorists, moderate degrees of complexity within a hypothetical system gave rise to interesting, and originally unpredictable "emergent" behaviors.
My final and most general association to Freud was triggered by Waldrup's account of John Holland's "Complex Adaptive Systems." These are extended networks of interconnected "agents," each one adapting in parallel with all the others, to its own "envirionmment" of immediate neighbors. Real world examples of such systems presumably include immune systems, economies, ant colonies, and nervous systems. Such systems, according to Holland, have no single or centrally localized agency of control; instead, control is dispersed throughout the system. Further, such systems are "dynamic" and constantly changing, and while each agent constantly acts so as to optimize its state, there is no single "perfect" or permanently optimum response. As Waldrup summarizes, "The space of possibilities is too vast; [the agents] have no practical way of finding the optimum. The most they can ever do is change and improve themselves relative to what the other agents are doing" (p. 147).
This reminded me of what I have always taken as Freud's ultimate and most general conception of the human condition - namely, that people are constantly beset by irreconcilable conflicts between "inner" demands originating in the body's physiology, "outer" demands imposed by the physical environment, and communal demands imposed by society and internalized as a sense of "conscience." One need not accept the specific concepts of id, ego and superego in order to appreciate the general logic of this conception, or of Freud's conclusion that the totality of these conflicting demands is too complex to allow for any "perfect" or absolutely optimum responses. Responses must inevitably be compromises, selected out of a vast range of possibilities and always imperfect to some degree. No "solution" - whether an anxiety-reducing symptom, a wish-fulfilling dream, or even a generally effective pattern of defense mechanisms - is ever permanent, or without costs along with its benefits. As Freud ruefully concluded in "Analysis Terminable and Interminable," no psychoanalysis can ever be "complete."
In the past I have told my students that, whatever one may think about the specifics of Freud's theory, in general it is one of the few to do justice to the "complexity of the problem" of psychological analysis. I was here using the word "complexity" in its everyday sense. I am still not certain how deep the connections I have discussed today will turn out to be. It now seems at least possible, however, that there is a new and more technical sense in which the term "complexity" is also appropriate.
Amacher, P. (1964). Freud's neurological education and its influence on psychoanalytic theory (Psychological Issues, Monograph 16). New York: International Universities Press.
Fancher, R. E. (1971). The neurological origin of Freud's dream theory. Journal of the History of the Behavioral Sciences, 7, 59-74.
Fancher, R. E. (1977). Brentano's Psychology from an empirical standpoint and Freud's early metapsychology. Journal of the History of the Behavioral Sciences, 13, 207-227.
Freud, S. (1954). The origins of psycho-analysis, Letters to Wilhelm Fliess, drafts and notes: 1887-1902, M. Bonaparte, A. Freud and E. Kris, eds. New York: Basic Books.
Waldrup, M. M. (1992). Complexity: The emerging science at the edge of order and chaos. New York: Touchstone.