(Engl. original:)
Free-Associations and Honeybee Dancers: The Unconscious
and its Place in Nature©
By D. L. Smith
(Copyright © D. L. Smith 1998)


I would like to begin this paper with a note of caution. To my mind there is little in the field of psychoanalysis and psychotherapy generally that deserves to be counted as genuine knowledge. Papers on psychoanalysis convince more through rhetoric than fact (Note: I regard psychoanalysis as a form of psychotherapy, and will use the two terms interchangeably). The psychoanalytic belief system to which one adheres is often more a matter of taste than of evidence. I regret this state of affairs as much as I am mired in it. So, please bear in mind that I offer the following reflections not as some definitive statement of truth but rather as a suggestion about how psychoanalytic research might be pursued. All such ideas must ultimately be brought before the tribunal of evidence, and in the absence of unambiguous corroboration should be regarded as conjectures, although hopefully stimulating and productive ones.

Free Association

What is the purpose of free-association? This question may sound strange, because we are accustomed to thinking of free-association as the conforming to a rule, the ‘fundamental rule’ of psychoanalysis. In this sense, the purpose of free-association is to drive the psychoanalytic process. In one of his earliest discussions of this topic Freud (1904) characterized ‘associations’ as: ‘¼ the involuntary thoughts (most frequently regarded as disturbing elements and therefore ordinarily pushed aside) which so often break across the continuity of a consecutive narrative (251). In this account, free association is a natural phenomenon: something that just happens. The free-association method is a way of enhancing this phenomenon, of offering nature some help.

 To secure these ideas and associations.... he [Freud] insists that they must. include...whatever comes into their heads, even if they think it unimportant or irrelevant or nonsensical; he lays special stress on their not omitting any thought from their story because to relate it would be embarrassing or distressing to them. (Ibid)

 The theoretical rationale for free-association is famously described in The Interpretation of Dreams as follows:  In the psycho-analysis of the neuroses the fullest use is made of these two theorems¾ that, when conscious purposive ideas are abandoned, concealed purposive ideas assume control of the current of ideas, and that superficial associations are only substitutes for suppressed deeper ones. Indeed, these theorems have become basic pillars of psycho-analytic technique. When I instruct a patient to abandon reflection of any kind and to tell me whatever comes into his head, I am relying firmly on the presumption that he will not be able to abandon the purposive ideas inherent in the treatment and I feel justified in inferring that what seem to be the most innocent and arbitrary things that he tells me are in fact related to his illness. (531)  So, although the rule of free-association has as its purpose the facilitation of the psychoanalytic process, the process of free-association - which the rule is designed to encourage - has as its purpose the expression of derivatives of unconscious mental contents. Now, the term ‘expression’ can mean several things. In the context of his theory of free-association, Freud most often uses ‘expression’ to mean something like ‘discharge’. In this view derivatives - and the free-associations that instantiated them - function to relieve inner tension: to maintain mental balance. Notice two features of this story. First, notice that the function of free-association has been specified purely with respect to the mind producing the associations. One could be in solitary confinement or morooned on a desert island and still discharge tension in this way. Second, free-associations so understood have no communicative function. This is not just a corollary of the previous point, because it is at least logically possible that a free-association might be a message to oneself. According to the discharge story, free-association does not occur in order to transmit any information. Of course, the discharge model does allow for free-associations carrying so-called ‘natural’ meaning along the lines of ‘where there’s smoke there’s fire’: they carry meaning in much the same way that a discharge of lightning carries meaning about atmospheric conditions.

There is an alternative outlook on the purpose of the free-associations that views them as messages, to oneself or to others. According to this story, free-association is a communicative act. This approach reached its clearest articulation and development in the work of Langs.


Irrespective of its being generally discredited, the fact remains that the discharge theory of free association is part of an over-arching meta-theoretical framework. In this section I will offer some suggestions about how this problem might be remedied. I will propose that the general idea, of adaptation and, more specifically, Ruth Millikan’s biological approach to mental content, provide valuable intellectual tools for refining and clarifying the communicative theory of free-association.

The notion of adaptation is the leitmotif of neo-Darwinian thought. Essentially, when we call some feature of an organism an ‘adaptation’ we mean that that feature exists to discharge some special function or functions with respect to its environment. Adaptation is therefore a sort of ‘fit’ between organism and environment, a fit that has become established through a process of selection. Plotkin (1993) claims, strikingly and poetically, that adaptation is a form of knowledge expressing an innate biological understanding of the world around one. A polar bear’s white coat is an adaptation to its snowy environment. In Plotkin’s view, the bear ‘represents’ the whiteness of its world in the whiteness of its coat.

It is perhaps warranted and informative to replace Plotkin’s claim with the idea that adaptations are not just knowledge, but are a kind of wisdom. Adaptations do not merely provide information about the world, they are all about how living things have come to grips with the dilemmas of life.

Adaptations are marks of biological intelligence. If we can make the case that the propensity for free-association is an adaptation and that free-association is the expression of unconscious mental processes, we will have made a case for unconscious intelligence.

Ruth Millikan is a philosopher who uses neo-Darwinian ideas to clarify issues in the philosophy of mind and philosophy of language. In one of her naturalistic accounts of mental content, Millikan distinguishes between tacit assumptions, intentional icons and mental representations. Tacit assumptions are the most primitive forms of content, representations are the most sophisticated, with intentional icons occupying an intermediate position.

 Tacit assumptions are ‘hard-wired’ adaptations to fixed environmental features. The polar bear’s white coat tacitly assumes a white environment. Tacit assumptions can also be variable (within a pre-set range). The camouflaging ability of the chameleon provides an example of this. Notice that tacit assumptions do not require an interpreter: neither the polar bear nor any other creature needs to make sense of the whiteness of the bear’s coat in order for the coat to serve its evolved purpose. Contrast this with the threat display of a male lizard, which needs to be ‘understood’ by other male lizards in order to work.

Intentional icons are a special class of adjustable adaptations. As well as being changeable in response to environmental contingencies, intentional icons possess two additional characteristics. First, there must be some mechanism the proper function of which is to generate the icon, and there must be some way that this mechanism is supposed to ensure that the icon maps onto the right environmental feature. Millikan calls this mechanism the producer of the icon. The second feature of intentional icons is that they have consumers. A consumer of an icon is a mechanism, the operations of which are guided by the icon.

Millikan’s paradigm case of intentional iconing is the Schwänzeltanz or ‘waggle-dance’ of the honey-bee. The honeybee is a social species of bee that can frequently be observed performing a curious ritual in its hive. In the presence of an excited audience, a bee regurgitates nectar and then moves round and round in a figure-of-eight pattern whilst buzzing and shaking vehemently. Karl von Frisch, an Austrian zoologist, decoded the meaning of the waggle dance in the early 1920’s. Von Frisch found that the dance is a message to the other worker bees encoding the location from which the nectar was obtained. In brief, the speed of the dance encodes the travel time from the hive to the nectar. Thus, if the bee completes eight or nine loops in fifteen seconds, it means that the nectar is two hundred metres away from the hive, if the bee completes only four loops in fifteen seconds, the dance means that the nectar is 1500 metres away, and so on. The angle of the straight part of the bee’s run represents the angle between the sun and the nectar source relative to the hive, transposing a horizontal bearing to the vertical plane of the hive, with the direction of gravity corresponding to the position of the hive relative to the sun. While she performs her dance, the workers gathered around her observe the honeybee dancer with great care. Observation gives way to understanding, as the workers make a bee-line for the food source represented in the dance.

In the case of the honeybee, the dance is the icon, produced by some special instinctual system within the dancing bee, and consumed by a corresponding device within the observing bee. When an organism produces an icon that it consumes itself, this can be described as the organism sending a message to itself. Here is an example given by Millikan (1993).

 To be able to negotiate physical movement among the objects in its world, an animal must be prepared to deal with any of innumerable arrangements of objects surrounding it. In the case of many sighted animals, these arrangements are mapped onto patterns of ambient light impinging on the moving animal and brought to a focus through a lens in the eye. A description of the rules by which variations (transformations) in the pattern of the light impinging on the animal correspond to variations in the arrangements of the surrounding objects is exceedingly complex. But the mapping is there, and the animal uses it. It uses the pattern, presumably, by projecting it internally to become a pattern in the nervous system which again maps onto the arrangements of external objects. Only in this way could the movement be guided according to the arrangement, guided as, in part, a function of this arrangement (111-112).

Representations are reflexive (or reflective) intentional icons. When we represent something, in Millikan’s special sense of ‘represent’, we know what it is we are representing: we are aware of how our intentional icons map onto the world. Millikan calls this the process of ‘identification’. She argues that representation entails having multiple maps of the world that are coordinated with one another to yield new information.

Imagine a creature that carries about in its head intentional icons that are three-dimensional maps of various places it has been - not maps in three spacial dimensions, of course, but dimensions in some neurally realizable medium. It has a map of the locale in which it last found water, and another of the locale in which it last saw lions. On each of these maps its den is marked. Now imagine that it overlaps these maps, using its den as a pivot, and arrives at a third map showing the proximity of lions to the source of water. Reacting to this new map, it seeks a new source of water rather than going back to the lion-infested source on its map....Two vehicles of information have been combined, using a middle term, so as to produce a third vehicle containing new information (113-114) Millikan’s way of putting things allows one to make some subtle distinctions. Notice that reflectiveness, the main property of representation, is not the same as sending a message to oneself. One canin principle ‘consume’- make use of - one’s own icon without identifying the icon. Notice also that consciousness of the meaning of an icon, presupposes its status as a representation.

Unconscious Communication

I will now return to the main subject of this paper and argue that unconscious communication (through the vehicle of spontaneous free association), can be fruitfully understood as a process of intentional iconing. I have already mentioned that the discharge model has been the dominant psychoanalytic approach for understanding the phenomenon of free association (perhaps because explanation in terms of proximal causes has been mistakenly regarded as more ‘scientific’ than functional or teleological explanation). Freud seemed quite reluctant to entertain any notion of unconscious communication. It is perhaps significant that most of his examples of unconscious communication are taken from outside the psychoanalytic setting. Here are two examples, cited by Boudreaux (1977), from The Psychopathology of Everyday Life (1901).

Here is another example, this time from ‘Constructions in analysis’ (1937). In contrast to his views on unconscious communication, Freud was much more open to the idea that we are unconsciously able to infer the unconscious meaning of the behaviour of others, but considering his general advocacy of the discharge model, this should probably be understood as an inferring rather than a genuine decoding. Freud probably believed that our unconscious understanding of one another was based more on an ‘if there’s smoke, there’s fire’ sort of reasoning than on a grasp of the denotation of symbolic behaviour.

To continue further with this argument, I will need to provide a summary of Langs’ communicative theory of unconscious mental processing and interaction. I have already mentioned that I regard Langs’ work as the most explicit and consistent articulation of the communicative understanding of free-association. Langs calls his theory the Communicative approach (I will henceforth distinguish between the generic and Langsian senses of ‘communicative’ by spelling the latter with an upper-case ‘C’).

According to Langs’ the mind is so constituted as to encode unconscious perceptions of interpersonal events by means of apparently unrelated narratives. These narratives are unconsciously interpreted by a corresponding mental system. The encoding and decoding devices are assumed to be components of a more inclusive mental system which Langs calls the ‘deep unconscious wisdom system’. A third hypothetical component of the deep unconscious wisdom system is an unconscious analyzing system.

If the Communicative account is substantially correct, it would seem reasonable to assume that the analyzing component of the deep unconscious wisdom system has been selected by mother nature to monitor the highly complex social environments peculiar to human existence. Essentially, the analyzing component appears to be dedicated to keeping track of others’ attempts to take advantage of one for their own benefit. It is therefore particularly concerned with such issues as the maintenance or erosion of interpersonal boundaries, honesty, the honoring of commitments and what Cosmides (1989) has called ‘the logic of social exchange’ - that is, the notion that no one in a social system should avoid paying their dues. Communicative theory uses the term ‘frame’ to denote social arrangements and the rules by means of which they are constituted.

These ideas were initially generated through studies of psychoanalytic and psychotherapeutic interchanges. The psychoanalytic situation is a relatively simple, two-person social system constituted by the ground-rules of psychoanalysis, such as neutrality, relative anonymity, privacy, confidentiality and so on (Langs, 1993; Mooij,1991). The concern for these issues evidenced by the operation of the deep unconscious wisdom system is of obvious adaptive value for intelligent creatures incessantly encountering competing interests within complicated social systems.

The existence of this unconscious monitoring or analyzing device is inferred on the basis of what happens when the rules of a social system are violated. This is most easily studied within simple social systems like the psychoanalytic situation that contain few variables. Langs found that when a basic, constitutive social rule is violated by a psychotherapist, the patient may or may not respond to this consciously and manifestly. When patients do respond consciously, their responses may differ dramatically from one person (and indeed one instance) to the next. In constast to this, free-associative responses to such events normally possess certain properties in common. They will usually possess a negative thematic tone and represent scenes that correspond to the violation and its implications Communicative theory claims that analysands prove to be particularly sensitive to any indication that their analysts’ self-interest takes precedence over an altruistic commitment to their (the patients’) own well-being This process is quite lawful (although - like any biological processes - not exceptionless) and enables anyone sufficiently well versed in the theory to generate reasonably specific predictions about the thematic content of narrative episodes occurring during psychotherapeutic sessions. According to Communicative theory, the very same processes can be found in any social situation.

It might be supposed that such unconscious responses are brought about by the violation of any social rules; i.e., that unconscious responses are contingent upon the rules that one adopts. This does not seem to be the case. According to Communicative theory, we all harbour genetically given social expectations. We are born with an innate social wisdom, acquired by means of natural selection over many generations. According to this view, we enter this world equipped with deep psychological structures for social interchange in much the same way that we are born with an innate understanding of the deep structure of language. These structures appear to operate independently of our conscious social commitments and the specific rules of the culture in which we dwell. They seem to be universal rather than culturally relative, although they are probably realized through diverse cultural forms. Furthermore, these unconscious structures seem to be highly stable and relatively impervious to influence.

Perhaps an example taken from a psychoanalytic hour will illustrate these points. Berns (1994) provides an example in which an analyst has failed to explain to his patient her role responsibilities (more specifically, he has failed to explain the ‘fundamental rule’ of free- association). Berns reports that he had formed the impression that the analysis under consideration had come to a standstill, although he had no idea why this should have been the case. During the session in question, the patient reported that she had begun to notice that her students cooperate and work best if she makes demands on them and tells them what they have to do and she wondered why this was the case. Guided by the theme of people requiring instruction in order to work well, the analyst recalled that because the patient spoke fluently and had been in a lengthy analysis some years before, he had neglected to explain to her the fundamental rule of free association. Treating the patient’s narrative as unconscious portrayal of his violation of a fundamental social rule (in this instance, the rule that someone offering a service must make explicit the associated role-requirements) he went on to say to her that he recalled that he had not made it quite clear what her analytic ‘job’ was and that she too might function better if he treated her in the same manner as she did with her pupils. He then went on to explain the rule of free-association.

According to Communicative theory, the recognition and restoration of such ‘framework violations’ should be confirmed by an appropriate unconscious response in the patient. The patient should generate a narrative unconsciously portraying the analyst in a positive light, a ‘derivative validation’. In this example the patient said, after a pause: (Editor's Note: Unpublished patient material omitted.) (Note: This is from page 13 of an unpublished English translation of Berns’ paper.)

It is essential to remember that Communicative theory claims that these processes are law-like, and that it would therefore have been very unlikely for the patient to have produced a predominantly negatively toned association following the intervention reported above.

There is a good deal of Communicative literature on processes of unconscious meaning analysis (Dorpat and Miller, 1992) and on the generation of ‘derivatives’ - free-associations expressing these unconscious meanings. There has been much less work done on our purported ability to unconsciously ‘decode’ the associations produced by others. This is a very significant lacuna. Why it is that we have evolved the capacity to unconsciously express information in the form of free-associations? What purpose does this serve? Surely, Mother Nature did not cultivate this capacity in anticipation of the creation of Communicative psychoanalysis! We are in a position analogous to that of an entomologist who understands the meaning of waggle-dancing but does not understand why it is that these dances are performed; i.e., he has not yet understood the significance of the dance for the spectator bees.

Following the honeybee analogy, the obvious hypothesis must be that unconscious communication is intended to regulate the behaviour of others. Clinical experience does not immediately support this conjecture. Anecdotal evidence suggests that analysands can produce pertinent associations until they are blue in the face without this having any constructive impact on the behavior of their analysts. Using Berns’ example, the analyst would have been unlikely to have clarified the ground rules of the analysis without conscious efforts to decode this material in light of a specific theoretical model. In fact, the more unconscious meaning an analysand injects into a session, the more frantically his or her analyst will be inclined to obliterate that meaning (Langs, 1982, 1985; Smith,1991). This is a tragic paradox of the psychoanalytic situation. Here is a reasonably typical example. The patient is a ‘psychotic’ adolescent boy named Dominique. The analyst is Françoise Dolto (Dolto, 1974) This is an excerpt from the very first meeting between Dolto and Dominique. What is important about it for the purpose of this exposition is that just before the meeting, Dominique’s mother had met with the analyst, a meeting from which the patient was excluded. According to communicative theory, such events are virtually always understood to possess conspiratorial qualities. After introducing herself, Dolto asks Dominique if there is anything he wants to say to explain his feelings. The interview then proceeds as follows:

With a frozen, anxious smile he says ‘ Well me, I’m not like everybody else. Sometimes when I wake up I feel like I’ve lived through a true story.’....

I SAY TO HIM: A story that made you untrue.

DOMINIQUE: That’s it! But how do you know?

ME: I don’t really know. But that’s what I think when I see you.

DOMINIQUE: I thought I was in the dining room as a little boy again, afraid of burglars. They can take money, they can take silverware. Don’t you think they could take almost anything?

ME: They might take your little sister?

DOMINIQUE: Oh you, how do you know everything?

ME: I don’t know anything at first, but it’s because you say things to me in your own words and I listen as well as I can. It’s you who knows what’s happened to you, not me. Together we might be able to understand.

Silence. I wait a long while, then:

ME: What are you thinking about?

DOMINIQUE: I’m trying to figure out what’s wrong in life. I’d like to be like everyone else. For instance, when I read a lesson several times over, I still don’t know it the following day. Sometimes I feel I’m dumber than everyone else, or I say to myself that I can’t go on like this. Things are going wrong. I’m talking nonsense (29-30).

Let’s consider this vignette from a standard Communicative perspective. Although such analysis must remain purely conjectural, and does not aspire to exhaustiveness, I think it provides a model of an interesting and useful way of thinking about human interaction.

The fact that Dolto feels prepared to interpret Dominique’s cryptic opening statement might be taken to imply that she has based the intervention at least in part on information obtained from Dominique’s mother. From an evolutionary perspective, it is reasonable to assume that his mother’s version of his ‘story’ will inevitably be self-servingly biased insofar as their interests diverge (Slavin and Kriegman, 1992). The printed page cannot capture the emotional inflection of Dominique’s response. We do not know if he was (consciously) amazed, frightened or sceptical. It seems clear, though, that he wants to know how she knows. At this point, the analyst dissimulates. She claims that her interpretation was derived from her her careful listening to him. The implicit confluence of interest between mother and analyst, which the analyst conceals, may have alerted Dominique to the presence of cheating and exploitation. He mentions thinking of being a little boy in the dining room. I take this to express the theme of being dependant on the adults around one, a theme that is both biologically poignant and very pertinent to Dominique’s relational situation. The story of the burglars expresses the them of an intruder secretly and illicitly securing his own advantage at someone else’s cost. Understood as an intentional icon, this burst of free-association can be seen as Dominique telling Dolto that he is aware of and disturbed by her alliance with his mother. It’s biological purpose, I suggest, was to end this situation. Dolto responds by interpreting the account as an unconscious reference to a past relational dilemma not involving her and based on information provided by Dominique’s mother. In short, she ignores the adaptive aspects of Dominique’s statements and attempts to persuade him that he is concerned with something else entirely in which she is not at all implicated. This seems to indicate a strategy of deception. Again, Dominique wants consciously to know how Dolto knows this, and again Dolto attempts - presumably without realizing it - to throw sand in his eyes, claiming that her interpretation is primarily derived from what Dominique has said to her. Dominique falls silent and, when probed, expresses considerable confusion. I submit that Dominique’s silence and confusion may well have been iatrogenic phenomena. They may have been consequences of Dolto’s successful disruption of Dominique’s attempts to get an adaptive grip on an exploitative situation.

 Free-Association As Intentional Iconing

There are examples in the Communicative literature that do support a honeybee analogy, although they are rather thin on the ground. I think that many of us familiar with the theory can report such anecdotes from everyday life. Here is one such example. (Editor's note: Personal information omitted.)

I will grant that this vignette might be understood in a variety of ways. For the sake of the discussion, though, let us treat it as an example of unconscious communication. Mrs. A has violated the social rule of reciprocity amongst equals by dominating the conversation. The deep unconscious wisdom system of Mrs. B portrays this action by means of the free association about her child being bullied. This icon is unconsciously ‘interpreted’ by a consuming device in Mrs A, which then regulates her behaviour accordingly. On the honeybee analogy, we do not have to assume that Mrs A unconsciously identifies Mrs B’s narrative as a portrayal of her invasive behaviour. The narrative does not have to function as a representation. Regarded as an intentional icon, it is enough that the narrative regulates A’s behaviour in the right sort of way.

The story of Mrs A and Mrs B seems to be a clear counterexample to the typical situation in psychoanalysis as described earlier. Now, it is perfectly consistent with biological reasoning to claim that although free associations are designed by Mother Nature to regulate human behavior, this function is not always realized. Perhaps the analytic situation is biologically deviant, providing an environment in which the deep unconscious wisdom system is unable to function properly (just as my computer will not function properly in a sauna). I think that it is too easy just to say that the deep unconscious wisdom system is imperfectly adapted to the psychoanalytic situation. After all, the idea of the deep unconscious wisdom system was formulated on the basis of studies of interactions within that very setting! Might there be a single principle that explains both the psychoanalyst’s imperviousness to his patient’s unconscious communication and Mrs. A’s cooperative response to Mrs. B’s free associations?

Let us compare the analyst/analysand interchange with the interchange between Mrs. A and Mrs. B. It seems intuitively obvious that Mrs A and Mrs B are approximately on a level, whereas in the analytic relationship there is clearly an element of asymmetry. Insofar as the analysand has a greater investment in the analyst than vice versa, and therefore has more at stake than the analyst, the analyst can be regarded as the dominant partner. Might it be, then, that the normal reaction to an intentional icon pertaining to a social interchange will vary as a function of the relative dominance of the person consuming the icon? Given the simplest case of a two-person interchange, this hypothesis suggests that an icon portraying an act of social violation committed by the dominant participant normally results in that participant attempting to override the objections of the less dominant partner. This is just an instance of the innate tendency to press one’s advantage. Analysts are insensitive to the unconscious criticisms offered by their patients not because they are ‘defensive’ - at least in the standard Freudian sense of the word - or because they are pathological sadists - but because, like all of us, they are biologically disposed to exploit others. The purpose, then, of the psychoanalysts’ response is to cause the submission of the patient, and it seems to work a good deal of the time. Although morally repugnant, this would appear to be an adaptive response: when someone catches your interests overriding theirs, you are inclined to press your advantage. Furthermore, this reaction is just as much a reflection of the deep unconscious wisdom system as was the original encoding. Following the honeybee analogy, imagine a bee mutant that responds to a dance by disrupting it, in order to obtain the nectar for herself.

I have not yet considered the possibility that the ‘consumer’ of an unconscious icon is part of the same person as its ‘producer’. If we imagine this as operating in much the same manner as interpersonal communication a situation is implied in which unconscious icons - in the form of free associations - are read by an unconscious interpreting device that then uses these messages to regulate behaviour. Again, this model seems to be strongly falsified by observations made within the psychoanalytic setting. Analysands who unconsciously believe that their analysts are behaving inappropriately and hurtfully, and who express these impressions by means of derivatives (and through free associations), do not normally heed their own advice. Just as analysts seem driven to override patients’ objections to their behavior, so analysands seem fatally attracted to those very situations that they unconsciously believe to be harmful. This pattern can, and has, been interpreted as the expression of an unconscious need for punishment or an attempt to find a cure through nefarious comparison. I wonder, though, whether this it is not an example of an adaptive, self-regulatory process becoming derailed by the peculiarities of the psychoanalytic situation. It is certainly my impression that most people are better able to ‘trust their intuition’ outside psychoanalysis then in it.

A moment’s reflection shows how this unconscious self-communication might work. Let’s go back to the example of Mrs. A and Mrs. B. We can assume that while Mrs. A was talking to her, free-associative thoughts began to arise in the mind of Mrs. B.: unpleasant thoughts of someone being bullied. It also seems safe to assume that these thoughts - expressing B’s unconscious analysis of A’s behavior - colored her mood. In thinking of her son being bullied, B began to feel bad in the presence of A.. I submit that such mood alterations might, under such circumstances, issue in the appropriate behaviour. Mrs. B would be inclined to move away from Mrs. A., just as she imagined removing her son from the school. Now, I am well aware that things do not always pan out in this way, but it is my distinct impression that the mood-altering effects of free-associations ‘work’, i.e., produce the appropriate behavioural effects, enough of the time to give this process a clear adaptive significance.

 The Problem of Altruism

I have claimed that free-associations are designed to convey information about the violation of social rules. Another way to put this is that we engage in free association primarily to signal to others that we are ‘onto’ them: that we understand that they are taking advantage of us. An individual at the receiving-end of such a message might respond in several ways. They might attempt to trick the victim into revising their opinion, they might attempt to intimidate the victim, they might sever the relationship or they might cease attempting to exploit the victim.

There is a sizable neo-Darwinian literature dealing with altruism, exploitation and trickery (Badcock, 1994). To be brief, biologists distinguish between kin and reciprocal altruism. Both of which are ultimately expressions of self-interest. Kin altruism is an expression of self-interest from a gene’s eye view, as activities that support one’s close blood relations may enhance the likelihood that one’s genes will proliferate. Reciprocal altruism operates on the principle of ‘you scratch my back and I’ll scratch yours’.

Reciprocal altruism is an advantageous strategy, particularly in conditions of scarcity. Even greater advantage can be secured, however, by cheating: by surreptitiously getting more than one’s fair share from the process of social exchange. Of course, cheating can only work if it remains undetected. Trivers (1976), a biologist, was the first to suggest that this has lead to the rendering unconscious of many self-serving motives and strategies. What better insurance could their be against giving the show away than remaining unaware of one’s own dishonesty? The evolution of cheating and the repression of the awareness of cheating may have led to a biological ‘arms race’ with the evolution of ever more sophisticated forms of cheating and cheating detectors (Trivers, 1985).

In a situation where we are consciously recognized (‘represented’ in Millikan’s terminology) as cheating, we may attempt to change our victim’s mind - to convince him that we are not really exploiting him - to maintain our advantage. An iconing situation requires a different approach because organisms are by definition unconscious of the meanings of the intentional icons that they produce and consume. The way to override an intentional icon - an ‘unconscious perception’ if you prefer - portraying an act of exploitation is to prevent the icon from leading to some action that ends the exploitation. This is a subject that deserves very considerable investigation in the context of psychotherapy. There is evidence to suggest that certain types of intervention (e.g., asking questions) consistently disrupts the production of these messages, and that in common with other biological systems the deep unconscious wisdom system can be induced to malfunction when subjected to sufficient interpersonal stress (Langs, 1993). It would seem that psychoanalysts are driven to disrupt and ignore the unconscious communications of their patients just because there is something about the situation that lets them get away with it. I will avoid the temptation to speculate about just what this ‘something’ is. Unconscious communications seem to be much more likely to be heeded in egalitarian situations - as in the conversation between Mrs. A and Mrs.- that promote reciprocal altruism.

Given that there is something about the analytic situation that permits analysts to get away with cheating, it follows that there is something deeply unnatural about the attitude demanded of practitioners, particularly practitioners of the Communicative approach who are supposed to recognize, interpret and heed analysands unconscious messages about them. Langs has often described Communicative psychoanalysis as an ‘unnatural act’ running counter to human nature. In a similar vein, Slavin and Kriegman (1992) write that:

It is an extraordinarily complicated and, we must admit, unnatural process by which one human being comes to know and become deeply allied with the interests of another to a degree that is sometimes achieved in the analytic relationship (244). I think that it is this factor that makes Communicative psychoanalysis so difficult to learn and to practise. The altruism required of the psychotherapist does not come easily or naturally. It can only be marshalled at the cost of tremendous conscious effort and vigilance. So, there is a sense in which the very structure of the psychotherapeutic situation operates against the aims of Communicative psychotherapy.

There is a further implication of all of this that might be used to test the hypothesis. The logic of my argument suggests that unconscious communications are most likely to be heeded when they pass down the gradient from the person with the least investment to the person with the greatest investment. Furthermore, the consistency of unconscious responses should directly vary with the steepness of this gradient. So we should find, for example, that patients are normally strongly inclined to cooperate with the unconscious imperatives of their therapists. We should also find that child analysts are particularly heedless of the unconscious messages of their patients, but it does not necessarily follow that children are especially responsive to the unconscious messages transmitted by their analysts.


I have argued that the purpose of free-association is to unconsciously convey information about the violation of the rules to which we unconsciously expect social systems to conform. These communications are of a general type found throughout the animal kingdom: they are intentional icons. The purpose of the production and transmission of intentional icons is to regulate the behaviour of the organisms consuming them. I have suggested that the asymmetrical features of the psychoanalytic setting which discourage reciprocal altruism disrupt this regulatory process and that it is exceptionally difficult for analysts to transcend the opportunities for exploitation and ‘cheating’ presented by the analytic situation.

I have attempted to express my central hypothesis in a readily falsifiable form. Should it be corroborated, the present paper suggests that research is urgently required to develop forms of psychotherapy better able to harness our ancestral unconscious wisdom and less inviting of interpersonal exploitation..


I would like to thank my wife, Prof. Emmy van Deurzen-Smith, who helped me to appreciate the significance of biology - the science of life - for psychoanalysis.

Berns, U. (1994) The tally interpretation: an outcome of evaluation analysis. Forum der Psychoanalyse, 10(3): 226-244.

Badalamenti, A.F. & Langs, R.J. (1992) Stochastic analysis of the duration of the speaker role in the psychotherapy of an AIDS patient. American Journal of Psychotherapy, 46: 207-225.

Badcock, C. (1994) PsychoDarwinism: The New Synthesis of Darwin and Freud. London: Harper/Collins.

Boudreaux, G. (1977) Freud on the nature of unconscious mental processes. Philosophy of the Social Sciences, 7:1-32.

Cosmides, L. (1989) The logic of social exchange: Has natural selection shaped how humans reason? Studies with the Wason selection task. Cognition, 31: 187-276.

Dolto, F. (1974) Dominique. London: Souvenir Press.

Dorpat, T.L. and Miller, M.L. (1992) Clinical Interaction and the Analysis of Meaning. Hillsdale, N.J.: The Analytic Press.

Freud, S. (1901) The psychopathology of everyday life. Standard Edition, 6.

----- (1904) Freud’s psycho-analytic procedure. Standard Edition, 7.

------- (1937) Constructions in analysis. Standard Edition, 23.

Langs, R. (1982) The Psychotherapeutic Conspiracy. New York: Jason Aronson.

----- (1985) Madness and Cure. Emerson, NJ: Newconcept.

----- (1993) Science, Systems and Psychoanalysis. London: Karnac.

----- & Badalamenti, A.F. (1990) Stochastic analysis of the duration of speaker role in psychotherapy. Perceptual and Motor Skills, 70: 675-689.

Millikan, R.G. (1993) On mentalese orthography, part I. In White Queen Psychology and Other Essays for Alice. Cambridge, MA: Bradford/MIT.

Mooij, A. (1991) Psychoanalysis and the Concept of a Rule: An Essay in the Philosophy of Psychoanalysis. Berlin: Springer-Verlag.

Plotkin, H. (1993) The Nature of Knowledge: Concerning Adaptations, Instinct and the Evolution of Intelligence. London: Allen Lane/Penguin Press.

Slavin, M.O. & Kriegman, D. (1992) The Adaptive Design of the Human Psyche: Psychoanalysis, Evolutionary Biology, and the Therapeutic Process. London: Guilford.

Smith, D.L. (1991) Hidden Conversations: An Introduction to Communicative Psychoanalysis. London: Routledge.

----- (1992) Where do we go from here? International Journal of Communicative Psychoanalysis and Psychotherapy, 7(1): 17-26.

Trivers, R. (1976) Forward to The Selfish Gene by R. Dawkins. Oxford: Oxford University Press.

----- (1985) Social Evolution. Boston: Addison-Wessley.

Notes:  i- According to Slavin and Kriegman (1992), ‘The universal, evolved design of the mind is...profoundly shaped by hundreds of thousands of generations of exposure to repeated, historical social selection pressures....The child is...heir to this legacy of ancestral, encoded "wisdom"....’ (287).

ii- There is some dispute about how this process should be understood. Langs (1993) believes that the process should be understood as a separate perceptual pathway, whereas I have advocated that the system in question just analyses consciously perceived information in a distinctive fashion.

iii- Mooij (1991) argues that ground-rules possess a constitutive function for social systems. It follows that the violation of a ground-rule is a critical event for any patient-therapist system. Langs (1993) has argued eloquently for this on clinical grounds.

iv- I follow Millikan’s argument that biological explanation adverts to norms rather than laws. A biological norm is just that which a living system was designed (by natural selection) to do. All biological systems can malfunction, can fail to perform their proper function, for a variety of reasons. They may, for instance, be operating under environmental conditions to which they are not adapted. Some biological systems, e.g. spermatazoa, only occasionally fulfill their functions. Contrast biological norms to exceptionless physical laws. Also, note that the notion of malfunction cannot by expressed in the language of physics. Purely physical systems, such as atoms, cannot malfunction. Artefacts can malfunction because, like organisms, they have been designed.

v- This term is taken from Slavin and Kriegman (1992)

vi- Unfortunately, many of these expecations are geared towards the hunter-gatherer mode of social organization (in which we have lived for over 99% of our evolutionary history). So, for example, the provision of psychotherapy under the NHS is unconsciously regarded as a provision for which the price has not been specified and is therefore felt to be suspect and exploitative. Although wise in many respects, the deep unconscious system can also be stupid in relation to circumstances with which it has not been adapted to deal.

vii- The deep unconscious social norms identified by communicative psychoanalysis should not be regarded as exhaustive. It seems at least likely that certain of these rules are relevant to the class of social interactions of which psychoanalysis is an example. By the same token, a number of relational issues are not replicated within the psychoanalytic situation (e.g., intergenerational conflicts). Also, whilst not advocating cultural relativism with regard to these issues, it seems likely that the interaction between cultural norms and unconscious social expectations is quite complex (Smith, 1991).

viii- This example is discussed in greater detail in Smith (1991). However, I regard the present analysis as superseding the earlier one insofar as they differ.

ix- It is not difficult to understand why this should be. The analysis was arranged by Dominique’s mother. The analyst was employed by Dominique’s parents. There are no mechanisms built in to this analytic scenario that constrain the analyst to represent Dominique’s intersts insofar as they diverge from those of his parents and those of the analyst. Slavin and Kriegman (1992) provide a very useful biological account of the inevitable clash of interests between parents and offspring.

x- Of course, given his mother’s involvement, this may also typify a relational pattern in Dominique’s family. However, the vignetted does not provide any evidence for or against the hypothesis that Dominique’s mother is able to successfully override his unconscious perceptions of her behaviour.

xi- Langs and Badalamenti (1990) and Badalamenti and Langs (1992) have garnered evidence indicating that this dominance extends to speech rate patterns. In their sample, the analyzands’ rate constants consistantly conformed to those of their analysts. However, this should not blind us to the possibility that linear influence also can flow from analysand to analyst (see Langs, 1993).

xii- The conventional solution to this problem is to attribute the analyst’s power to ‘transference’. This claim is impossible to evaluate unless one commits oneself to a definite notion of what ‘transference’ denotes. ‘Transference’ is used in so many senses that I think it is best to avoid using the term in an explanatory context.


The author of this article, can be reached by E-mail: