Wittgenstein and Freud

Freud saw psychoanalysis as a natural science, but his critics deny that it is scientific. This is a complex issue, but it is worth exploring in detail. Two points seem hard to dispute. Firstly, it is clear that psychoanalysis aims to be a truth-focused undertaking. It puts forward explanations of why people do what they do (and feel what they feel) that we can either accept as correct or reject. The statement, “Bob forgot to send the invitation to Ben because he unconsciously did not want him to attend the event”, explains Bob’s mistake, and we can accept this explanation, reject it in favour of another explanation or deny that Bob’s mistake needs this kind of explanation. Secondly, psychoanalysis takes a systematic approach. It does not simply generate isolated claims. Rather, it develops models of the mind that bring together multiple explanations in an articulated whole. These points explain the position of the defenders of psychoanalysis. A systematic attempt to establish the truth in a specific area might well be described as a scientific undertaking.

But psychoanalysis is very different from the natural sciences. It is not focused on generating testable hypotheses, making predictions or establishing law-like regularities. Its evidence relates to individual cases, each of which is recognised to be unique. Furthermore, the situations that generated the evidence on which it is based cannot be replicated. An incident in a session cannot be replicated by another analyst or indeed by the first analyst involved. Psychoanalysts do, of course, share and discuss their clinical material, but the most important aspects of this material (the exact tone of voice, facial expression, bodily movements of the patient, the impact these had on the analyst, any countertransference feelings the analyst may have experienced, etc.) cannot be shared in a neutral form. Any description of a session in human terms (“the patient reacted negatively to my interpretation”) presents what happened in a certain light (was it a definitely reaction to the interpretation? Was the reaction entirely negative? Was it entirely genuine?). This means that analysts may disagree with each other not just about how what happened in a session should be explained, but also about how it should be described. 

As these points suggest, there is radically more scope for disagreement in psychoanalysis than there is in the natural sciences. In fact, there is no process within psychoanalysis for eliminating disagreement. This applies both to specific judgements and to general claims. One analyst may claim that the patient’s behaviour during a session was driven by unconscious envy, and another analyst may deny this. The two may, in the course of discussion, come to a shared view, but this may not happen. The first analyst may be tempted to say: “You would have had to be in the room”. But even if the other psychoanalyst had been in the room, they might still disagree. Similarly, there is no process for resolving the disagreement between analysts on general issues, e.g. as to whether (destructive) aggression is a primary or secondary driver of human behaviour. After years of clinical work, some analysts are firmly committed to one view, some to the other.

So, why does psychoanalysis differ from the natural sciences in the ways described? The answer lies in the nature of its interest in human action. To understand this point, we need to go back to everyday discussions about human action. When we do not understand why someone did something, we ask the person concerned, who offers an explanation of their action. They may be able to explain their action in a word or a short phrase, but especially in complex situations, they may need to tell us a long story to help us understand why they did what they did. If I ask my friend Bess why she broke up with Bob, she may say: “he was too controlling,” or she may give me a detailed account of what he did and how she felt about it. Bess explains her action without conducting an investigation and without exploring a range of hypotheses. She also does not need to introspect – unless by “introspect” you mean reflect, and even then, we are usually able to explain our actions without needing to reflect. 

The practice of asking for and giving reasons is a central part of human life. We ask people for their reasons because we relate to them as human beings (who have feelings and intentions, etc) and because we want to understand them. We want to be able to make sense of their actions, and the conclusion we reach about the correct explanation for their action will shape our relation to them. I may be very sympathetic to Bess if I believe that she broke up with Bob because of his controlling behaviour, and less sympathetic to her if I believe that the real reason she broke up with him was because she discovered that he was not going to inherit his father’s wealth. Understanding those around us helps us predict what they are likely to do, but the effort we put into understanding people’s actions is not about trying to create a maximally predictable environment. It is about relating to others as human beings and understanding them as people.

One way of capturing these points is to say that reasons are not causes. Against this, it may be pointed out that we often use the words “reason” and “cause” interchangeably in relation to human action. But it is not about these specific words. What matters is the nature of the understanding at stake here. Bess may say that Bob’s controlling behaviour “caused” her to end their relationship, but her explanation provides a way of making sense of her action rather than presenting it as an instance of a law-like regularity. This should not be taken as implying that there is no question of truth here, that it is just a matter of different possible ways of seeing the action. Establishing the correct account of why someone did something is vital not just in court cases, but in all our interactions with people. Why people do things matters to us, and we put a lot of time and effort into determining the meaning that should be assigned to their actions.

Despite these points, we may be tempted to argue that a reason is only relevant if it is operative and being operative can only mean being the causal factor. Here, the word “operative” is being used to confuse the issue and smuggle in the assumption that only causes really explain. If Bess’s anger at Bob’s controlling behaviour played no role in her decision to end their relationship, then it was not the reason for her decision, but the link between her action and the reason is created by her citing it (and by us accepting this if we do), not by evidence that a previously dormant causal factor has become operative. Donald Davidson (and other philosophers) are wrong to see the “because” that links the reason and the action as mysterious. What makes the reason relevant to the action is the claim that it provides the correct way of understanding it. If I come to believe that Bess left Bob because she discovered that he was not going to inherit his father’s wealth, I have changed my view on why she did what she did, but it would be ridiculous to present this as my discovering that the causal mechanism operated differently than I had been led to believe.

So far, we have focused on the reasons for action given by the person who performed the action. This is the aspect of our language game with reasons that most obviously lacks any equivalent in the language game of explaining physical events in causal terms. But our language game is not just based on the reasons the agent gives. In fact, we often reject the account the agent gives.  Sometimes this is because we think that they are being insincere (i.e. offering an account that they know is incorrect). Sometimes it is because we think that they are deceiving themselves. In rejecting their account, we do not jump from the language game of reasons to the language game of causes. We are still trying to make sense of their action, we are just rejecting their account in favour of a different one. We base our judgement on what we know about the person concerned (other things they have said and done) and on our wider experience of human life (our interactions with other people, what we have read, etc). No two people will have exactly the same basis for making a judgement, and therefore it is not surprising that we often disagree. My friend may say, “I just can’t believe Bess would leave Bob for financial reasons”. To which I may reply: “Well, you obviously don’t know Bess very well” or “You obviously don’t know people very well”.

Most of us try to get better at understanding other people (and understanding ourselves). But we do not usually seek to capture our views in a systematic framework. Nonetheless, we all have views on human nature and a range of general views on human action. Most of us would agree that people do crazy things for love, but it would be strange to call this a theory, and it is certainly not a causal theory. It is not based on clear definitions and a rigorous process of investigation. We hold lots of views of this kind and express them in different ways. When we watch a film or read a novel, we may say, “That’s not realistic. People wouldn’t do that in that situation”, or “that’s out of keeping with the character’s personality. It makes no sense for him to act like that”. It is not surprising that we make such claims and develop such views. People’s behaviour matters to us not just practically but emotionally. It is natural, therefore, that, in addition to making specific judgements, we develop general views on why people do what they do. Everyone has views on human nature, either in a developed or a less developed form. 

Psychoanalytic explanations of human action differ from everyday explanations in several important ways. Firstly, they refer to unconscious thoughts and feelings, whereas, traditionally, thoughts and feelings were by definition conscious. Secondly (and relatedly), they treat people as complex rather than unitary agents. For example, an analyst may claim that Bill consciously wanted to succeed, but unconsciously wanted to fail. Finally, they treat as significant aspects of human behaviour (slips of the tongue, mistakes, dreams, etc.) that have traditionally not been so treated. Despite these differences, psychoanalytic explanations are of the same kind as everyday explanations of human action. The correct explanation is the one that does the best job of bringing together the evidence (although in psychoanalysis, a broader evidence set is used than the usual one). As with our everyday explanations, there is no clear process for resolving disagreements. There are rival explanations, but deciding which one is correct is a matter of (individual) judgment.

What happens when we disagree underlines the difference between reason-like explanations and causal explanations. When we disagree about causes, we resolve the disagreement by collecting more objective evidence about the phenomena in question and by doing experiments which pit the different explanations against each other. In everyday disagreements about human action, we may also seek to bring in more evidence, but that evidence is likely to be just as contentious as the original situation. My friend might say: “If you had seen how Bess looked at Bob that night, you would realise that Bess would never break up with Bob over money”. To which I might respond: “But I did see Bess looking at Bob that night, and it didn’t convince me that she had deep feelings for him.” My friend might continue: “But people like us never break up over money”. To which I might respond: “Well, perhaps we do. Or perhaps Bess is not like us.” This sort of discussion may end in agreement, but often it does not. The two individuals may conclude that they will never agree on this particular case. Perhaps they generally judge human action differently. Maybe with more experience, their judgements will become more aligned. Or maybe not.

Disagreements in psychoanalysis are more complex than everyday disagreements about human action, but they have the same underlying characteristics. The two parties may disagree on the correct interpretation of any evidence and on which cases are relevantly similar. When an analyst cites some clinical material to support their position, another may dispute the interpretation the other analyst puts on the clinical material, or may deny that the new material is relevantly similar. They may reach agreement, but there is no guarantee this will happen. Each may hope that the other will change their view as they undertake more clinical work, but it is possible that further clinical work will simply confirm them in the views they currently hold.

To sum up, psychoanalysis is about making sense of what people do (and what they feel). It aims at the same kind of understanding as provided by our everyday accounts of human action, except that it takes a more systematic approach and is based on the claim that people have unconscious as well as conscious feelings. Like our everyday accounts, it is about understanding people, but it treats people as more complex than traditional approaches. Both approaches are based on a commitment to there being correct accounts of human action (just as the natural sciences are based on there being correct causal explanations of natural phenomena). Psychoanalysis is an attempt to capture the truth about human nature, but it is not a natural science. Whether it is described as a science or not depends on how the word “science” is defined, but it is not about the causes of our actions, it is about their meaning.