Thursday, September 27, 2012

Michael Lacewing - Could Psychoanalysis Be a Science?

Back in 1895, Sigmund Freud began writing his Project for a Scientific Psychology, in which he sought to establish psychoanalysis as a natural science (this paper takes a modern look at the theory). Within three decades, his inner circle had ruptured with Carl Jung (Analytical Psychology), Alfred Adler (Individual Psychology), and Otto Rank (Object-relations) all leaving to follow their own path and develop their own models of psychotherapy.

Freud's hope for a natural science rested on his belief in the biological origins of drive theory and an intuition about energy use in neurological function.

Modern psychoanalysis is greatly influenced by neuroscience, attachment theory, and infant/child development. This newer relational, intersubjective psychoanalysis (as well as neuropsychoanalysis)  is now the most scientific of the psychodynamic models currently in use.

In this article, Michael Lacewing attempts to analyze whether or not psychoanalysis is an actual science, mostly from a philosophical perspective.

Could Psychoanalysis Be a Science?

Michael Lacewing
University of London - Heythrop College

September 20, 2011

Fulford, K. W. M. et al (eds) Oxford Handbook of Philosophy and Psychiatry (Oxford: Oxford University Press) (Forthcoming)


Could psychoanalysis be a science? There are three ways of reading this question, which will structure our discussion:

1. Is psychoanalysis the kind of investigation or activity that could, logically speaking, be ‘scientific’? If we can defend a positive answer here, then it makes sense to ask:

2. Is psychoanalysis, in the form in which it has traditionally been practiced, and continues to be practiced, a science? If there are good reasons to doubt its credentials, then we might ask:

3. Is psychoanalysis able to become a science? This is a question about what is needed for the necessary transformation.

I shall argue that psychoanalysis can be a science (§1), but that the historical debate raised important challenges to its methodology, viz. confirmation bias (§2.1), suggestion (§2.2), and unsupportable causal inference (§2.3). I argue that recent developments (§3.1-2) meet these challenges, and conclude with some reflections on the interdisciplinary nature of psychoanalysis (§3.3).

Full Citation:
Lacewing, M. (2011, Sep 20). Could Psychoanalysis Be a Science? Fulford, K. W. M. et al (eds) Oxford Handbook of Philosophy and Psychiatry (Oxford: Oxford University Press) (Forthcoming). Available at SSRN:

Here is a little bit of the introduction, in which he rules out psychoanalysis as a natural science, then turns to whether or not it is a social science.

1. Could psychoanalysis (logically) be a science?
This is a question about what psychoanalysis is and what counts as science, a question about our concepts of ‘psychoanalysis’ and ‘science’. Psychoanalysis involves both the clinical encounter and the production of psychoanalytic theory. It is a mistake to restrict the meaning of ‘psychoanalysis’ to the interaction between analyst and analysand (§1.3), though this remains central to psychoanalysis. Psychoanalytic theory is a theory of the nature and functioning of the human mind, especially in relation to motives. Much of its evidential base rests in the clinical data – neurotic symptoms, dreams, present thoughts and emotions – but psychoanalysis has always gone beyond clinical data to appeal to data from other fields of enquiry (§3.3).

Psychoanalysts have been active in producing some of this data, e.g. in child development or psychiatry, and in integrating the results into new psychoanalytic theory. It is the generation of psychoanalytic knowledge that is of central interest here, and so our immediate question is ‘Are the form of psychoanalytic knowledge and the method of its generation of a kind that could (logically) qualify as scientific?’.
On the meaning of ‘science’, I shall proceed pragmatically. I assume we have a rough conception that enables us to identify paradigmatic examples, and I shall argue by comparing psychoanalysis and ‘established’ sciences, especially social psychology (§1.2) and the social sciences more broadly (§2.2).

1.1 But first, could psychoanalysis be a natural science? The most plausible defence of an affirmative answer rests on Freud’s ‘economic’ model of human psychology. The aim of his Project for a Scientific Psychology (1895) was ‘to discover what form the theory of psychical functioning will take if a quantitative line of approach, a kind of economics of nervous force, is introduced into it’ (296). Freud intended ‘to furnish a psychology that shall be a natural science’ seeking to understand human psychology in terms of a ‘conception of neuronal excitation as quantity in a state of flow’ (296), governed by biological principles of homeostasis. The model he applied, popular at the time, was that of the reflex; energy in requires energy out, to prevent energy from building up dangerously within the system, which Freud argued is experienced as pain. And so neurones and the nervous system as a whole have a tendency to divest themselves of energy.
If the economic model were the core of psychoanalysis, there would be reason to consider it a natural science. But there would also be good reason to reject it. First, many of the claims of the economic model have been superseded by neuroscience. The nervous system does not operate on a reflex model, and does not tend to divest itself of energy. Neurones generate their own energy metabolically, rather than receiving it from outside stimulation, which therefore modulates, rather than creates, nervous system activity. Sensory surfaces are not conductors, but transducers, of external energy, converting it into electrochemical impulses of negligible energy but with varying frequency – and so the nervous system cannot be swamped by energy from outside, nor can energy be trapped in it. The energy within the system is not conducted – it is not a quantity in a state of flow – but is transmitted by propagation. Finally, the quantity of energy involved bears no correlation to the psychological state of the person; the nervous system uses information, not energy, to structure its activities (see Hobson 1988, esp. pp.284-6; Holt 1965). Second, and perhaps most central to our enquiry, given this last claim, the clinical methods of psychoanalytic enquiry are inappropriate for generating knowledge of neural functioning.
Fortunately, psychoanalysis survives the demise of the economic theory. Freud repeatedly drew upon the economic model in his later psychological theorizing, e.g. he modelled psychic conflict as involving forces, understood associative links in content as involving energy pathways, talked of psychological ideas and experiences as cathecting neural networks, analysed psychic phenomena such as condensation and displacement in terms of transpositions of energy. This all needs to be reformulated just in terms of psychological content and processes. Freud sometimes approaches clinical questions in economic terms, e.g. narcissism (1914a), mourning (1917), and masochism (1924), and to the extent that psychoanalytic theories of these phenomena rest implicitly or explicitly on a mistaken conception of human beings as closed systems of fixed amounts of undissipated libido, the theories must be rethought. The theory of instincts, ‘our mythology’ as Freud put it (1933, p.94), must be translated into psychological terms, abandoned, or radically amended in the light of recent biological and neuropsychological investigations. All this can be or has been done.

These remarks leave open the possibility that psychoanalysis abuts neuroscience as a discipline, and there may be fruitful exchange (it may even be that neuroscience formulates a workable version of an ‘energy’ concept). But psychoanalysis does not qualify as a natural science in its own right. This does not rule out the possibility that it may qualify as a social science, and it is to this question we now turn.

No comments: