What are the major metapsychological alternatives to Kleinian theory? Here I will give a brief, philosophically motivated enumeration of the basic kinds of metapsychology that have been proposed. Metapsychologies sort themselves chiefly in terms of two issues: the relation of the clinical to the theoretical components in psychoanalytic thought, and the relation of psychoanalytic theory to the physical sciences. The possibilities may be grouped as follows:

1. The first group is comprised of metapsychologies undertaking reconstructions, to varying degrees reductive, of psychoanalytic theory in terms of concepts and methods of the physical sciences. See Maze, The meaning of behaviour, ch. 6; Gill and Pribram, Freud's ‘Project’ re-assessed; and the writings of Emmanuel Peterfreund and David Rapaport. Also included are attempts to subsume psychoanalytic theory under cognitive psychology (see 2. 11 n56).

2. Anti-metapsychological positions overlap in some respects with those just described. Here the characteristic strategy is to separate the metapsychological from the clinical components in Freud's writings, condemn the former and salvage the latter. See the papers in Gill and Holzman eds., Psychology versus metapsychology. Gill, for example, argues that the job of metapsychology is properly done by biological and neurological theory, and that metapsychology, considered as a higher-order abstraction from clinical psychology, consists only of ‘pseudo explanations’. Rubinstein sees clinical hypotheses as requiring metapsychological support, on the grounds that clinical theory, in its ‘simple’ form, consists only of ‘unproven’ presupposition or ‘dogma’, and, in its ‘extended’ form ‘poses a problem but offers no solution’. Rubinstein therefore suggests that a reformulation of psychoanalytic metapsychology based on neurophysiology, and employing a depersonalised causal language, is needed, and has yet to be produced.