The Theory of the Parent-Infant Relationship—Contributions to Discussion1

I think we may all agree that it is an absolutely necessary condition for any reliable psycho-analytical theory to be based on facts which have been obtained in a setting in which transference is in existence. Transference is the basis of every analytical observation, and we may say, roughly although perhaps not quite exactly, that when there is no transference there is hardly any possibility of analytical theory. Now this is exactly the great difficulty with any theory of infancy; the relationship between mother and child is so strong, so intense, that the influence of any third object is negligible, and therefore there is hardly any possibility of transference. In order to cope with this inherent difficulty in psycho-analytical theory several attempts at solution have been made. I shall briefly recapitulate them, though not in the historical sequence of their appearance.

One is Klein's attempt at a metapsychology of early infancy which is based on an extrapolation of data obtained at a later age. There is another attempt, which perhaps is best represented by Greenacre, which is based on data obtained from biology. On the whole one may call it the theory of maturation. There is yet another attempt which is based on sociological data represented mainly by Bowlby, which may be called the ethological approach. And lastly there is the other attempt, which was historically the very first, and here because of my still unresolved transference I am rather sad that nobody mentioned the name of Sándor Ferenczi, who first called our attention to the fact that the formal elements of the transference and the whole analytic situation derive from very early infant-parent relationships. This idea, and this approach, was represented to us today by Winnicott, who stated in so many words in his paper that what he will say in it will be derived from observations of patients in the analytical situation.

Perhaps the most important lesson that we can der