If, as Jung says, ‘theories are the very devil’, what then counts as analytic knowledge? Elsewhere Jung suggests in similar vein that:
Initially, Jung seems to suggest that the multiplicity of theories is a temporary state of affairs, necessary until a clearer picture of the psyche's complexity emerges. However he goes on to imply that the multiplicity of theories is a necessary state of affairs due to the very nature of the psyche. And he concludes that theories do not constitute knowledge as such but are merely instruments of knowledge—the heuristic devices he refers to in the previous quote.
Clinical knowledge is therefore more like a skill than a body of factual information. It provides a way of understanding the patient in analysis but it does not provide evidence for any pre-existing, generalizable laws about the human psyche. Rather analytic theories and the supposed knowledge they contain are more like a series of imaginative constructions that have been generated by analysts over a period of more than a century as ways of understanding clinical phenomena and which therefore represent the crystallized experience of the analytic profession. They are like metaphorical maps to the ever-shifting territory of the psyche, subject to the idiosyncratic descriptions of the mapmakers and only roughly applicable to the particular psychic territory which the analyst is likely to meet. Just as those who know the territory well are less likely to need maps so, as they become more experienced, analysts are less likely to refer to theory and more to the hard-won learning from their own clinical work with patients.
....Nevertheless, I still felt that becoming an analyst would enable me to know things about the psyche that others did not know, as if it would initiate me into a form of esoteric knowledge. Gradually, though, I modified this view—it was not that I knew things that others did not, it was rather that I understood things that others did not. However, I would now qualify this still further and suggest that it is merely that analysts have a way of understanding (the psyche) that others do not but that this is not necessarily the only way. Indeed it is patently obvious that there is now a far greater plethora of analytic theories than ever there was in Jung's day.
This is troubling only to those who expect analytical psychology to conform to the criteria of science in which interpretations are a means of arriving at some kind of fundamental truth, however unattainable that may be in practice. It may be possible to create outcome measures which show that some forms of psychotherapy are more effective than others, but this is likely to be much less to do with the content of the therapist's interpretations than with the therapeutic goals which interpretation furthers. Having a language with which to understand oneself and one's relations with others is much more important than what that language happens to be.
If theories and the interpretations which they generate are more like useful tools than statements of factual knowledge, then it is, as Jung argues, a positive advantage to have many different theories and multiple potential interpretations of the same phenomena. In this regard, the clinical practice of analysis is more like literary criticism where there can never be any definitive interpretation of what a text ‘means’. On the contrary, the greater the work of art, the more variable its potential meanings. Works of art are symbolic productions which are most valuable when they have multiple, indeterminate and potentially infinite meaning (Langer 1951; Rycroft 1979). (pp. 199-201)
Warren Colman (2009). Theory as Metaphor: Clinical Knowledge and its Communication. Journal of Analytical Psychology, Vol. 54, pp. 199-215