The important thing to note in the move from primordial image to dominant is that the innate structure, whatever it is called, is regarded as more and more powerful, to the point where it becomes actor rather than acted upon. There is a shift in Jung's view of the balance of power between pre-existing structure and personal experience. Jung was also in reaction to Freud, to psychoanalytic causality and to what remained of trauma theory, so it was therefore important to him to move away from a case history approach and to strengthen his own position in the debate concerning patient recall of childhood experiences. Briefly, Jung felt that certain primal fantasies did not arise from real experience, but were better conceived of as projected into so-called memories. Primordial images and the dominants of the collective unconscious were the sources of these later fantasies (cf. Samuels, 1982). In 1919 Jung introduced the term archetype. Any consideration of the ways in which primordial imagery is transmitted over time runs foul of the Lamarckian fallacy. As applied to psychology, this suggests that fantasies are memories of specific, prehistoric experiences and that their content is inherited from previous generations. In the same way that biologists cannot accept that acquired characteristics are inherited, it is impossible for psychologists to hold that mental imagery or other contents can be passed on in that way. However, it is perfectly reasonable to argue that, while content is not inherited, form and pattern are; the concept of archetype meets this criterion. The archetype is seen as a purely formal, skeletal concept, which is then fleshed out with imagery, ideas, motifs and so on. The archetypal form or pattern is inherited but the content is variable, subject to environmental and historical changes.
— Andrew Samuels
Samuels is pinpointing the moment the archetype hardened — the shift from primordial image, still porous to personal experience, to dominant, which acts rather than being acted upon. Something closes when Jung introduces the term in 1919: the inherited structure becomes the motor, and whatever the individual actually lived gets reclassified as projection into memory rather than memory proper. Notice what that move required. Jung needed to separate himself from Freudian causality, from the case history, from the patient's recall of what happened to them. The archetype arrives partly as a theoretical defense perimeter.
Samuels saves the concept by distinguishing form from content — the pattern is inherited, the imagery is not, which sidesteps the Lamarckian trap neatly. That distinction is defensible and probably right. But it also means the archetype, properly understood, is empty — a purely formal, skeletal structure waiting to be clothed by culture, history, environment. If that is so, then the enormous explanatory weight Jung and his successors load onto specific archetypal contents — the Great Mother, the Wise Old Man, the Hero — belongs to culture and history, not to the inherited form itself. The form guarantees that something will gather around these patterns; it does not guarantee what. That is a much more modest claim than the tradition usually makes with it, and the modesty is where the concept's actual precision lives.
Andrew Samuels·Jung and the Post-Jungians·1985