Modern psychology and psychiatry are very well grounded in scientific principles. Both, however, have a history in various analytical philosophies. Jung had the notion of an archetype, a universally known symbol. While these are perhaps fanciful concepts, I think that it can be interesting to try to reconcile them in terms of current knowledge.
A modern author (Knox 2001) makes the following assessment:
I suggest that although, like Freud, Jung contradicted himself at times, there is a lot of evidence that he viewed archetypes as innate predispositions with an organizing function. I think it is one of the possible models for archetypes that we need to research extensively by investigating the ways in which an understanding of 'frames' and image schemas may advance our concept. We need to gain a clearer understanding of the extent to which archetypes are 'hard-wired', genetically inherited structures or, on the other hand, self-organizing patterns that emerge as a central part of the development of meaning by the human brain as...recently proposed.
I think the arguments about genetically driven versus self-organizing patterns are fascinating, but in this question, I'd like to presuppose the latter.
If these patterns were to exist, what sensory modalities would be stored in combination? Are they exclusively visual, or is there a tight binding between the visual and auditory components of these memories? Could different memories be combinations of modular archetypal "bases" or would each be unique?
Knox, J.M. (2001). Memories, fantasies, archetypes: an exploration of some connections between cognitive science and analytical psychology. Journal of Analytical Psychology, 46:613-635.