However, another option remained open to her: the reverse path, which starts with the primacy of the memory-trace in Freud's thought (especially in his early sketch for a scientific psychology), and then only proceeds to Beyond the Pleasure Principle and its analysis of trauma as a breakdown in this theory of memory. Derrida, the person who made available to Caruth the language of marking, takes this path. On this reading, trauma isn't a non-experience because what has been traumatic has imprinted itself indelibly on the subject, but rather because it constitutes a break in an apparatus that constantly is laying down memory-traces which are traces, that is, lay themselves down in order to defer themselves through experience. This means that the non-experience is that of a non-memory qua trace. Thus experience is led back to the memory-trace in order to trace it. This is traumatic repetition, the return to lay down as if for the first time an experience that one supposedly already had. Caruth explains repetition as the inability to get over an experience that formed itself into a gap because it was over-full: that is, it is a memory, which for Freud is absolutely not the case.
This is a bit sketchy (and of course I'm over-simplifying Caruth... though actually not much I think), but I will revise this and outline it more.
The thing that is probably the clearest proof that Caruth is mistaken in her view, and Derrida more correct, is that for Freud the psyche is bodily. Freud (famously) says in The Ego and the Id, "the ego is first and foremost a body-ego. It is not merely a surface entity, but is in itself a projection of a surface." He also (and also famously) writes in a brilliant late note: "the psyche is extended, knows nothing about it" (in Schriften aus dem Nachlass, the last volume of the Standard Edition). And--and here is the key thing I'd like to personally suggest--also all his continual references to grounding psychoanalytic findings in biology and neuroscience, far from being just empty speculation or falsities, can be seen as meditations on this embodied state of the psyche: they indeed presuppose this to be the case in order to be speculations at all. All this noted, we can see that for Caruth, trauma remains a penetration of the bodily into the mental, the outside into the inside, while for Freud the bodily nature of the psyche actually already constitutes any "inside" or "outside." In fact, the embodied nature of the mind is precisely what allows trauma to happen to it: trauma does therefore not stem from the instant the mind is disturbed and made too aware of its bodily nature. So in Caruth we have the paradox that a theory of trauma as marking, imprinting, etc. would end up being radically unable to account for the body within trauma, i.e. would remain more idealist or subjectivist than ever.