Butler & Freud on melancholy

Butler repeatedly asks how a subject could internalize lost attachment to an object prior to the constitution of interior/exterior distinctions. I guess I'm not sure about Butler's relationship to Freud: did Freud give melancholy its founding, originary role, or did he conceive of it as one possible psychic phenomenon among many? Is the paradox Butler seizes on inherent to Freud, or is it the result of her own understanding of melancholy as the founding turn of subjectivation?
I'm intrigued by the idea that interpolation works by failure that Butler extrapolates from the structure of melancholy (197). I understand this to mean that a sort of melancholization of foreclosed attachments both forms the conscience and implies that the subject is not completely, or at least not flawlessly, constructed by pre-given social or linguistic categories. That said, I'm confused what Butler means later on the same page when she says that power itself is the withdrawn object of melancholy.

On 196 Butler writes that the withdrawn object of loss may be 'an ideal, a country, a concept of liberty, in which the loss of such ideals is compensated by the interiorized ideality of conscience,' and later on she talks about foreclosures and unspeakables (I'm something along the lines of the 'never-never' of the heterosexual's lost/loved male [I'm also thinking of the snarly 'off to never never land!' in Metallica's 'Enter Sandman,' but I don't think that anyone cares]) qualifying as withdrawn objects as well. That's probably the kind of power she has in mind, and probably also the kind of 'loss' or 'withdrawal' she has in mind.

It may help to think about 'loss' in terms of a passage where Butler actually answers the question she posed in the first sentence of your post:

'The psychological discourses that presume the topographical stability of an "internal world" and its various "parts" miss the crucial point that melancholy is precisely what interiorizes the psyche, that is, makes it possible to refer to the psyche through such topographical tropes. The turn from object to ego is the movement that makes the distinction between them possible, that marks the division, the separation or loss, that forms the ego to begin with. In this sense, the turn from object to the ego fails successfully to substitute the latter for the former, but does succeed in marking and perpetuating the partition between the two. The turn thus produces the divide between ego and object, the internal and external worlds that it appears to presume' (170).

So I think that 'loss' or 'withdrawal' basically is just the division or distinction between ego and object, in our case subject and power (e.g. notice how 'the separation or loss' is tacked on appositively to the clauses about distinction and division). To distinguish ego from object is already to have lost the object, hence the melancholic-interpellative scene of object-incorporation and identification. And while there's still a failure of that incorporation or interpellation or whatever, since the object isn't just cleanly imported and substituted fully for the ego, Butler's phrasing of the failure is sort of oddly ambivalent: she doesn't write that the turn 'fails to successfully substitute' but that it 'fails successfully,' which suggests that the 'failure' is sort of built-in and constitutive, and that the 'successful failure' is one that does involve less than a 1:1 between ego and object. Would the 'failed failure' be the case of overidentification? And I guess I'm also not 100% on from whose POV, i.e. ego's or object's, these can be said to be failures or successes to begin with.

--Guattari Hero