I will begin by outlining Hume's treatment of the experience of time. Then, I will pose a criticism (in the spirit of William James) followed by 2 related questions for discussion.
As is well known, temporality as the substance of consciousness is emphasized by Hume. He speaks of the mind as a "bundle of impressions" arranged in temporal sequence. To illustrate, he likens the mind to a "theatre, where several perceptions successively make their appearance; pass, re-pass, glide away, and mingle in an infinite variety of postures and situations." (Treatise, pp. 252f.) Such a succession of different data and items gives rise to both the "impression" and "idea" of time.
Here we do not have to deal with Hume's principle thesis concerning space and time, viz., his contention of infinite divisibility and his assertion that both space and time have an atomistic structure. We are concerned, rather, with the experience of time based on Hume's theory.
The "perceptions" (i.e., complexes of impressions and ideas) which fill the "indivisible moments of time" and "whose succession forms the duration and makes it conceivable by the mind" are held by Hume to be of a perfectly *determinate* nature. Both their qualitative and quantitative attributes have precisely determinate degrees in every respect. "Perceptions," furthermore, are conceived as *discrete* to each other. Hume repeatedly emphasizes the principle that "whatever objects are different are distinguishable and separable by thought and imagination."
Keeping this in mind, let us examine the sequence of two impressions, i.e., the experience of time passing in its simplest form. What is given at the second phase of such a sequence is an impression 'B' plus an idea 'a', the copy of that impression 'A' which occupied the first phase of the sequence. Impression 'B' and idea 'a' are different, and consequently they are separable. Their being together is extrinsic to either of them and is a mere accidental coincidence. There is no feature, no tinge in the impression that would result from its being accompanied by this idea rather than by another; the impression would remain in every respect what it is, even if the actually concomitant idea were replaced by another one. The fact that it appears in a certain context has no bearing whatever upon the nature of any "perception." The sequence in question consists then in that impression 'A' is superseded by a complex state composed of impression 'B' and idea 'a', both contemporary but with no intrinsic connection with each other. On these grounds, time passing is a sequence of perfectly definite, self-sufficient and intrinsically unrelated "perceptions" (according to Hume).
When different "perceptions" succeed each other and thus time is experienced, no new impression of the senses is aroused in addition to those which form the sequence. Nor does any impression of reflection arise. Hume explicitly stresses that by the observation and contemplation of a succession of "perceptions, no emotion in the mind ... no affection of any kind" is produced. The succession of "perceptions" is not accompanied by a *new original impression* of any kind, from which a *new original idea* (namely, that of time) might be derived. In other words, the succession of perceptions is more than a necessary condition of the experience of time ---- it is this very experience itself. Besides those "perceptions" which succeed each other, there is no "primary and distinct impression," no specific fact standing for what is called awareness of time passing.
To all this one might reply, following William James, that a "succession of impressions, in and of itself, is not an impression of succession." In fact, for a succession to be experienced as such, it is not sufficient for it only to take place. If, when a later phase of a succession is present, there were no simultaneous knowledge, in some form or other, of previous phases, consciousness would be confined to the actually present phase alone (i.e., the punctual "now") and there would be no experience of the succession, although the succession did effectively take place.
1. Is Hume's theory open to this criticism? After all, Hume's theory does provide knowledge of the previous phase at the moment of the later one (i.e. impression 'B' and idea 'a' are given simultaneously, despite their purely extrinsic connection). Consequently, at the present phase there is knowledge of the preceding one.
However, it still seems that all Hume has accounted for is the fact that one that one mental state *has succeeded* another rather than *the very succeeding* itself. That being said...
2. If the experience of time is confined to a succession of actually present "nows," with no intrinsic relation to other moments in the sequence, then how is the perception of change, succession, and (indeed) time itself even possible? To make this clear: If what Hume says is true, then how are we to account for the perception of a temporally extended object, like a melody?