One way to approach the distinction between spirit and soul is to look at the different ways in which we engage with the psyche (which I assume to be the larger notion, comprising both of them). If there is an essential difference between spirit and soul, then there should be a corresponding contrast between spirit-work and soul-work.
1. In his discussion of psychological creativity in The Myth of Analysis, Hillman arrives at precisely such a contrast: “the paths of spiritual discipline and psychological development [i.e., soul-work] diverge.” (MA, 26.)
Soul-work, on Hillman’s view, requires always more than one individual soul. There is an “interdependence of psyches for the engendering of soul” (MA 25) — soul-making can’t be done with a single soul alone.
In the model situation of psychological analysis, there is an analyst and an analysand, and the interaction of their souls is what results (ideally) in the soul-building process. This, as Jung has frequently pointed out, is never only a one-way street, is not merely a therapist working on a patient’s psyche without undergoing any transformations themselves. But even the personal path of individuation (“me, working on my soul”) cannot do without other people to interact and relate with:
[…] psychological development stops in isolation; it seems unable to forgo the context of other souls. […] concern with soul immerses us in immanence: God in the soul or the soul in God, the soul in the body, the soul in the world, souls in each other or in the world-soul. (MA 26, 27)
It’s different with spirit-work, however:
Although spiritual disciplines may begin with personifications of the goal and may stress the importance of community and the master, these personifications must later be dissolved in experiences of higher abstraction and objectlessness. Persons and involvement are at best secondary. […] Besides, in spiritual disciplines even the community and the master are ultimately transpersonal abstractions. People are never as real as the spirit. The vale of the world is transcended through retreat, meditation, and prayer. (MA 26)
2. How helpful is this as a criterion for distinguishing between soul-work and spirit-work?
When we confuse soul and spirit, we may likewise confuse soul-work and spirit-work. The result may be a lot of misguided effort on the wrong path: a hardened conflict between two voices within oneself (neurosis) cannot be solved or transcended in Zen sittings or yoga exercises — for the latter focuses on spirit, whereas the problem lies in soul. Similarly mistaken would be a view of psychotherapy (which is care for the soul) on the lines of a master guiding one on one’s path (which belongs to spirit-work). We would be “confusing psychotherapy with yoga and the analyst with the master” there, as Hillman puts it (MA 26). (And obviously, such a mistake might be made by the patient or the therapist, or even both.)
Unfortunately, this latter mistake could not be avoided by Hillman’s criterion alone, for in that scenario, it still appears as if there are two souls in interrelation.
The question, under this criterion, is therefore not so much to distinguish between soul-work and spirit work, but between soul-work and the initial stages of spirit-work (which appear similar to soul-work), or perhaps between soul-work and spirit-work disguising as soul-work. (The latter formulation is perhaps not as implausible as it sounds, for from Hillman’s point of view, one of the problems with contemporary psychological culture is precisely an overabundance of spiritual offerings and a widespread confusion of them with soul-work. It is, however, also question-begging, if we’re asking whether there actually is any distinction between soul-work and spirit-work in the first place.)
3. This problem with Hillman’s criterion becomes especially pressing when we learn that the interrelatedness of souls, though necessary for soul-work, is not the point of it, either:
Human relationships may be an indispensable condition, but still the opus remains the soul. Neither relationships, nor feeling, nor any of the human context in which the psyche finds itself should be mistaken for the soul-making opus. When we make this mistake, we focus upon the instruments and means and not upon the end. Improving relationships and making connections with feeling is not at all what is meant by psychological creativity. (MA 25)
But then, we may ask, what is the difference between relationships and connectedness as “only a means” for soul-work and personalizations and the master as “only secondary” for spirit-work? Wasn’t the interrelatedness of souls the defining mark, supposed to differentiate soul-work from spirit work?
In this questioning, however, lies the clue. For spirit-work, personalizations and the master are secondary in the sense that they are not even necessary. We often see them on the spiritual path, it’s true; but that is only initially, and it’s not actually required. The path can be walked (and often has been) entirely as a solitary pursuit. Personifications, community, the master and all that are neither necessary nor sufficient for spirit work: they are entirely accidental. In contrast, human relationships and “the other soul” are (instrumentally) necessary for soul-work — although not sufficient, if Hillman’s point in the citation above holds.
The refined criterion to distinguish between spirit-work and soul-work, then, is this: if it can be done alone and (especially in the later stages) tends to be an individual, even impersonal pursuit, then it is spiritual; whereas if there is always a necessary personified component to it and we have to do it in immersion with other souls, the body, and the world, then it is soul-work.