There are 2 theories of psychological symbolism and psychological symbolisation which interest me at the moment. One we owe primarily to Freud, the other primarily to Langer.
Freud's is a substitutive theory. It parallels our understanding of conventional (non-psychological) symbolism - in which one extant thing (the substitute) stands for some other extant thing (an intolerable wish or fear, for example) - except, in the substitutive psychological case, the relation (created by a 'mechanism' of displacement) is causally rather than intentionally constructed. (This makes it capable of being unconscious.) The intolerable wish/fear can now be avoided by a focus on the tolerable substitute which focus is invested with the affective charge proper to the intolerable wish/fear.
Langer's is a non-substitutive, expressive, theory. It offers 'symbolisation' as the name of the natural process of (let's-call-it) the animal soul whereby confounding experience is metabolised through its various expressions, or whereby the joyous exuberance of the individual bodies forth in spontaneous creativity. Symbolisation, that is, is the largely non-conscious expressive metabolism of such predicaments as we are sunk into, or of the ecstatic creation of new meaning in play and art. Without it the engine of the animal soul seizes up and we become flung along or frozen by life's vicissitudes. These expressions may include conventional expressions (e.g. linguistic representations that mention an intentional object of the now fully-fledged wish or fear) but may include non-conventional expressions. Yet any expressive shape may do the job: gestures, song lyrics, singing expressively without words, play scenarios, visual-artistic productions, rational emotionally-alive thoughts about the situation one is in, the imaginative self-talk of young children.
So an important difference between the two models is how much each relies on defences as the defining motors of symbolisation, and how much each relies on the idea of an extant representation being repressed and transformed. Substitutive theories rely on a notion of an already extant representation being worked over by a defence which leads, by psychologically real 'primary processes' which themselves may not be motivationally driven (i.e. they may just characterise 'the unconscious'), to the creation of a substitutive symbol. The symbol here really does take the place of something which already obtained, if only 'in the unconscious'. This 'taking the place of' is part of what characterises the particular meaning of 'symbol' at play here.
There are various significant and obvious benefits, at times, of operating within this region - i.e. of finding an expressive outlet for a feeling which outlet also constitutes a truthful self-ascription of that very feeling. Yet there appear to be disbenefits as well when that mode takes over (think of those rather annoyingly poised 'well-analysed' people who only tend to express their feelings through well-controlled self-ascriptions), disbenefits such as creative deficits, lack of real playfulness, lack of inspiring passion - in short, disbenefits that seem to amount to a lack of full humanity since one part of the soul has here hijacked its other parts. It seems to me that the substitutive model also fosters that unhealthy 'hermeneutic of suspicion' that can underlie a certain kind of psychoanalytic vision: a vision which risks inexorably imposing a normative model of the 'actual intentional object' onto others' expressive projects which are seen now as inexorably defensively motivated. To be sure, we do well to take note of, and issue with, those who unfairly take out their feelings on others. Yet many a work of, say, visual art may express feelings toward objects which are not depicted on the canvas without their creation itself being motivated defensively.
One of the risks of using self-ascription and descriptive object reference ('I am angry - with her for hurting my friend') to articulate feeling is that it can quickly become affectless. We risk falling into 'intellectualisation', i.e. into situations in which - however much one's words may count as true descriptions of situations and relationships, and as true self-ascriptions of feelings towards those situations - one's self-ascriptive words no longer count as expressions because they do not express the feelings in question. (Here I follow Finkelstein who describes unconscious states as those which cannot themselves be expressed by means of true self-ascriptions.) Situations in which, I'd say, the words no longer convey an understanding or show a knowledge of the described situations, since true understanding and knowledge do not reside merely in words but rather propagate right through, and take their shape within, the lived body's reactive dispositions - reactive dispositions which here are lacking. (A pseudo-mature person 'knows' that one ought to save drowning men (knows to say 'one ought to save a drowning man' and to draw apt inferences about such situations) but yet on encountering one walks on by: such a person does not, I think we should want to say, really enjoy moral understanding or embody moral knowledge.) Situations of quiet affect yet truthful judgement which are mistaken for genuinely coming to terms with one's predicaments.
So we (these unreflective selves who are us) often do better to express our feelings in a variety of creative and, relative to truthful self-ascription, oblique ways. This keeps us inwardly alive and prevents us from falling into a cul-de-sac of affectless (non-expressive) true self-ascriptions. Our unconsciousness (or at least: our not here expressing our feelings through self-ascriptions) is now not a problem. All that really matters is that we express ourselves creatively and don't take our feelings out on undeserving others. Relatedly we will do well, too, to not allow our rites and rituals to become too instrumentally intelligible, for this will corrode their true function for us. Our gods, too, ought not to become the objects of intelligible predications, for they surely die once that happens (caught in the Enlightenment's spotlight, for example). And whilst we need our theologians to be self-conscious of when they're issuing intelligible pronouncements and when they're talking nonsense, it would be devastating for the meaning of our faiths if they tried to substitute the former for the latter.