The idea of automaticity — a notion that is often invoked in the context of affective processing — is a pretty tricky one. The issue is, of course, not limited to affective processing and is encountered in several cognitive domains (for example, word processing). Reading some of cognitive literature it feels that many (all?) of the processes that at some point were deemed automatic were shown to be capacity-limited once the system was pushed hard enough. It is in this context that I particularly like the quote by Moors and De Houwer (2006, p. 321):
“Every process is uncontrolled, efficient, unconscious, and fast, to some degree.”
In other words, an all-or-none view of automaticity is untenable, and a continuous approach is needed (as eloquently outlined by Moors and De Houwer). We thus need frameworks for understanding the continuous nature of cognitive/affective processing, for instance, as suggested originally by Norman and Bobrow (1975) and again by Nakayama and Joseph (1998).
I have briefly outlined related ideas in the context of affective processing in a recent talk at a meeting organized by Gilles Pourtois, Ernst Koster, and colleagues at the University of Ghent, Belgium.
Moors A, De Houwer J (2006) Automaticity: a theoretical and conceptual analysis. Psychol Bull 132:297-326.
Nakayama K, Joseph JS (1998) Attention, pattern recognition, and pop-out in visual search. In: The Attentive Brain (Parasuraman R, ed), pp 279-298. Cambridge: MIT Press.
Norman DA, Bobrow DG (1975) On data-limited and resource-limited processes. Cognit Psychol 7:44-64.