Jaynes Definitions of Consciousness
Jaynes defines 'consciousness' two different ways:
1) denotative: 'that which is introspectable'. In other words, pointing to one's own introspection and saying 'this is what I mean by consciousness'. Or asking someone to point to their own introspection and saying 'that is what I mean by consciousness'.
2) connotative: 'analog 'I' narratizing in mind-space'. This definition explains consciousness as a bootstrapping of our everyday perception through metaphor.
First, take perception: when we see a table, for instance, and say 'I see the table', Jaynes maintains that this need not be evidence of consciousness as he means it, but is simply perception. The table is an object 'out there' in physical space, and the personal pronoun 'I' is merely a placeholder indexing the person who is doing the seeing or perceiving. The world is then for us a story or narrative of perceived objects, others, and events, related to the person indexed by 'I', which I will call the 'personal I'.
The second important ingredient for definition (2) is metaphor. Jaynes claims that language and even understanding both proceed by way of metaphor. The central place of metaphor in language and cognition in general has been strongly reinforced since Jaynes, by for example Lakhoff and Johnson and their popular book Philosophy in the Flesh (1999).
Jaynes then puts together the generative power of metaphor in human understanding, with his account of perception, to arrive at his notion of consciousness. Just as we perceive the table in physical space, we 'see' metaphorically our inner states and mental world as occupying a mind-space. And just as there is a personal 'I' which perceives tables and things, so we create by the process of metaphor an 'analog I' which perceives in this mind-space.
Finally, just as our everyday world out there is a narrative of perceived things happening and interacting with our personal 'I', so through this bootstrapping process of metaphor we arrive at Jaynes's notion of consciousness as an analog 'I' narratizing in mind-space.
I find these definitions of 'consciousness' of Jaynes very satisfying. But there is I believe a property of consciousness thus defined which is under-emphasized, both by Jaynes and by others since - and that is self-reference. Jaynes himself recognized this, writing for instance that the relationship between the 'analog I' and the 'metaphor me' had 'profound problems'.
In fact, self-reference seems to me such a necessary property of consciousness, as defined above, that I think it can even be defined a third way in self-referential terms.
Jaynes occasionally uses a simple syntax to diagram his statements about consciousness, using an operator '>'. He also uses 'I' (in single quotes) to mean the analog 'I', and I without quotes to be the I of perception (what I have called above the personal I).
For instance, in writing about Bertrand Russell's example 'I see the table', Jaynes maintains that what Russell is really discussing is only the perception [I see the table]. Consciousness needs the analog 'I', whereas seeing the table is an act of perception using the personal I. So Jaynes diagrams what Russell is really discussing as not the perception [I see the table] but the act of consciousness of the perception 'I' > [I see the table].
'I' > [I see the table] is not yet self-reference, since in this formulation we have the analog 'I' on the left and the personal I on the right. But we can use Jaynes's syntax to create self-reference as follows:
What can go on the left side of the '>' operator other than the analog 'I'? So in terms of Jaynes's syntax one can then perhaps define the analog 'I' self-referentially as:
Definition: The analog 'I' is the value of the variable x that satisfies x > x.
The only solution to x > x is then x = 'I', and we arrive at another definition of Jaynes's 'consciousness', using his own syntax, as:
3) Consciousness is the self-referential relationship 'I' > 'I'.
The Errant Daughter
Jaynes himself sometimes seems to deny such a formulation when made explicitly.
In a paper on the Jaynes website, Jaynes is asked whether he really means by his definition (2) above 'self-awareness', and after some preliminary clarification he denies that he is meaning some version of my (3). In his answer he gives an example of worrying about a late night returning daughter, which he claims to be an example of consciousness ('narratizing in mind-space') but he then says: 'I canít see how that could be called self-consciousness or self-awareness.'
I think it can be. Perhaps in the initial worry about the errant daughter he is not thinking of the analog 'I' specifically, but in thinking of the issue he surely does observe his own thinking, and the analog 'I' is at least always in the periphery of this thinking, if not the center.
I think therefore that it is not just a case of ['I' worry about my daughter], which is what Jaynes maintains. (Jaynes makes clear it is the analog 'I' here, not the personal I of perception as in Russell's [I see the table]). Thus paraphrasing Jaynes's own critique of Russell, what Jaynes is really saying is 'I' > ['I' worry about my daughter], which reduces to my formulation (3).
The Benefits of Self-Reference
If Jaynes's explicit denial of (3) is ignored, and one accepts that self-reference is at the heart of Jaynes's work, which I believe it to be, then several benefits follow. Not the least of these is that Jaynes's work might then be more acceptable to current mainstream work in consciousness studies, and philosophy and psychology in general.
In my conversations with both academics in the field and others, the assertion that humans at the time of the Iliad and for all time before were not conscious, is an immediate show-stopper. One then attempts to define away most everyday meaning of 'consciousness' to arrive at Jaynes's definition. This can be avoided by asserting up front instead that humans at the time of the Iliad and before were not self-aware, or conscious of themselves being conscious. This is a tenable position, and accepted as such by everyone I have spoken to on this issue, whether they argue against it or not.
I think it would take minimal effort to restate Jaynes's work from a self-referential starting point such as my (3) above. I do not regard such a re-formulation as diluting Jaynes's position in any sense, but in fact as strengthening it and highlighting the essence.
Thanks to Brian McVeigh for his comments on the original draft of this essay.