Three facts come to mind:
(1) Children focus on the speaker’s intensions in order to learn the meanings of words.
(2) It is easier for children to extend the use of a word (e.g. ‘dog’) than a sign (e.g. a drawing).
(3) Novel actions or words (unlike signs) are easily learned and extended in use.
This suggests a relation between intensionality and extendibility. It is easier to extend the use of a sign when it is associated with a speaker’s intensionality than when it isn’t. But not only, it (obviously) also suggests a lack of intensionality (or at least, a lack of recognition of intensionality) in physical, external, subject-independent signs (e.g. a drawing). Children’s theory of mind easily picks up the distinction between physical and mental; it is difficult, however, to have a clear-cut criterion for borderline cases, such as texts, drawings, and works of art. Thus, initially, there is a big set of objects, which are not recognized as having any intensions. Infants might start with humans as endowed with intensions. Adults certainly recognize intensions in objects that exist independently of any particular human individual (e.g. texts, drawings, music, and more). What is the mechanism underlying this difference?
Here’s a hypothesis, a very simple, obvious, hypothesis: we start associating mental properties with human individuals (there’s a good chance that infants have the capacity to distinguish human from non-human objects), and then extend certain features to other objects (e.g. animated objects) by making them, somehow, subjects (i.e. objects with intensions). These extensions are difficult to make, but they are essential.
Another more interesting case is reading. Reading can be seen as the result of one of these extensions: i.e. extending intensionality to physical, external, subject-independent signs. This extension, however, is special. It extends to inanimate objects. Thus, intensionality is attributed indirectly. Unlike speech, gazes, and other behavior, a drawing and a text can be perceived without the presence of a speaker. This is, I guess, what is so strange about reading (e.g. texts, drawings, paintings, etc).
Through writing (e.g. texts, drawings and else) humans have found a way to put their intensions out in a more ‘independent’ manner. This trick, however, has a cost. No text, drawing, or painting is as transparent as speech, gazes and other behavior in terms of the speaker’s mental state. Texts are better seen as tracks left by some intentional system at some given time. That this is true can be seen by anyone who dares to read any piece of text that was written by ‘her’, enough time in the past. Whatever it is that we read in those cases is certainly not what we currently think. Reading ourselves looks more like reading someone else. In trying to externalize our mental states we inevitably detach them from the process that they form a part of. They are detached from whatever it is that makes them part of a mental life. Thus, it can be said, in this very simply (non-metaphysical) sense that in writing texts, whatever it is that we put is something strange to ourselves.
It might very well be, however, that we have no better way to perceive, and eventually know or believe, something about ourselves. It might be, then, that all we have is the chance to collect these tracks, hints and clues that we leave behind. We then come up with a story, and stick to it. Though, as we’ve seen, the story will always, most certainly, be about someone else.
Furthermore, that is in fact the only thing we can do. For a complete track of us from ourselves is impossible. We would fall directly (and literally) into the third man’s pit, we would need to stand above, aside, or outside of ourselves; but who would stand above that external individual perceiving us? Perhaps that’s way, as a matter of fact, we rely on others to keep track of ourselves. This will be a nice interpretation, a close one, perhaps the best one, but it will still be nothing more than the recollection of hints, tracks, and clues mentioned above.
What is the problem here? Why are our theories, believes and desires about ourselves so detached from what we are? The answer must be simple: to theorize we must first perceive. When the same organism that is supposed to be perceived theorizes at the same time that it is supposed to perceive, then all you can get is delayed theorizing about delayed perception. Perhaps that’s a fact of all theories, concerning all phenomena. It is a peculiar idiosyncratic feature of humans that they feel they should have a better, infallible theorizing of their selves. Something would be wrong if they could.