Your theory is plausible sounding (hybrid between interface characteristics and signal source) but is untestable and thus worthless unless you provide a set of criteria on which to evaluate it.
For example: I theorize that a person's soul is the source of their personality. Thus, if the soul is damaged somehow, the person's personality should change.
I suggest an experiment in which people sign the bottoms of hidden documents. The control group will sign pieces of paper that are blank. The test subjects will sign papers that condemn an innocent person to death, or the cutting of funding to an orphanage. (For the sake of rigor, we should actually carry out whatever horrible act is proscribed by the signed papers)
If the "soul" theory is correct, such soul damaging activities should result in measurable personality changes to the test subjects' personalities without corresponding changes to the control group.
So yeah. I disagree with "we do not know". We "do not know" all sorts of stuff, like why gravity works or why the universe exists, but that doesn't stop us from determining things about them based on their observable characteristics. (i.e. gravity exists and is tied somehow to mass) We can also remove possibilities that fail to show evidence of existence (i.e. there is no ether in space).
We know that almost everything that makes a personality can be explained by various functions of the brain, with the only gaps being left in some of the more complex interactions between our systems of consciousness. We know which part of the brain creates emotion, which part recognizes faces, which part lets us analyze our own thoughts, and a bunch of other things that were once attributed to "the soul".