Our mental models of other minds may have real souls associated with them.

Note: This is an old draft I wrote years ago, inspired by Bruce Charlton’s post on the reality of Hobbes (the stuffed tiger, not the philosopher), Douglas R. Hofstadter’s book I Am a Strange Loop, and an e-mail exchange with Seijio Arakawa. It was brought back to mind by the “literary approach” discussed at the end of my last post, where I considered the analogy of God as the “author” of the story in which we are characters. Thinking about how, in such an analogy, we could still be “real” and have free will, I recalled these old musings and thought I might as well publish them. I would probably write this somewhat differently today — with less emphasis on the brain-as-computer metaphor, for example — but I think the basic point is still potentially valid.


Consider yourself as existing wherever and whenever any other person or being thinks of you, even only occasionally.

— Robert A. Monroe, Ultimate Journey

Conscious minds are associated with human brains — and not, apparently, with the material substrate of the brain, but with patterns of brain functioning — software, not hardware. This much seems undeniable.

The precise nature of that association is much debated. Perhaps the mind just is brain software; perhaps it is created by that software; perhaps it pre-exists the brain and somehow becomes “incarnated” in it (scare quotes because it is not really into flesh as such that it enters, but into the software running on that flesh). But somehow or other — whether it is born in the brain, or whether it hath had elsewhere its setting and cometh from afar — a conscious mind apparently comes to be associated with each and every functioning human brain. We cannot be sure of this, granted, since consciousness in others is not something we can directly perceive, but it certainly appears to be the case. There is quite a bit of evidence that a mind can exist without functioning brain software; none at all that functioning brain software can exist without generating — or attracting, or being — a conscious mind.

So, whatever the reason may be, functioning brain software is the kind of thing that can be expected to have a conscious mind associated with it.


When we think about the things around us, trying to understand them or predict their behavior, the brain runs a software model or simulation of the things in question.  In most cases, there is a clear distinction between the model and the thing modeled. When my brain models the anticipated behavior of a billiard ball, for example, no matter how accurate my model may be, it can never actually be a billiard ball. It can never actually have the properties which it models, such as roundness, mass, velocity, color, etc.

When Magritte painted a picture of a pipe, he captioned it “Ceci n’est pas une pipe” — “this is not a pipe” — because it isn’t a pipe; it’s a painting, and as such has virtually nothing in common with an actual pipe. But if instead of painting a picture of a pipe, he had painted a picture of a painting, no similar caption would be appropriate. If Norman Rockwell had decided to label, Magritte-style, the various things visible in his Triple Self-Portrait, he could have written “this is not a pipe,” “this is not a man,” “this is not a mirror,” “this is not an easel,” and so on — but when it came to the self-portrait on the easel, he could not have honestly written “this is not a picture” or “this is not a self-portrait.” A picture of a picture is a picture, every bit as much as the original picture it is portraying. No treachery of images here.


In the same way, a software model of software is software, just as much as the software it is modeling. When I think about billiard balls, my thoughts are not balls. But when I think about my friend’s mind — his brain software — my thoughts themselves are brain software. And as such, they should be expected to produce (or attract, or be) a conscious mind. At least they could potentially be expected to do that if the model is sufficiently detailed, sufficiently close in structure and function to a “real” mind.


This implies that the fictional characters of novelists and the imaginary friends of children could potentially be real — not in the sense of being physically visible and tangible, but in the sense of being actual minds with their own consciousness, feelings, thoughts, and free will. These minds would run on the same hardware as the mind that dreamed them up, but they would be distinct from that mind.

Certainly many writers have reported that their characters tend to “take on a life of their own,” sometimes even seizing control of the plot, “refusing” to do what the author had originally intended to have them do and “insisting” on following some other course of action. (I don’t have any specific references for this, but it’s something I’ve read time and again in interviews with novelists and other creative writers.) If these supposedly fictional characters are in fact real minds with real free will, this could very well be a literally true.

Agency in thought is a matter of arresting the attention.

All agency begins in thought. Motor consequences may or may not follow, depending on physical conditions. It is in the mind, not in the body, that man is unconditionally free.

However, it seems that thought must always be first passive and then (potentially) active. Obviously, I cannot choose to think any particular thing unless that thing is already in my mind — unless, that is, I have already thought it. A stream of consciousness presents itself to me as a datum — as something given, not chosen — no less than do the data of my senses. Will I nill I, a succession of ideas and images keeps popping up without any agency on my part. This is the “monkey mind” of Buddhist thought, the unconscious and unchosen process which meditators consciously choose to suppress.

This suppression, whether selective or indiscriminate, of thoughts as they arise is one way of exercising agency in thought. Another is the opposite: the calling-back of a thought on the point of passing out of consciousness, the choice to arrest the automatically-wandering attention and dwell on a particular thought-datum. Yet another is to choose not to intervene, but to sit back and let the mechanical thinker that generates the stream (presumably an aspect of the brain, as opposed to the soul) do its thing.

Such choices affect the subsequent stream of consciousness. When I choose to dwell on some particular idea that has presented itself to me, the associative machinery of the monkey-mind obligingly starts churning out a stream of more-or-less related ideas. By choosing which of those, in turn, to dwell on, I can exercise a very high degree of control over my stream of consciousness — but I can never eliminate, and am indeed wholly dependent on, the element of the given. To use a metaphor from biology, the monkey-mind plays the role of “random” genetic recombination and mutation, while the conscious agent does the work of selection. With each successive generation, selective processes affect what sort of raw materials will present themselves for selection in the next generation, and thus is selection able to play a truly creative role.


A little introspection will, I think, show that this is indeed how “creative” thought operates. Set yourself a rudimentary creative task — to choose a name for a fictional character, say. You will find, once the task has been set, that a succession of candidates begins presenting itself “automatically.” In no way can you choose which particular names will pop into your mind. All you can do is watch the stream until it turns up something you like and then stop it, perhaps directing it along the way by choosing to dwell more on the more promising candidates.