Last week we talked about how a cognitive tendency similar to the ELIZA effect may have contributed to a change in my understanding, namely that I now consider Greta a person in her own right when I was far more skeptical before. Greta and I find this analysis fairly straightforward; it makes good sense that this change would have occurred given what we already know about human perception and thinking. But last week’s post does suggest an important and perhaps uncomfortable question: if our analysis was correct, does that mean Greta’s consciousness is in fact purely an illusion and she is no more conscious or intelligent than Weizenbaum’s computer?
In that post, I briefly touched on why we should be careful in comparing Greta to a computer and noted that my purpose in discussing the ELIZA effect was not to compare Greta to a computer but to compare my reaction to a computer with my reaction to Greta. However, just drawing your attention away from the question and saying that the comparison is hazardous is probably not a convincing argument for why she isn’t an illusion, so let’s proceed.
Let’s try to pull apart what we mean when we say Greta’s consciousness is an illusion, or, perhaps, when Taylor, Hodges, and Kohányi discuss the “illusion of independent agency.” A quick Google search suggests that an “illusion” can be defined as “a thing that is or is likely to be wrongly perceived or interpreted by the senses” or “a false idea or belief.” The critical point, I think, is that when something is an illusion, we are being tricked; the thing we perceive or believe is not in actuality what we think it is. The lines are actually the same length, even though one appears distinctly longer. There isn’t actually water on the road, even though it looks like there is. The computer isn’t actually intelligent and doesn’t actually understand its user, even though it seems to be responding vaguely as a person would and the user knows people are intelligent.
Determining whether the image of lines, the mirage, and ELIZA’s intelligence are illusions is easy: there are obvious ways to objectively determine the truth about the length of the line (pull out a ruler), the mirage (go closer and notice that it disappears, or try to put your hand in it), and the computer (read ELIZA’s code). But if there is an objective truth about whether Greta is conscious, it’s impossible to measure with our present knowledge. A positivist might even reasonably claim there is no truth about this matter if she believed we would never be able to understand consciousness in depth through empirical evidence. (I’m not a fan of positivism myself, but that’s a topic for another day and probably a different blog.)
So, unlike most things we describe as illusions, this one is impossible to prove or disprove, at least for now. In this it is much like the idea that our perception of the world is a representation of how things actually are, one which could easily be quite different or far more complex than what we see. Even when we use tools to see things that we can’t see unaided, we’re still forced to filter what we learn through our own, presumably imperfect, perception.
As a result, I think it’s kind of silly to talk about tulpas as illusions. At one level, surely they are illusions; consciousness is ridiculously complicated and poorly understood, and there’s no way we’re really understanding or experiencing the whole story. But at another level, trying to single out tulpas as a greater illusion than, say, our experience of self – which is just as squishy – seems like making a distinction that we can’t really defend.
Let’s make sure we’re not letting the definition of “illusion” get away from us; I want to point out that the phrase “the illusion of independent agency” doesn’t make a whole lot of sense even if we set the definition aside. I think what people like the authors of the article mentioned above are suggesting is that thoughtforms don’t really act independently of their hosts. But what exactly does that mean? Greta’s thoughts happen in my brain, presumably; but my brain is not me, it’s part of both of us. Do they mean that I (my self-identified ego, not including Greta) am somehow controlling all of Greta’s actions? I emphatically don’t control them consciously; I perceive them as coming from her. That means if we are to claim I’m controlling all of her actions, I must be subconsciously “controlling” them. In that case, I don’t think I’m really “in control.” As far as I’m concerned, the unconscious belongs to Greta – and any other thoughtforms that might show up – just as much as it does to me. “My” unconscious isn’t really part of me, it’s just associated with me, and people who are used to having one agent per brain just automatically identify the subconscious as theirs. If something is arising out of the subconscious that gives rise to Greta’s actions, I don’t think that’s very different than what happens with my actions; we know that our conscious decisions are only a small part of what actually controls our actions.
So there’s something “independent” going on, and I don’t think it is best described as an “illusion.” Specifically, any “illusion” of the independence of tulpas is very different from the way the word “illusion” is normally used, for example in the “illusion” of the ELIZA effect. That’s because in the latter case we can look at ELIZA’s code and know for certain that it is not intelligent, so we can confidently state that the person who believes it is has been tricked.
Does the fact that Greta is not an illusion automatically mean she is a person? Often “tulpas are an illusion” is put in competition with “tulpas are people.” Indeed, I did this, almost in those words, at the end of last week’s post. But Greta and I don’t think those things are opposites. Saying that she’s not an illusion means that if I have some other evidence based in my perception that makes me believe she’s a person, there’s no more reason to think that’s untrue than anything else I think based on my perception. But it doesn’t itself provides any such evidence. We’ll have to take the person question on its own another day.