Tim Harford has a nice post summarizing some work on on-line interactions and how (potentially) they affect us. On some results I’m a bit meh, but not having read the studies (because, really, I have tons of other crap to do, so there is a limit to how much I can procrastinate), I can’t say much that is intelligent. Let’s just say, I’m not really at all sure that the conclusion that the net is not really good for the soul is warranted.
But, he starts out talking about Joseph Weitzenbaum’s ELIZA. The little LISP program set up like a mirroring therapist that fooled some people into all kinds of disclosure. I first encountered ELIZA in one of Doug Hofstadters books, and, if I recall right, I had to build a proto-ELIZA in one of my undergrad LISP classes.
I also used to throw the original paper at my clinical students in our theory of science seminars. The chapters (Bem & deJong) that the seminar was attached to summarized the philosophical thoughts around the early cognitive revolution and Artificial Intelligence, including discussions on Dennet’s stances (very similar to Marr’s levels), Fodors Language of Thought, the rise and fall of AI, etc. I had found, earlier, that the students were not very motivated to engage, so I hoped that this would get their attention. I included a link to a latter-day, Java emulation of ELIZA, and told them to try it out, and to go easy on it as, after all, they were all almost finished clinical psychologists, and ELIZA was just a program.
And, I think it did. I always had fun in the seminars (probably more fun than the students – we have done away with them now, so they can spend more time doing stats and methodology). But, one thing that struck me was how they anthropomorphized this program. ELIZA was unempathetic, stupid, broke easily, just a really crappy therapist, and of course had none of the important qualities that are needed for a good therapist. And, well, it is just a program of course, but it was really amazing how they imbued this thing with agency!
I tried to take that as a beginning to then try to probe the science and philosophy behind a project like ELIZA, like an instantiation of the ideas that have permeated AI and cognitive science, and cognitive psychology: the mind as a computer, the language of thought. How you can find those ideas in the anxiety and attention research which I have read quite a bit about (well, when I started, doing emotion was something more fit to clinicians, so they had data and paradigms). Try to understand them. Try to look at the limits. Look at the merits.
One of the issues Tim brings up (because it was researched) was whether it is just the posting, or if a response is required for facebook posts. Seems like the posting matters, but not necessarily the interaction . (That may need to be probed more). But, it also made me think more of how humans can imbue their surroundings with life and agency. Getting angry at chairs (well beyond the supposed animistic age at 4), punishing rivers, yelling at computers, falling in love with My Little Pony. All of Paul Blooms work on attaching essential meanings to paintings, and shirts, and lovey-animals (I still have my christening teddy-bear).
This anthropomorphizing, ascribing minds to other entities is, of course, a known (but occasionally forgotten) trouble with psychological research. Louise Barrett’s book was filled with this. Not too long ago, Joseph LeDoux suggested that we call the brain-circuits subserving emotion not emotion circuits but survival circuits, because using the folk-psychological term really adds a lot of fuzzy characteristics that will obscure the research, and, strictly speaking, we have no idea what a freezing defecating rat feels, and whether it is anything like the human conscious subjective feeling of fear. Cordelia Fine brings up the problem with calling certain variants of cognitive style found among those on the autistic spectrum as male brains, because we are liable then to attach a whole lot of cultural garbage onto the notion, and miss important phenomena. And, in the “evolutions empress” book I just got, one of the authors also commented that calling forced copulation rape, rather than something less (kinda?) loaded like, I don’t know, forced copulation (made that up, but it is descriptive) likely also brings in a great deal of cultural garbage that may make us feel either that we understand it better than we do, or reject the ideas because even thinking about it is just simply vile.
With that detour, I wanted to link into an old post of mine, where I discuss Tyler Cowen’s “Create your own economy”. He has a much cheerier view of the internet life.