This is a response to an article in Nature which describes the findings of Mel Slater in "A Virtual Reprise of the Stanley Milgram Obedience Experiments".
The study states,
This, it seems to me and my trusty sidekick Logic, is quite silly and rather dangerous.
It has been clear for a very long time that at is very real to us. It was as much a debate to Plato and Aristotle as it is to us today. Immersive environments are interesting precisely because they offer a different kind of correspondence to reality. It is both silly and illogical to suggest that
- this test is valid because it recreates the conditions of the original study
- this test is ethical because it fails to recreate the conditions of the original study
It *is* true that the more recent test measures something different than the Milligram test. It is an interesting result. But if the Nature article and the scholarly publication are any indication, the researchers seem to completely misunderstand what it is that they're actually measuring.
Jeremy Bailenson is quoted as saying that "What Slater's research is showing is if you make your virtual reality good enough, you can go back and ask all these questions".
This is silly. This is *not* what Slater's research is showing. The significant difference between Slater's research and the theatrical simulations of the past is not the fact that these are computer simulations. The difference is that Slater's research is not a double-blind. This is freshman stuff. Don't they understand this?
This simple logical failure allows them to mischaracterize the ethical question. The ethical question with in Milgram's study did not rest in the actors who pretended to be tortured. It resided in the double-blind nature of the test and its moral/psychological impact upon those whom the test encouraged to torture other people.
By telling subjects about the test, Slater ruined the results and destroyed any useful correspondence to Milgram's study. If his test is valid, then his results seem to indicate that even in the absence of a double-blind, people take this stuff seriously. This supports the argument that Slater's test was also unethical.
But we can't really know, because Slater introduced too many new variables. If he just wanted to test the double-blind effect, he would have used actors, as did Milgram. But he also introduced a virtual reality system. So his results have to do with the nature of double-blinds, but they also have to do with the believability of virtual reality.
Bailenson seems to think that virtual reality is believable enough to give us insight into human reactions in the actual situations simulated. But this is the central issue with Milgram's study: whether simulation is ethical or not. Replacing human actors with digital actors does not change the ethical problem. If it is someday established that humans react to certain digital environments in nearly-identical ways to physical reality, then the potential for unethical testing increases, due to the greater possibilities of simulation. Would it then be possible, for example, to recreate rape in a digital environment, measure the results, and claim ethical high ground since physical rape is not perpetrated? Of course not. The Milgram study provides a case of far better simulated realism than any computer will ever be able to create. And anyone with any basic understanding of Internet culture has likely read the 1993 case study, "A Rape in Cyberspace", which suggests that even textual encounters can have a significant psychological impact, a conclusion supported by Slater's research.
The popular misconception that videogames and televisions, by merit of their nature as simulations, are effectless, is dangerous and troubling. If they were truly nonexperience, they would not be popular. Each year, hundreds of computer games put millions of people in situations even more ethically questionable than that of participants Milgram's famous study.
IMHO, Slater is trolling. It's the good ol' funding and publicity garnering standby of "let's do X -- with computers! "(in space, underwater, with six foot tall dancing cockroach hobos, etc). The latest, most depressing variant of this is the Arden Project, which has almost nothing to do with theatre or Shakespeare's plays. Why can't people obtain a basic idea of the issues they're dealing with? And why do other people fund them? Sigh.
I'm just getting warmed up. Rest and further study have resulted in such gems as "Why Christians Need to Read The Communist Manifesto". But I'll let that wait for a future date.
Also note that I had the reprise on simulation and stimulation ready to go in my blog queue several days before I came across this issue. Weird, eh?