The interesting thing about this whole class of studies (including all studies directed at whether or not pornography foments callousness) is that they have an uneasy intersection with the problem of free will. If free will exists, then one of its primary places of action is in identification with other people. There is a switch where, if you go one way, you end up taking an I-Thou standpoint toward someone (or a class of someones) represented by an image, and if you go the other way, you end up taking an I-It standpoint (terms from Martin Buber, I and Thou - Wikipedia).
Can there be any tendency intrinsic in human biology that imposes a trend on how likely someone is to choose I-Thou over I-It? If so, we know little or nothing about it, except that we can see that children usually love their parents, which might suggest an automatic I-Thou pitch in childhood. That’s reasonable enough: it probably helps infant survival significantly. It’s hard to conceive that that sort of evolutionary bias extends to something so non-biological as looking at an image - but it’s not impossible. For example, there could be some ape-like level of normal inter-human affinity vs. hostility that would bear on how people respond to images of random Others. Whether this would then alter the way they perceive real people is, as far as I know, not addressed outside of the subject of pornography, thus making any trend in this area a potential confounder of any results obtained from pornography studies. Images in newcasts and even Tiktoks may in theory be no more or less alienating than pornographic images. Has any study of heterosexual porn effects ever included a control group that just looked at a similar number of photos of random female celebrities from tabloid coverage?
Now, what about cultural impositions? Could trends in culture cause people to tend to be mentally respectful of, or disrespectful of, subjects shown in images? Andrea Dworkin is famous for suggesting that the reductive nature of an image allows men, in particular, to imagine themselves in a role of subjugation over those whom they view in erotic imagery, and that this reinforces their tendency to extend the same reduction to real, live women. Male culture may further reinforce this, she believed, with typical ‘hey, hey’ wolf talk about ‘babes’ and ‘broads.’
No one who has thought about genocidal events like the Holocaust would question that culture can temporarily make large numbers of people switch from I-Thou to I-It with a group of people targeted for stigma. Reductive images like anti-Semitic cartoons may even contribute to these processes. However, if you were to do a scientific study on the pure topic, “do cartoons showing Jewish people engender or prevent anti-Semitism,” you would run into what I call the “polling problem.” In effect, you’d quite possibly be taking a poll about a temporary cultural situation rather than addressing any underlying causal question about human psychology. Cartoons showing Jews might seem causally linked to anti-Semitism and horror for five years, and then, after the defeat of the Nazis and an upsurge in sympathy towards the Jewish victims and Jewish people in general, might then switch to seeming causally linked towards sympathetic attitudes towards Jewish people. The kind of peer pressure that causes mass cultural swings toward I-It and I-Thou in attitudes among human subgroups might have no causal input from the fact that images or other simulacra appeared reflecting the trends.
Many people know that the reproducibility of psychological studies was dramatically called into question by the Reproducibility Project in 2015, as it attempted to replicate studies from 2008 and found that a large proportion were irreproducible (Reproducibility Project - Wikipedia). I was intrigued that discussion of this problem was almost entirely focused on statistical issues and only minimally on the possibility that some of the items studied might genuinely have changed, either over time, or among the populations tested. And yet, the majority of psychological studies are precisely constructed as polls, surveys of subjective responses.
Do we need to be concerned that a study of sex dolls vs. callousness (inhumane or dehumanizing attitudes) might be skewed by cultural trends? Will the same results be generated in an every-person-for-himself-and-damn-the-hindmost subculture like MAGA Americans, and in a let’s-cooperate-and-give-everyone-a-leg-up subculture like liberal Danes? And even if the study population is found to have a significant swing toward I-It perceptions of real people, associated statistically with desire for or use of sex dolls, will there be any stability in this trend over the course of time? In principle, any tested individual actually taking the improbable stance of “well, the human is just a livelier version of the doll, so let me just have my way with that livelier one, instead” (I-It) could have a crisis of conscience at any moment and say, “wait, there’s an actual human, just like me, involved in the latter case,” switching to I-Thou. Can statistical trends that last over the course of years truly be generated for the likelihood of this happening or not happening, in a way that cuts across enough cultural trends to make it a statement about the psychology of human relations?
I’d be interested to know if there’s a theoretical review paper addressing concerns about this type of study - studies that essentially propose that there are trends in inter-human identification that statistically eliminate the effect of free will, or even support its non-existence. Anyone got one on hand?