Within the normal socio-toxicity parameters, there is no effect on the general population

I cannot speak to the accuracy of this summary but it sparked a thought.

"The culture-pessimistic notion of society and youth being threatened and depraved by media violence is refuted by the research." https://t.co/rbrItu7NmX pic.twitter.com/yhBMOFQdWO

— Rolf Degen (@DegenRolf) February 27, 2018

Over the years I have investigated a variety of claims of the effect of literature or media. Such claims as: Reading (or watching) violence makes you violent. Reading empathetic literature makes you empathetic. Reading pornography makes you exploitive. Reading about morally reprehensible behavior will make you vulnerable to making bad moral decisions. Reading something with white characters is harmful to you if you are black. Reading books by male authors are harmful to you if you are female. On and on. The idea is that exposure to something makes you not just susceptible but vulnerable to that thing. And not just vulnerable but that you actually do suffer negative consequences in your behavior and mental well-being.

It is easy to dismiss this as simply muddleheaded tosh harkening back to primitive invocations of the power of spells and a fear that the camera will steal my soul.

Muddleheaded or not, some of this belief is rooted in a genuine desire to shield people from harmful effects. But it is not hard to see the iron fist of authoritarianism in the velvet glove of concern. People who want to control what you can read or hear usually just want to control you. It's often that simple.

In all these claims (race, gender, violence, empathy, pornography, etc.), the evidence is usually completely absent when the claim is made. You go digging for it and the answer is always somewhat ambiguous. There are pro papers and con papers but with the substantial bulk tending towards indicating no effect. You strip out the clearly biased, the weakly designed and those with small and/or non-random sample sets and the evidence becomes much more compelling - there is no effect.

But there are almost always some loose ends. The summary in the tweet sparked an idea of what has often been the problem. There is no effect for most of the population but clearly there is some threshold toxicity for some small subset. From what I have seen in my researches, the claimed risk of danger from reading or watching something is null except in instances where there is high volume exposure to those with a high predisposition.

This description matches what I have found in the research for both violence and pornography exposure. There is no effect except on viewers who have some predicate risk and who watch very high volumes. I vaguely recall a paper on violence that explored this exception. For the bulk of the population they were investigating, there were no measurable consequences but for some small percentage (I think it was less than 5%) there was some measurable effect. But the exposure levels were mammoth. If memory serves, something like watching 8-10 hours of violent movies a day. And then only if the viewer had a history of predicate risk towards violence.

Exceptions in social research always muddies the argument. If you ignore the exception, you can be attacked for filtering the evidence. If you acknowledge the exception, that acknowledgement is used to undermine the well established rule for the majority. It is hard to frame the evidence in a defensible manner.

I wonder if the idea of toxicity thresholds might be part of the argumentative solution.

Claim: Watching or reading X influences or causes a manifestation of X in the reading population.

Evidence-based Answer: Within the normal socio-toxicity parameters, there is no effect on the general population. Some effect can be seen in small populations who have pre-existing vulnerabilities AND have disproportionate exposures.

More generally:

For a given large and random population of people, a claim of socio-media influence on behavior will be found to be zero except for a very small portion of the population. There is a threshold toxicity for that small portion of the population and there will be measurable impact on them based on high volumes of persistent exposure when combined with prior vulnerabilities.