zeogold: There's a difference between "Doesn't Islam allow pedophilia?" to "lol look at dem child rapists nuke 'em prez Trump".
The difference here is that the former sounds like a honest question, which is always a valid investigation, and supposes that a rigorous research would bring the answer. But this is not how racist discourses function. Racism requires and pre-supposes the answer.
What I call racism (broadly, independantly from the notion of biological racialism), is the attribution of trait A to a group of people which is defined by trait B, with a wrongly determinist assumption that A implies B.
For instance : If your skin is black then you are a criminal. If you are a muslim then you are an islamist. If you are homosexual then you are a pedophile. Etc.
These claims benefit from our brain's natural quest for categories, simplifications, and "economical" (effort-wise) understanding of the world. To be fair, one has to explain stuff like the reasons of the over-represented poverty in violent crimes, and the over-represented "blackness" in poverty, and the pervasive cognitive trap of "x% of A are B implies that x% of B are A". Or expose the diversity of islam and the problem of salafism. Or point out heterosexual pedophilia and the nature of most homosexual relationships. But these discussions operate in a context of honest questionning. Many people avoid this work, for many (self-serving) reasons, and are even encouraged to avoid it (demagogic propaganda). This is where racism resides.
In honest contexts (that is, outside manipulative rhetorical constructions), "a rapist has raped", is not racist. "A mexican is mexican" is not racist. "This mexican is a rapist" is not racist. "Are mexicans rapists ?" is not racist. "Mexicans are rapists" is racist. "Most mexicans are rapists" is racist. When the underlying intention is to convince people or oneself that there is a reason to assume that a random mexican is a rapist, it is racism. And there are many reasons to have this intention : To protect one's identity as member of a group who always claimed it ("my daddy/party is right"), to avoid facing the complicated issue of migrations ("anyway they are evil"), to feel superior ("at least i am not mexican"), to gather votes ("i will protect you against that enemy"), to grasp the world easily ("ah so the rapists are the mexicans are the rapists, okay, next"), etc. And, given these motives, actual factual scientific arguments miss the point. They don't weaken it. And what reinforce it is "everyday thinks that" and "i am not a bad person for indulging in these thought processes". Both are encouraged by propaganda.
You escape racism when you cease to assume that such vaguely defined large heterogeneous categories determine irrelevant traits. When you complexify it. And you encourage racism when you deliberately strengthen such essentialist beliefs through rhetorics (distored, filtered, selected, simplified, decontextualized data, etc). It's a matter of methodology, but, above all, of the underlying intention that defines it.
zeogold: These types of people would be where I draw the line, since they promote violence against these particular groups (again, I do understand what you're getting at though, that stirring up these kind of thoughts is what breeds violence).
The problem -the truly difficult problem- is where to draw the line. Because, even in the cases of genocides, there is a
long process behind it. The first steps are about shaping "common sense" around some unquestionned association of a group identity and a nasty trait (or series of nasty traits). Once this is established, you can process to violences. But, without going that far, you also favour a whole range of self-righteous self-justified discriminations, in attitudes, in sense of human responsability, in degree of empathy, in expectations and interpretations, etc. And as far as I am concerned, these are already very bad things. Which should be avoided. Which can be avoided, by discouraging the underlying racist worldviews.