It is tempting to define a group you do not like by the worst people associated with it, but this can lead to committing the fallacy of guilt by association. To illustrate, conservative protests sometimes include people openly displaying racist symbols and this can lead leftists to conclude that all the protestors are racists. As another example, protests against Israel’s actions sometimes include people who make antisemitic statements, and this leads some people to categorize the protests as antisemitic. While this is often done in bad faith, people can sincerely make unwarranted inferences about protests from the worst people present.

Since people generally do not make their reasoning clear, it often must be reconstructed. One possible line of bad reasoning is the use of a hasty generalization. A hasty generalization occurs when a person draws a conclusion about a population based on a sample that is not large enough to adequately support the concussion. It has the following form:

Premise 1: Sample S (which is too small) is taken from population P.

Premise 2: In Sample S X% of the observed A’s are B’s.

Conclusion: X% of all A’s are B’s in Population P.

This is a fallacy because the sample is too small to warrant the inference. In the case of the protesters, inferring that most conservative protesters are racists based on some of them displaying racist symbols would be an error. Likewise, inferring that most people protesting Israel are antisemitic because some of them say antisemitic things would also be an error. At this point it is likely that someone is thinking that even if most conservative protesters are not open racists, they associate with them—thus warranting the inference that they are also guilty. Likewise, someone is probably thinking that people protesting Israel are guilty of antisemitism because of their association with antisemites. This leads us to the guilt by association fallacy.

The guilt by association fallacy has many variations but this version occurs when it is inferred that a group or individual has bad qualities because of their  (alleged) association with groups or individuals who have those qualities. The form of the fallacy is this:

Premise 1: Group or person A is associated with group or person B

Premise 2: Group or person B has (bad) qualities P, Q, R.

Conclusion: Group A has (bad) qualities P, Q, R.

The error is that the only evidence offered is the (alleged) association between the two. What is wanting is an adequate connection that justifies the inference. In the conservative protester example, the protesters might be associated with protesters displaying racist symbols, but this is not enough to warrant the conclusion that they are racists. More is needed than a mere association. The more is, as one would imagine, a matter of considerable debate: those who loath conservatives will tend to accept relatively weak evidence as supporting their biased view; those who like the protesters might be blind even to the strongest evidence. Likewise for people protesting Israel. But whatever standards are used to judge association, they must be applied consistently—whether one loathes or loves the group or person.

As noted above, people who have protested Israel have been accused of association with antisemites. But the same standards applied to conservative protesters need to be applied: to infer that because some protesters have been observed to be antisemitic then most (or all) are as well would commit the hasty generalization fallacy. Naturally, if there is evidence showing that most conservative protesters are racist or evidence showing that most (or all) people who protest Israel are antisemitic, then the fallacy would not be committed.

To infer that those protesting Israel are antisemitic because some associated with the protests are antisemitic would commit the guilt by association fallacy, just as the fallacy would be committed if one inferred that conservative protesters are racists because they are associated with racists. Obviously, if there is adequate evidence supporting these claims, then the fallacy would not be committed.