Identity Signaling
Identity signaling is Jonah Berger and Chip Heath's finding that consumption choices are read as claims about who a person is, and that people abandon tastes when the wrong group takes them up. The signal matters more than the object, which is why a product can be killed by the wrong people liking it. The direction of the effect is well supported. Its size and generality are not, because the evidence rests on modest samples at a single university with no large preregistered replication behind it.
What it is
Berger's classroom demonstration is the cleanest way in. Tell a room that somebody drives a minivan and they will confidently infer an age, children and a set of politics. Nothing about the vehicle carries that information. The inference is doing all the work, and everyone knows the inference is being made, which is why the choice was never neutral in the first place.
Berger and Heath published the divergence work in the Journal of Consumer Research in 2007 and in the Journal of Personality and Social Psychology in 2008. The finding is that people move away from tastes associated with groups they do not want to be mistaken for, and that they do so in domains where consumption is taken to signal identity rather than mere preference.
There is a mechanism underneath which is worth crediting properly. Nick Feltovich, Rick Harbaugh and Ted To described countersignalling in the RAND Journal of Economics in 2002: signals form a hierarchy, and once a signal is widely adopted the people who were using it to mark distinction move up a rung. Casual friends signal warmth to distinguish themselves from enemies; close friends signal unfriendliness to distinguish themselves from casual friends. Kevin Simler and Robin Hanson apply that structure to taste abandonment in The Elephant in the Brain, and it is the sharpest available reading of why a taste dies when the wrong group adopts it. The abandoning group is not rejecting the object. It is moving up a rung to keep the distinction legible.