Those few categories of speech that the government can regulate or punish—for instance, fraud, defamation, or incitement—are well established within our constitutional tradition. See United States v. Stevens, 559 U. S. 460, 468 (2010). Aside from these and a few other narrow exceptions, it is a fundamental principle of the First Amendment that the government may not punish or suppress speech based on disapproval of the ideas or perspectives the speech conveys. See Rosenberger v. Rector and Visitors of Univ. of Va., 515 U. S. 819, 828–829 (1995).
The Government may not insulate a law from charges of viewpoint discrimination by tying censorship to the reaction of the speaker’s audience. The Court has suggested that viewpoint discrimination occurs when the government intends to suppress a speaker’s beliefs, Reed, supra, at ___–___ (slip op., at 11–12), but viewpoint discrimination need not take that form in every instance. The danger of viewpoint discrimination is that the government is attempting to remove certain ideas or perspectives from a broader debate. That danger is all the greater if the ideas or perspectives are ones a particular audience might think offensive, at least at first hearing. An initial reaction may prompt further reflection, leading to a more reasoned, more tolerant position. Indeed, a speech burden based on audience reactions is simply government hostility and intervention in a different guise. The speech is targeted, after all, based on the government’s disapproval of the speaker’s choice of message. And it is the government itself that is attempting in this case to decide whether the relevant audience would find the speech offensive. For reasons like these, the Court’s cases have long prohibited the government from justifying a First Amendment burden by pointing to the offensiveness of the speech to be suppressed.