The authors, professors of Constitutional Law, argue that the Supreme Court’s approval of immigration stops based partly on language choice creates a distinct First Amendment harm—chilling the speech of entire multilingual communities who may choose silence over the risk of government attention…

The Supreme Court, in its shadow docket order in Noem v. Vasquez Perdomo, stayed a district court ruling that had temporarily barred federal immigration agents in the Los Angeles area from conducting roving patrol stops based upon a handful of proxy factors—including “[s]peaking Spanish or English with an accent.” A separate concurrence by Justice Brett Kavanaugh made explicit what the majority left unstated: the Fourth Amendment permits immigration agents to consider language use, ethnicity, location, and job type together to conclude that reasonable suspicion exists to conduct these stops. 

The immediate legal dispute in Noem concerned the scope of reasonable suspicion under the Fourth Amendment in the immigration context. But the case also raised a second constitutional question that has received far less attention: when the government treats language choice as a suspicion-generating factor, does it chill speech protected by the First Amendment?

Noem arose from large-scale immigration enforcement operations conducted in the summer of 2025. Agents claimed to rely on four main factors in deciding whether to stop and question individuals about immigration status: (1) apparent race or ethnicity; (2) speaking Spanish or English with an accent; (3) presence at certain locations, such as car washes, tow yards, or bus stops; and (4) the type of work an individual appeared to perform. For example, the record in the district court showed that agents stopped one plaintiff, Jason Gavidia, a Latino U.S. citizen, while he was working on his car at a tow yard in Montebello, California. They relied on the surrounding location and Gavidia’s perceived characteristics to justify the stop. In a sworn declaration, Gavidia stated that immigration officers “seem to target non-white, Spanish-speaking workers, regardless of whether they have long-standing ties to the community or lawful presence in the United States.” Concluding that the plaintiffs were likely to succeed on their Fourth Amendment claims, the district court emphasized that seizures must be grounded in individualized, articulable facts rather than generalized assumptions. After the Ninth Circuit declined to stay the district court’s ruling, the government sought emergency relief from the Supreme Court. 

In Kavanaugh’s view, the circumstances warranted a straightforward application of United States v. Brignoni-Ponce. There, the Court held that a stop based solely on the appearance of Mexican descent violated the Fourth Amendment. Here, Kavanaugh maintained, the stops in question were based upon more, and he regarded as “common sense” the proposition that immigration officers might, in determining whether to stop individuals, consider the high number of unlawful immigrants in a region, the locations where they may congregate, the types of jobs they often hold—and whether they speak English, or speak it without an accent. 

Embedded in this understanding of what are now often referred to as “Kavanaugh stops” is a separate constitutional concern. In extending the limits of the Court’s decision in Brignoni-Ponce deep into the interior of a major metropolitan area—where speaking Spanish or accented English is common among both U.S. citizens and lawful residents—Kavanaugh justified the transformation of ordinary speech into a potential liability. Alone among the factors he concluded might justify immigration stops, one’s choice of language enjoys distinct constitutional protection. And Kavanaugh stops risk chilling an otherwise protected speech interest: in many everyday urban settings, multilingual speech is commonplace. As Justice Sotomayor observed in her Noem dissent, more than 37 percent of Los Angeles County residents speak Spanish at home, and over 55 percent speak a language other than English. 

The Court has long defined “speech” as direct communication conveyed through spoken or written language. Language choice fits squarely within this understanding. Indeed, there is no plausible basis for treating communication using a language other than English as beyond the First Amendment’s reach. Simply put, regardless of the language in which a speaker chooses to communicate, they are engaged in the effort to convey ideas and information, the actions that lie at the heart of the First Amendment’s protection of free expression. 

Recognizing language choice as speech, however, is only the first step; the more difficult question is how enforcement practices that treat particular speech choices as suspicious implicate another core First Amendment interest: the willingness of individuals to speak at all. The Court has often referred to the potential “chilling effect” of government action that deters lawful expression even absent a direct prohibition. The constitutional concern is ex ante deterrence: when individuals reasonably refrain from speaking to avoid potential investigation, surveillance, or sanction. Because such deterrence suppresses protected speech before it occurs, chilling effects have been recognized as a cognizable First Amendment harm. 

This concern may be especially acute where government policies operate through discretionary enforcement, as in the immigration context. In that context, in other words, the constitutional concern is not limited to the factors that officers lawfully may consider before initiating an individual stop, but how the prospect of such consideration may shape individual behavior before any encounter occurs. If speaking Spanish or accented English increases the likelihood of being questioned or detained by immigration officers, individuals will have strong incentives to alter their speech—or to stop speaking at all. Indeed, it would not be irrational for many lawful residents to choose silence, linguistic self-policing, or withdrawal from public spaces altogether. The chilling effect arises not from any single stop, but from the predictable operation of an enforcement regime that treats language choice itself as suspicious. 

In this way, the factors that Justice Kavanaugh sanctioned as legitimate in Noem effectively burden First Amendment rights without direct censorship. The Supreme Court has recognized similar dynamics in other contexts. In NAACP v. Alabama, for instance, the Court held that compelled disclosure of membership lists violated the First Amendment because of its deterrent effect on lawful association. In Bantam Books v. Sullivan, it concluded that informal government pressure, even absent formal censorship, could induce self-censorship and chill speech. And in Watchtower Bible & Tract Society v. Stratton, the Court invalidated a municipal ordinance that subjected door-to-door canvassing to enforcement scrutiny, emphasizing that even modest exposure to law enforcement oversight can deter protected speech. Taken together, these decisions reflect a consistent First Amendment concern with government practices that discourage people from speaking or associating by making lawful activity riskier—even when those practices fall short of formal prohibition. 

Importantly, while the precise contours of First Amendment protection for individuals without lawful immigration status remain unsettled, the deterrent effects of language-based policing are not confined to them. Such practices may well burden the speech of U.S. citizens and others lawfully present who speak Spanish or English with an accent. Citizens no less than anyone else may self-censor or cease to communicate in public to avoid attracting the interest of immigration officers.

While Noem ostensibly concerned the basis for reasonable suspicion to stop an individual in the immigration enforcement context, Justice Kavanaugh’s approval of stops based in part upon language choice has exposed a distinct First Amendment cost. When language choice becomes a reason for law-enforcement attention, freedom of speech may be burdened across entire communities. Even if true, Kavanaugh’s assurance that lawful individuals would be promptly released once questioned does not mitigate that harm. The First Amendment injury occurs earlier—at the moment any individual decides that speaking in a particular language in public is not worth the risk of government attention. We should have little doubt that, in many instances and in many communities, silence will prevail and the marketplace of ideas will be that much poorer for it. 

Lauren Gearty teaches Constitutional Law and Immigration Law at New England Law | Boston, where Lawrence Friedman also teaches Constitutional Law. 

Opinions expressed in JURIST Commentary are the sole responsibility of the author and do not necessarily reflect the views of JURIST’s editors, staff, donors or the University of Pittsburgh.