I noticed, in the wake of last week’s unrest in Ferguson, several references to an exchange between Ta-Nehisi Coates and Jonathan Chait about the role of looting and riots in social progress. It began with Chait’s discussion of the implicit (perhaps, in some cases, unwitting) justification for riots advanced by a few left-wing voices. In response to these justifications, Chait wrote: “Property damage and looting impede social progress. They do so in their proximate impact (destroying the town and livelihoods of residents of Ferguson) and in their long-term impact of fostering a backlash. There is no contradiction in opposing a response to injustice that creates more injustice.”

In a longer piece about Obama’s weary response to the unrest, Coates singled out Chait’s remark: “What clearly cannot be said is that violence and nonviolence are tools, and that violence—like nonviolence—sometimes works,” he wrote. Rebutting Chait’s claim that “property damage and looting impede social progress,” Coates responded: “Taken together, property damage and looting have been the most effective tools of social progress for white people in America. They describe everything from enslavement to Jim Crow laws to lynching to red-lining.” He went on to cite the role that violence and the threat of violence played in guaranteeing abolition and civil rights, and added: “‘Property damage and looting’ is a fairly accurate description of the emancipation of black people in 1865, who only five years earlier constituted some $4 billion in property.”

If my anecdotal survey of left-wing Twitter is any indication, this retort was pretty popular. But since both Chait and Coates seem to me broadly correct, it’s doubtful that their positions are as far apart as at first they might appear. A lot depends on what is meant by “social progress,” a word that’s doing quite different work in these two pieces.

Coates is obviously correct that white Americans have historically enriched and empowered themselves at the expense of black Americans. That is a form of progress, albeit a sick one, predicated on the suffering and oppression of others. It’s progress in the same way that I progress financially by stealing your wallet, or in the same way that I progress academically by cheating on the exam and tossing your copy in the trash.

At a societal level, an entire group may profit from oppression of this kind, but Chait’s use of “social progress” seems to refer to something different. One of the specific fears Chait cites, besides the destruction of Ferguson and its residents’ livelihoods, is “backlash.” This is a tellingly democratic word, because it suggests that riots impede social progress when they impede social unity, or when they inspire a resentful, reactionary response. “Backlash,” in this sense, is political backlash: The same “law and order” response to civic unrest that helped birth the modern conservative movement, with all its submerged (or not-so-submerged) racial pathologies. The fear of backlash, in this instance, stems from a fear that marginalized voices will be de-legitimized by the opportunistic seizure of the law-and-order mantle. In other words, the worry that riots could spark a counterproductive backlash stems from a desire to maintain the place of marginal voices within public discourse.

I stress this point because the underlying concern with democratic politics and public opinion suggests a different meaning of social progress—one that is by definition inclusive. Advancement that is exclusive or oppressive, that benefits one group at another group’s expense, cannot, on this definition, be called social progress. Coates is right that violence as well as nonviolence has advanced the cause of racial equality, and that violence has been a key tool of white advancement. But this form of social progress is not the only form of social progress, and we need a vocabulary that can acknowledge that complex reality.