Political violence is rarely only about what happened. It is also about what the event is called. A thousand people can gather in the street and, depending on who is describing them, become citizens demanding rights, protesters challenging authority, an unruly crowd, religious fanatics, rioters, miscreants or a mob. The bodies are the same and the event may be the same. But the political meaning changes almost completely.
That is because language does more than describe political reality…it organizes it.
A sentence such as “Thousands of people took to the streets to demand their rights” grants the crowd a political identity. The people in it are citizens. They have grievances. They are making claims on the state and their presence in the street is part of a negotiation, however disorderly, between rulers and ruled.
Now change the sentence: “An unruly mob brought chaos to the city.” The people disappear almost immediately. A ‘mob’ does not have demands. It has impulse. It does not participate in politics, rather it threatens order. Once a crowd has been transformed into a mob, the question is no longer why those people came into the street. It is how quickly they can be removed from it.
That transformation is the essence of political narrative.
States, political parties, media institutions and powerful groups construct narratives by selecting some facts, emphasizing others and assigning names that tell audiences how the event should be understood. Over time, the language begins doing much of the political work.
Edward Herman and Noam Chomsky called one version of this process the manufacture of consent. The phrase is sometimes misunderstood as a theory of crude propaganda: a government or media organization tells people what to believe, and people obediently believe it. The actual process is usually more subtle.
Certain interpretations become normal because they are repeated. Certain terms become instinctive. Certain kinds of experts appear again and again. Certain images become representative of an entire event. A protester becomes an agitator and the agitator becomes an extremist. The extremist becomes anti-state and the anti-state figure becomes a terrorist.
By the time force is used against him, the political argument for that force may already have been made linguistically. And it is a predictable chain: an event occurs; certain facts are selected; a label is attached; the label is repeated; fear grows; an “us” and a “them” emerge; public consent follows; state action becomes easier to defend.
Few political labels perform this work as efficiently as “mob.”
The word itself carries an old suspicion of ordinary people. ‘Mob’ developed from the Latin phrase mobile vulgus, referring broadly to the fickle or excitable common people. Embedded in its history is a class judgment: the educated individual is presumed capable of reason, while the crowd is susceptible to passion.
Colonial governments found this distinction particularly useful.
In British India, peasants in revolt, workers in agitation, religious gatherings and anti-colonial demonstrators could be classified administratively as a “mob,” a “riotous crowd,” an “unlawful assembly” or a “seditious gathering.” Such terms did not merely describe the size or behavior of a gathering. They located it outside legitimate politics.
Call something a peasant rebellion and one naturally asks what the peasants were rebelling against. Call it labor resistance and the conditions of labor become relevant. Call it anti-colonial protest and colonial rule itself enters the frame. Call it a ‘mob’, and the political questions begin to vanish.
A mob, by definition, need not possess a coherent grievance. It need not have a history. Its very existence can become the problem. The state then does not have to answer the crowd. It merely has to control it.
Independence did not erase this habit of mind. Across South Asia, colonial institutions were nationalized far more quickly than colonial political vocabularies were dismantled. New flags flew above old administrative instincts. Governments continued to treat political unrest as a law-and-order problem, and crowds that did not fit recognized channels of politics remained especially vulnerable to suspicion.
Bangladesh inherited this language too. But here “mob” has acquired another function beyond delegitimizing protest. It can also obscure responsibility. That matters because many of the country's episodes of collective violence have not been as spontaneous as their initial descriptions suggested.
Reports of attacks on minorities, for example, often begin with a vocabulary like “an enraged crowd,” “angry worshippers,” “religious extremists,” “miscreants,” a “mob.” The wording creates an image of combustion. Something inflammatory happens; people become angry; violence erupts.
Then investigators begin asking questions. Who circulated the Facebook post? Who announced the protest? Who arranged the loudspeakers? Who transported people? Who identified the houses? Who repeated the rumor? Who gained possession of land later? Who weakened a political rival? Who benefited?
The crowd may have committed the violence. But the crowd may not explain it and that distinction is crucial.
Consider Nasirnagar in 2016.
The first version of the story was simple. A Facebook post had allegedly insulted Islam. Muslims became angry. A crowd attacked Hindu homes and temples. As a causal narrative, it was almost self-contained: insult, outrage, mob, attack.
Later reporting and investigation complicated that story. Questions emerged about local political rivalries, conflicts within the Awami League and the role of political actors in mobilizing people. Religious anger was undoubtedly part of the event. But it was no longer enough to explain the event. At that point, calling it only a religious mob attack became analytically inadequate.
Yet crowds do not create inflammatory Facebook accounts by themselves. They do not rent microphones. They do not arrange transport. They do not identify specific homes in advance. Those actions require people with names, intentions, connections and sometimes material interests. Once organizers exist, responsibility becomes more precise.
The same problem appears in Parbatipur, Dinajpur. After violence involving Hindu families there in 2016, a citizens' investigation committee accused a local Awami League figure, Emdadul Islam Chowdhury, of attempting to seize their land and alleged his involvement in the attacks.
The significance of the allegation went beyond one local politician. It raised a recurring question in Bangladesh: When violence looks communal, is religion always the underlying motive?
Sometimes it is. But sometimes religion is the language through which another interest operates. Land is one such interest. A minority family's religious identity can make it vulnerable, but vulnerability can be exploited for reasons that are simultaneously sectarian, economic, political and personal. If analysis stops at “religious mob,” land grabbing can disappear from view. So can property disputes, local patronage systems and private revenge.
The violence against Santal people in Gobindaganj, Gaibandha, in 2016 makes the problem even clearer.
Land was central to the conflict. A case brought by members of the Santal community named specific actors, including a local Awami League member of Parliament, government officials, a Union Parishad chairman and sugar-mill officials. Allegations concerned both the attack and the decisions surrounding the eviction. Questions later arose about police officers setting fire to Santal homes.
Describe that event as a “clash between locals and Santals,” and a large chunk of politics disappears. The state disappears. The land dispute disappears. Officials disappear. Local political authority disappears. All that remains are two groups apparently colliding with each other.
Rangpur's Thakurpara violence in 2017 followed a similar narrative pattern. The first reports emphasized religious outrage over an alleged Facebook insult, followed by an attack by an angry crowd. Police later said hatred had been spread over several days and that particular individuals were suspected of planning the violence. Some victims also alleged that driving Hindu families away from their land was part of the motive.
Not every allegation in such cases has been established in court, and political analysis should not pretend otherwise. But the legal standard of proof is not the only question worth asking. The analytical question is where inquiry stops. Does it stop at “the mob attacked”? Or does it continue?
Who produced the rumor? Who amplified it? Who called people together? Who selected the targets? Who knew in advance? Who failed to intervene? Who gained afterward?
The danger of “mob” is that it can end an investigation before the most important questions begin.
There is some convenience in collective guilt.
Let’s consider this thought process. A mob attacked. Who attacked? The mob. Who was the mob? People. Which people? Unknown. Who organized them? Unclear. Who benefited? That question rarely survives into the headline. Responsibility is first collectivized, then diluted. Everyone is implicated abstractly, which can result in no one being held meaningfully accountable.
This is one reason “mob violence” can be a misleadingly complete phrase. It appears to identify the perpetrator while actually making the perpetrator harder to see. Individuals can be prosecuted. Organizers can be questioned. Political patrons can be exposed. Police officers can be held responsible for failure or complicity. A person who seized land after an attack can be investigated.
A mob, by contrast, is temporary. It has no office. It issues no statement. It cannot be summoned next week. Once it disperses, the category remains but the people who may have directed it can disappear behind it. The essential question, then, is not whether mobs exist. Of course they do.
Spontaneous collective violence is real. Rumors can produce lynchings. Communal hatred can erupt without a centrally directed conspiracy. Crowds can commit terrible crimes because individuals within them feel protected by numbers and anonymity. The people in a crowd are morally and legally responsible for what they do. But it is equally naive to assume that every violent crowd is spontaneous.
Sometimes the mob is the perpetrator, but sometimes the mob is a weapon too. The difference matters. And the repeated refusal to distinguish between the two can itself become political.
Bangladesh's major parties are all capable of using informal power. Local Awami League figures did so during their years of dominance. BNP actors, Jamaat activists or locally influential men can do the same wherever they possess the right combination of organization, patronage and impunity. This is not fundamentally a question of one party's moral character. It is a question of how power operates where formal accountability is weak.
Religion can mobilize. Land can motivate. Political protection can enable. Rumor can trigger. A crowd can execute. All of these things can be true at once. Calling the final result a “mob attack” may be descriptively correct and politically insufficient.
The argument becomes more complicated in Bangladesh after a revolutionary-scale mass uprising.
Large political ruptures leave unstable societies behind them. The state changes faster in some places than others. Governments fall but institutions remain and old fears survive. New forms of public participation emerge before formal politics knows how to absorb them. That is why Bangladesh’s current debate about mobs should not be separated from its recent experience of mass protest.
When people faced lethal state violence and came into the streets, not everyone belonged to an organized political party. Not everyone carried a developed ideological program. And certainly not everyone spoke in the polished language of constitutional rights.
Some came out because their children, siblings, friends or neighbours were in danger and some came out because fear had turned into anger. They were not necessarily disciplined political cadres. At times they may have looked exactly like what educated observers are inclined to call an emotional crowd. But collective emotion can be political and indeed, in moments of upheaval, it can become political power.
That creates an unexplored contradiction today. The institutions of coercive power that accumulated serious allegations over many years—the police, RAB, DGFI and other security bodies—remain immensely powerful. Allegations of enforced disappearances, torture and lethal repression have made reform of these bodies one of Bangladesh's central democratic questions.
Yet institutional reform is slow and politically difficult. The state retains its laws. It retains weapons. It retains intelligence networks, prisons, administrative machinery and the authority to define threats to public order. At the same time, there is a growing temptation to delegitimize informal public gathering itself.
A demonstration organized by an established party is comprehensible. A protest attended by prominent lawyers, academics or civil-society figures is legible. A rally with tasteful placards and disciplined slogans feels recognizably political. But what about a gathering without familiar leaders? What about people whose language is crude?
What about workers who shout rather than issue statements? Farmers who do not speak the vocabulary of policy? Mothers who scream? Religious crowds whose idiom unsettles secular professionals? Poor people who do not know the etiquette of political negotiation?
Urban intellectual culture often has an aesthetic theory of legitimate dissent without admitting that it does. It prefers the protest of the “conscious citizen”: organized, articulate, photogenic, programmatic and linguistically familiar. When the people assembling do not conform to that image, the word “mob” comes easily.
That instinct carries a class judgment inherited from a much older political tradition. The articulate citizen is imagined to possess politics. The emotional crowd is imagined to possess only anger. But democratic politics cannot safely rest on such a distinction. This does not mean romanticizing crowds.
Murder is a crime. Lynching is a crime. Arson is a crime. Looting is a crime. Attacking someone because of religion is a crime. Attacking someone because of political affiliation is a crime. Destroying homes and seizing land are crimes.
No political theory of public assembly should obscure that. But neither should the prosecution of violence become a pretext for stripping every unorganized gathering of political legitimacy. Because the ability of people to assemble is not a minor democratic accessory. It is one of the few powers ordinary citizens possess that can rival concentrated institutional authority.
Individually, a citizen is weak before the state. Collectively, citizens can force the state to listen. That is precisely why the language used to describe collective action matters. If every spontaneous gathering becomes a mob, every angry protest a riot, every socially unfamiliar crowd a collection of miscreants, then language begins performing a form of political disarmament.
The state's coercive capacity remains intact and the public's capacity to stand together becomes suspect. That is an unhealthy balance in any democracy.
The most useful question in Bangladesh's debate over mobs, then, may not be whether the mob has a face. It plainly does. A crowd can contain hundreds or thousands of faces. The more important question is whether we are looking in the right direction.
In many incidents, the faces at the front are the easiest to see. They throw the stone. They break the door. They shout the slogan. Behind them may be only a handful of people. The organizer or the political patron. Or the instigator and the beneficiaries. Those are the faces that premature political language can make disappear.
If we see the thousand people in front but fail to identify the five people behind them, we may understand the fiasco without understanding the crime. And when those five people are close to political power, “mob” becomes a useful word.
Because responsibility then migrates from identifiable actors to an anonymous public.
—
Dilshana Parul is a leader of NCP