How contemporary political and media discourse redefines violence
Prologue: The Adjective Before the Act
There is a form of violence that leaves no bodies, no ruins, no photographs difficult to look at. It requires no weapons and no uniforms. It does not interrupt normality—it inhabits it. It is a quiet, administrative, almost elegant violence. It happens before anything happens. It happens in language.
In recent years, the word violent has changed its place. It once described a fact: an act, a concrete action, an observable harm. Today, with unsettling frequency, it appears before the fact, attached to people, ideas, or political positions that have not yet done anything. It no longer points to what has occurred, but to what might occur. It does not name an action, but a suspicion. The adjective arrives first. The act—if it ever arrives—no longer matters.
This shift is neither accidental nor neutral. When violence ceases to be something that is done and becomes something that is, political debate changes its nature. Discussion moves away from ideas, proposals, and consequences, and into the moral terrain: who is legitimate, who is dangerous, who deserves to be heard, and who must be contained.
This is not a denial of real violence. On the contrary. We live in a world where concrete violence—armed, criminal, terrorist, structural—has by no means disappeared. But something strange happens when that palpable violence coexists with another, purely symbolic one, which receives disproportionate attention. A hypothetical violence, anticipated and endlessly narrated, while the other is explained, contextualized, or dissolved into abstractions.
This essay does not begin as a defense of particular people, parties, or ideologies. Nor does it aim to establish a simplistic axis of “good” and “bad.” It begins from a more modest—and perhaps therefore more uncomfortable—observation: contemporary political language appears to have inverted the order of things. It qualifies first. It describes later—if at all.
In this inverted order, certain actors enter the public sphere already marked. Not by what they have done, but by what they are said to represent. Others, by contrast, appear stripped of agency, transformed into effects of context, victims of impersonal forces that explain—and sometimes excuse—their most extreme actions.
The result is not a more careful or empathetic debate. It is a profoundly asymmetrical moral field, where violence is distributed unevenly: not according to facts, but according to the narrative frame that surrounds them.
This text proposes to pause there. Not to shout louder, but to look more slowly. To examine how certain words—repeated, emptied, amplified—have stopped describing reality and have begun to pre-order it. And how, in that process, real violence can drift out of focus while imagined violence occupies the entire frame.
What follows is neither an accusation nor a closed theory. It is an exploration. Chapter by chapter, it attempts to dismantle that mechanism: how the “extremist” is manufactured, how agency is suspended for some and hyper-activated for others, and what all of this costs the very possibility of dissent without being treated as a threat.
Because before any act—before even disagreement—something decisive has already occurred: someone has been named.
Chapter I: When Violence Stops Being a Fact
For a long time, the word violence carried a relatively stable meaning. It was neither simple nor comfortable, but it was clear. Violence referred to an act that produced physical harm, direct coercion, or tangible destruction. There were identifiable victims, accountable perpetrators, visible consequences. Violence could be debated, condemned, judged. Above all, it was something that happened.
That anchoring in the fact did not make violence less complex, but it kept it within a verifiable frame. One could argue about its causes, its legitimacy, or its context, but not about its existence. A wounded body required no ideological interpretation to be recognized as such.
In contemporary political discourse, however, that anchoring has begun to erode. Violence no longer appears only in association with completed acts, but increasingly with positions, discourses, or identities. It becomes a latent property, a quality attributed prior to any action. Something that does not need to be demonstrated—it is presumed.
This shift is subtle but profound. It does not arrive all at once, nor is it openly declared. It seeps into everyday language, into headlines, into comments that appear harmless. “Violent rhetoric.” “Dangerous discourse.” “Ideas that incite.” Violence detaches itself from observable harm and migrates toward the terrain of intention—or worse, of possibility.
Under this new framework, nothing needs to have happened for something to be treated as violence. It is enough that it could happen. Or that someone claims it could lead to something undesirable. The fact becomes secondary to anticipation.
This displacement alters the basic logic of responsibility. If violence is an act, one is responsible for what one does. If violence is a prior quality, one is responsible for what one is—or for what one is said to represent. The boundary between action and thought becomes blurred, and with it a fundamental distinction weakens: the one separating real harm from disagreement.
This is not a denial that words have effects. Language matters; it shapes frames and orients behavior. But there is a crucial difference between acknowledging that power and equating language itself with violence. When that equation becomes normalized, the category stretches until it loses precision. Everything can be violent—and precisely for that reason, nothing is clearly so.
The problem is not merely semantic. When violence ceases to be a fact and becomes a floating label, political debate loses one of its few shared anchors. Discussion no longer revolves around verifiable actions, but around attributed intentions. Around imagined futures. Around risks narrated with enough repetition to feel real.
In that context, the accusation of violence no longer requires proof. It functions as a performative act: to name is enough. The accused is trapped in a permanent defensive position, forced to prove an absence—of intent, of danger, of threat—that can never be conclusively demonstrated.
Meanwhile, concrete violence—the kind that occurs without adjectives, without quotation marks, without theoretical framing—can become routine. It is explained, contextualized, dissolved into statistics or abstract structures. It does not disappear, but it loses narrative centrality. It stops shocking.
Thus a troubling paradox emerges: the more the symbolic use of the word violence expands, the harder it becomes to name real violence with precision. The term, saturated, no longer distinguishes. And when it no longer distinguishes, it no longer guides.
This chapter does not seek to impose a definitive definition, nor to romanticize a supposedly clearer past. It aims at something more modest: to recall that violence, in order to be confronted, must be recognizable. And that when language abandons facts as its point of departure, the ground is prepared for the category to be used not to describe the world, but to morally arrange it.
The next chapter will no longer treat this shift as a diffuse drift, but as a concrete tool: the adjective as a preventive weapon.
Chapter II: The Adjective as a Preventive Weapon
Once violence ceases to function primarily as a description of acts, it becomes available for another use. No longer anchored to what has occurred, the word can be deployed in advance. It turns from diagnosis into prevention—not of harm, but of legitimacy.
The adjective violent begins to operate as a warning label. It does not require evidence, only plausibility. Its purpose is not to describe behavior, but to shape perception before behavior can be evaluated. In this sense, the label does not follow reality; it prepares the ground on which reality will later be interpreted.
This preventive use of language relies on anticipation rather than observation. It frames certain positions as inherently risky, unstable, or dangerous, regardless of what they actually propose or do. Once the frame is in place, every subsequent action is read through it. Silence becomes suspicious. Moderation becomes tactical. Condemnations of violence become performative rather than sincere.
The adjetive works asymmetrically. It is rarely applied evenly across the political field. Certain actors are presumed safe until proven otherwise; others are presumed dangerous until proven harmless. The burden of proof shifts accordingly. Some speak freely and are interpreted generously. Others must constantly clarify, distance themselves, and preempt accusations that have not yet been made.
This mechanism is particularly effective because it appears reasonable. After all, who would oppose preventing violence? Who would argue against caution? The power of the label lies precisely in this moral insulation. To question its application is easily framed as irresponsibility, or worse, complicity.
Yet what is being prevented is not violence itself, but participation. The adjective functions as a gatekeeping tool. It marks certain ideas as beyond the boundaries of acceptable discourse, not because they have produced harm, but because they are said to carry a potential for harm. That potential is rarely specified in concrete terms. It remains abstract, elastic, and therefore impossible to falsify.
In this way, language performs a form of preemptive containment. Actors are not challenged on the substance of their arguments, but disqualified on the basis of projected futures. The debate shifts from what is being proposed to what might happen if these people are allowed to speak or govern. The imagined outcome replaces the actual content.
This anticipatory logic has a corrosive effect on political reasoning. It collapses the distinction between disagreement and danger. Opposition becomes suspect by default. Dissent is tolerated only insofar as it remains symbolically disarmed—stripped of intensity, ambition, or the capacity to disrupt existing arrangements.
Crucially, this process does not require censorship in the classical sense. No one needs to be silenced outright. The label does the work quietly. Once an actor is framed as “extreme,” “radical,” or “violent,” their speech remains formally allowed but substantively neutralized. Their words are heard, but no longer taken seriously. They are present, but already outside.
Over time, the preventive adjective reshapes the political landscape. It does not merely respond to threats; it creates them conceptually. By repeatedly associating certain ideas with danger, it produces a climate in which exclusion feels like protection and suspicion feels like responsibility.
What makes this mechanism especially effective is that it rarely announces itself as such. It presents itself as prudence, moderation, even care. It claims to defend democratic norms while quietly narrowing the range of what those norms are allowed to include.
In this sense, the adjective is not just descriptive. It is performative. It does not point to violence—it enacts a boundary. And once that boundary is established, crossing it becomes less a matter of action than of interpretation.
The next chapter will examine a complementary operation: how this same language that hyper-activates agency for some actors simultaneously suspends it for others, replacing responsibility with context and intention with structure.
A note on visibility, vulnerability, and selective naming
Preventive language does not operate in a vacuum. It relies on deeper assumptions about who counts as vulnerable, who counts as dangerous, and whose suffering demands narration. Nowhere is this asymmetry more visible than in how violence is gendered in public discourse.
In many societies, violence against women is treated as an exceptional moral category. It is named, foregrounded, ritualized in language. Violence against men, by contrast, is often treated as background noise: tragic, but expected; regrettable, but structurally invisible. The difference is not primarily statistical—it is narrative.
This does not mean violence against women is unreal or unimportant. It is real, and it matters. But the way it is framed reveals something deeper about how societies allocate moral attention. Certain victims are individuated, mourned, symbolically protected. Others are absorbed into aggregates, statistics, or assumptions about risk.
In countries like Mexico, this asymmetry becomes impossible to ignore. Men are vastly overrepresented among victims of homicide, disappearance, and organized violence. Yet the dominant media discourse repeatedly insists—often reflexively—that women are the primary victims of violence. The data does not disappear, but it is reframed, subordinated, or morally overwritten.
Here, language once again does its quiet work. Terms like femicide can function as analytical tools when carefully defined. But when deployed indiscriminately, they begin to operate less as classifiers and more as emotional triggers—words that pre-load interpretation, foreclose comparison, and redistribute attention regardless of empirical proportion.
The effect is not greater clarity, but moral simplification. Violence becomes legible only when it fits a preapproved victim profile. Men, meanwhile, occupy an ambiguous position: at once presumed agents of violence and disposable recipients of it. Their deaths are rarely framed as evidence of systemic failure; they are framed as outcomes—of crime, of risk, of environment.
This asymmetry is rarely discussed openly, perhaps because it cuts across ideological lines. It is easier to treat it as a moral necessity than as a cultural choice. Yet from a broader perspective, it reflects a deeper logic—one that predates contemporary politics.
Across human societies, male disposability has been a recurring pattern. In evolutionary terms, populations are sustained by reproductive continuity, not numerical symmetry. Risk, expendability, and exposure have historically been unevenly distributed. That history does not justify suffering—but it helps explain why some forms of suffering are narratively normalized while others are ritualized as intolerable.
The tragedy lies precisely there. What is normalized is not less tragic; it is merely less visible. When language repeatedly elevates some deaths while rendering others implicit, it does not reduce violence. It redistributes empathy.
And once empathy becomes selective, responsibility follows suit. Some forms of violence demand structural reckoning. Others are treated as ambient conditions—unfortunate, but unsurprising. Some victims must be named; others are expected.
This does not produce justice. It produces moral distortion.
Chapter III: Suspended Agency and Selective Victimhood
One of the quiet consequences of preventive language is that it does not distribute responsibility evenly. It assigns agency asymmetrically. Some actors are treated as fully intentional, morally saturated, and perpetually accountable. Others are progressively relieved of agency, reframed less as decision-makers than as outcomes of circumstance.
This asymmetry is rarely stated outright. It operates through implication. Certain forms of violence are immediately personalized: they have perpetrators with names, motives, and presumed intentions. Other forms are depersonalized: they are attributed to context, environment, history, or structure. In one case, agency is hyper-activated. In the other, it is suspended.
The suspension of agency is often presented as compassion. To contextualize is to understand; to understand is to humanize. But when context consistently replaces responsibility—when agency disappears entirely—the moral calculus shifts. Violence ceases to be something done by someone, and becomes something that simply happens.
This distinction matters because responsibility follows agency. Where agency is suspended, accountability thins. Where agency is inflated, suspicion hardens. The same action can be read as tragic inevitability or as moral failure, depending not on what occurred, but on who is narratively allowed to act.
Gender provides one of the clearest lenses through which to observe this mechanism.
In many contemporary narratives, women are positioned as paradigmatic victims of violence. Their suffering is individualized, named, ritualized. Men, by contrast, are overwhelmingly overrepresented in violent deaths—through homicide, organized crime, warfare, and state failure—yet their suffering is often treated as ambient. It is tragic, but unsurprising. Expected rather than interrogated.
This is not merely a discrepancy in data reporting; it is a discrepancy in moral framing.
Terms like femicide illustrate the tension. When carefully defined, such terms can clarify patterns of targeted violence. But when expanded beyond analytical boundaries, they begin to function differently. They no longer describe a specific category of crime; they pre-select the emotional register through which violence is to be understood.
In such cases, the word does not illuminate—it prioritizes. It signals which deaths demand outrage and which belong to the background conditions of reality.
In contexts like Mexico, this tension becomes stark. The vast majority of violent deaths are male. Young men, in particular, disappear, are murdered, or are absorbed into cycles of organized violence at extraordinary rates. Yet media and political discourse often returns to a singular refrain: women are the primary victims of violence.
The claim is not entirely false—but it is structurally misleading. It conflates vulnerability with visibility. It elevates one form of suffering to symbolic centrality while rendering another statistically dominant form narratively secondary.
Men, in this framework, occupy a paradoxical position. They are simultaneously presumed perpetrators and acceptable casualties. Their deaths are rarely framed as evidence of systemic collapse; they are framed as outcomes of risk, choice, or proximity to danger. Their agency is hyper-assumed in life and quietly erased in death.
This asymmetry is not accidental, nor is it purely ideological. It reflects older patterns—what might be called a logic of disposability. Across human societies, male expendability has been historically normalized. Risk-taking, exposure to violence, and sacrificial roles have disproportionately fallen on men, not because their lives are worth less in any moral sense, but because social continuity has often been organized around differential roles.
Recognizing this does not justify suffering. It contextualizes narrative behavior. Societies tend to ritualize the protection of what they perceive as reproductively or symbolically central, while naturalizing the loss of what they perceive as structurally replaceable. That perception may be archaic, but its imprint remains.
The tragedy emerges when these inherited patterns are repurposed within modern moral discourse without reflection. When language continues to sort victims according to inherited hierarchies of empathy, it reproduces blind spots. Violence against some must be named loudly to be recognized; violence against others must be extreme to even be noticed.
And here the earlier mechanism reappears. Agency is suspended where empathy is concentrated, and hyper-activated where suspicion already exists. Some actors are endlessly explained; others are endlessly warned against. Some deaths demand systemic reform; others dissolve into statistics.
This does not reduce violence. It misallocates moral attention.
A discourse that cannot acknowledge asymmetry without collapsing into accusation is a discourse that has lost its analytical flexibility. It confuses moral signaling with moral clarity. And in doing so, it risks reproducing the very injustices it claims to oppose—by deciding, in advance, whose suffering counts and whose is simply part of the landscape.
The next chapter will examine how this asymmetry is stabilized and reproduced: how media ecosystems manufacture “extremes,” flatten ideological diversity, and rely on a small set of linguistic shortcuts to keep the moral map simple—even when reality is not.
Chapter IV: The Manufacture of Extremes
Once preventive language becomes normalized, it requires a supporting environment to remain effective. Words alone do not sustain moral asymmetry; they need repetition, amplification, and simplification. This is where media ecosystems enter—not as a single actor, but as a distributed system of incentives that quietly favors certain framings over others.
Complexity does not scale well. Extremes do.
Modern media operates under conditions of speed, competition, and cognitive scarcity. Attention is fragmented, cycles are short, and narratives must be legible almost instantly. In such an environment, nuance becomes friction. Ambiguity slows transmission. The result is not deliberate distortion, but structural flattening.
Ideological diversity collapses into a small number of recognizable labels. Positions that differ profoundly in substance are grouped together under familiar moral shorthand: far-right, ultra, radical, extreme. These labels do not describe policy, structure, or institutional design. They describe distance from the accepted center, whatever that center happens to be at a given moment.
The function of these labels is not explanatory but cartographic. They draw a moral map with a clear inside and outside. Once drawn, movement across that map becomes difficult. Actors placed at the edge are read through a pre-established lens. Their differences no longer matter; their category does the work.
This is how heterogeneity disappears. A libertarian reformer, a religious conservative, a nationalist populist, and a technocratic outsider may share little beyond their challenge to existing arrangements, yet they are rendered interchangeable. Not because they are the same, but because they play the same narrative role.
The role is that of the destabilizer.
Importantly, this manufacture of extremes is not symmetrical. Certain ideological deviations are framed as risky by default, while others are framed as misguided but well-intentioned. Some forms of radicalism are treated as existential threats; others as moral excesses in need of better management. Again, agency is unevenly assigned.
The media does not need to lie to achieve this effect. Selection is enough. Headlines lead with labels before substance. Images precede explanations. Context appears later, if at all. Over time, the audience learns how to read these cues. Recognition replaces evaluation.
Once recognition is in place, the system becomes self-reinforcing. Audiences anticipate certain framings. Editors respond to expectations. Deviations feel irresponsible. The cost of resisting the shorthand rises, while the cost of repeating it falls.
In this way, “extremism” becomes less a description than a maintenance category. It preserves the intelligibility of the political field by limiting the number of morally acceptable positions. It reduces uncertainty. It tells the reader, before they engage, how worried they should be.
What is lost in this process is not only fairness, but information. When everything outside the center is compressed into the same moral container, real differences disappear. Policy becomes personality. Structure becomes tone. Outcomes become intentions.
This flattening has a predictable side effect: it shifts political debate away from consequences and toward symbolism. What matters is no longer what a proposal would do, but what it signals. Whether it reassures or alarms. Whether it aligns with or disrupts the emotional equilibrium of the narrative.
Here, the adjective once again does its quiet work. To be labeled “extreme” is not to be wrong; it is to be pre-interpreted. The debate is no longer about truth or efficacy, but about containment. The question becomes not “Is this correct?” but “Is this safe?”
And safety, once defined narratively, is difficult to contest.
The manufacture of extremes thus completes the circuit begun earlier. Preventive language marks certain actors as dangerous. Selective agency assigns responsibility unevenly. Media shorthand stabilizes the map. Together, they produce a political environment in which dissent is permitted only within carefully managed boundaries.
Outside those boundaries, speech remains formally free—but substantively inert. It circulates, but does not land. It is visible, but already disqualified.
The final irony is that this system presents itself as a defense against polarization. In practice, it deepens it. By denying legitimate difference, it pushes dissonant positions toward caricature. By refusing to engage complexity, it encourages radicalization through exclusion rather than persuasion.
The next chapter will turn to the cost of this arrangement: what happens to democratic reasoning, institutional trust, and collective decision-making when language is optimized for moral sorting rather than understanding.
Chapter V: The Cost of Missnaming
Language does not merely reflect political reality; it conditions the way decisions are made within it. When categories become blunt, when labels replace analysis, the damage is not confined to discourse. It accumulates slowly, and then all at once, in institutions, in trust, and in the capacity of societies to correct their own errors.
The first cost is epistemic. When words are used to pre-sort actors into moral categories, information loses its autonomy. Facts no longer stand on their own; they are filtered through prior classification. Evidence that confirms the label is amplified. Evidence that complicates it is discounted as anomalous, strategic, or bad faith.
Over time, this produces a degraded information environment. Not because truth disappears, but because it becomes unevenly legible. Some claims are treated as inherently credible; others require extraordinary proof. The burden of evidence shifts depending on who is speaking, not on what is being said.
This asymmetry erodes one of the few shared foundations of democratic reasoning: the assumption that arguments can be evaluated independently of their origin. When that assumption collapses, debate becomes performative. Positions harden not through persuasion, but through repetition and tribal reinforcement.
The second cost is institutional. Institutions rely on categories to function, but they also rely on flexibility. When language locks actors into predetermined moral roles, institutions lose their capacity to respond proportionally. Exceptional measures become normalized. Temporary safeguards harden into permanent mechanisms.
Preventive logic, once embedded, justifies its own expansion. If certain ideas are dangerous in principle, then extraordinary oversight becomes reasonable. If certain actors are unstable by nature, then preemptive restriction feels prudent. What begins as caution slowly turns into constraint.
This process rarely announces itself as repression. It presents as responsibility. As care. As the avoidance of risk. But risk avoidance, when untethered from concrete evidence, becomes a substitute for judgment.
The third cost is political. When dissent is consistently framed as danger, the space for legitimate opposition narrows. Actors who might otherwise operate within institutional channels are pushed outward—not necessarily toward extremism of belief, but toward extremism of posture. Exclusion produces its own feedback loop.
Those placed outside the moral boundary learn quickly that nuance offers no protection. If moderation is read as camouflage, there is little incentive to maintain it. The language that claims to prevent polarization thus actively contributes to it.
The fourth cost is ethical. Selective empathy distorts moral perception. When some forms of suffering are ritualized and others normalized, compassion becomes conditional. Tragedy is no longer tragic in itself; it must first qualify.
This is particularly visible in societies marked by chronic violence. When certain deaths are expected, they cease to provoke collective reckoning. When others are symbolically elevated, they absorb disproportionate attention. The imbalance does not reflect cruelty, but habituation. Yet habituation has moral consequences.
Over time, this hierarchy of suffering reshapes social expectations. Some lives are treated as losses; others as data. Some victims demand explanation; others demand accounting. The distinction is rarely explicit, but it is widely felt.
The final cost is corrective capacity. Societies that cannot name reality accurately struggle to respond to it effectively. Policies designed around distorted narratives misallocate resources, misidentify causes, and misjudge outcomes. Problems persist not because they are unsolvable, but because they are poorly described.
To name wrongly is not simply to offend. It is to misalign perception with reality. And when perception drifts far enough, action follows in the wrong direction.
This is why the stakes of language matter. Not because words are violent in themselves, but because they shape the conditions under which violence is understood, tolerated, or confronted. Precision is not pedantry; it is a prerequisite for responsibility.
The next and final chapter will step back from diagnosis and consequence to ask a more uncomfortable question: what would it mean to speak differently? Not more loudly, not more morally, but more carefully—without abandoning judgment, and without outsourcing it to labels.
Chapter VI: Speaking Carefully in a sea of Noise
At this point, the temptation is to offer a solution. A prescription. A new vocabulary that would fix what has gone wrong. But that impulse—the need to resolve discomfort quickly—is part of the problem this essay has been tracing.
There is no neutral language waiting to be discovered. No perfect set of words that will finally describe reality without distortion. Language is not a mirror; it is a tool. And tools shape what they touch.
What can change is not the presence of judgment, but the way it is exercised.
To speak carefully is not to speak weakly. It is not to suspend moral evaluation or retreat into relativism. It is to resist the shortcut that replaces description with categorization, and understanding with preemption. It is to accept the slower task of distinguishing where others prefer to sort.
This requires tolerating uncertainty. Allowing positions to remain undefined longer than feels comfortable. Letting actions occur before deciding what they mean. It requires separating danger from disagreement, and refusing the reassurance offered by labels that promise clarity at the cost of accuracy.
Careful speech does not deny that violence exists. It insists that violence be named where it occurs, not where it is anticipated. It insists that agency be attributed where it operates, and not suspended or inflated according to narrative convenience. It insists that empathy not be rationed according to symbolic value.
This posture is uncomfortable because it removes moral automation. It asks the speaker to judge repeatedly rather than rely on inherited frames. It asks the listener to remain open longer than instinct prefers. It offers no immediate sense of belonging.
But it restores something more important: the possibility of correction.
A society that can revise its descriptions can revise its actions. A society that confuses naming with knowing cannot. When language hardens into ritual, reality loses its ability to interrupt.
None of this guarantees better outcomes. Careful speech does not promise harmony, nor does it protect against conflict. What it does is preserve the distinction between error and malice, between threat and difference, between tragedy and inevitability.
That distinction matters because without it, dissent becomes danger, danger becomes identity, and identity becomes destiny.
The alternative is not silence, nor neutrality. It is a renewed commitment to proportionality: to letting facts weigh as much as frames, to letting actions speak louder than adjectives, and to recognizing that moral seriousness begins not with certainty, but with restraint.
This essay has not argued for trust in any particular actor or ideology. It has argued for distrust of shortcuts—especially those that flatter our sense of moral alignment while quietly narrowing our field of vision.
If there is a single claim to carry forward, it is this: language does not merely describe the world we inhabit; it helps decide which worlds remain thinkable.
To speak carefully, then, is not a matter of politeness. It is a political act in the deepest sense—not because it advances a program, but because it preserves the conditions under which programs can be meaningfully evaluated.
Before the act, there is the word. Before the word, there is the choice to name—or not to name—too quickly. That choice is never neutral. But it is still a choice.
