Your critical thinking gym
Home
The Commandments
Join the Club
Your critical thinking gym
● Propaganda technique
← from Latin de- ("removal of") + humanus ("human"): literally, the removal of human status
Strip a group of the qualities that mark them as human, and hurting them stops feeling like hurting a person.
Also called: subhumanization (the animal-imagery form) · German: Entmenschlichung
Dehumanization is the use of language or imagery that denies a person or group the qualities understood as core to being human, such as moral judgment, complex emotion, or individuality, so that harming them, excluding them, or simply not caring what happens to them stops registering as a moral problem.
TL;DR
Describing or showing a group of people as if they're animals, objects, or a disease instead of people, so hurting them stops feeling like hurting anyone at all.
Aliases
Psychologist Nick Haslam's 2006 review split the concept into two forms: animalistic dehumanization (denying "uniquely human" traits like refinement and self-control, casting a group as beast-like) and mechanistic dehumanization (denying basic human warmth and agency, casting a group as object-like or robotic).
Scope
Gets confused with
Scapegoating, Name-calling, Demonization. Precise distinctions in Section 09.
A single dehumanizing line or image is sometimes visible enough to point to, and on its own it can pass as a joke, an insult, or "just rhetoric." What keeps the score low is the third factor below: it works by building disgust and contempt, the exact feelings that switch off the part of your brain that usually asks whether something is fair to the person on the other end.
Primary bias
Reduced "mind perception" toward out-groups. A 2006 brain-imaging study by Lasana Harris and Susan Fiske found that when people viewed images of extreme out-groups already coded as disgusting in their study (such as the homeless or people using drugs), the brain region normally active when thinking about another person's mind showed less activity, while disgust-linked regions lit up instead.
Secondary biases
In-group/out-group bias: people automatically extend more trust and moral concern to whoever they categorize as "us."
Disgust-based moral contagion: once something is framed as disgusting or contaminating, people instinctively want distance from it rather than understanding of it.
Psychic numbing: the more a group is treated as a faceless mass rather than individuals, the less any one person's suffering registers emotionally.
How it activates
1. A group is repeatedly described or shown using animal, object, vermin, or disease language and imagery.
2. The brain's social-cognition response to that group weakens, while disgust-linked responses strengthen (Harris & Fiske, 2006).
3. Empathy and felt moral obligation toward the group drop, even though nothing about the group itself has actually changed.
4. Harm, exclusion, or violence against the group becomes easier to permit, ignore, or carry out without guilt (Kelman, 1973).
Susceptibility factors
Strong: real or invented threat. Competition over land, jobs, or safety makes dehumanizing framing land harder.
Strong: authority sanction. A trusted leader, broadcaster, or institution normalizing the language does most of the work.
Moderate: existing distance from the target group. The less contact you've had with real individuals in that group, the easier the label sticks.
Moderate: repetition over time. A slow drip of casual jokes and comments works better than one shocking statement, since it doesn't trigger the same alarm.
What does not predict it
A 2014 review by Haslam and Steve Loughnan found dehumanizing bias shows up in ordinary people in everyday situations, not just in extremists, and isn't reliably blocked by intelligence or education on their own. What helps more consistently is direct, sustained personal contact with members of the target group, see Section 08.
Primary purpose
Mobilize support for harm against a group, and manufacture consent for treating that harm as acceptable or even necessary.
Immediate effect
For the audience, empathy and willingness to intervene drop. For the target group, the documented effect is exclusion, fear, and loss of legal or social protection.
Long-term effect
What counts as "shocking" keeps shifting. Treatment of the target group that would have seemed extreme before the campaign starts seeming reasonable, even necessary, after enough repetition. This is part of why dehumanization tends to precede, not just accompany, organized violence.
Cui bono
Regimes, movements, or factions that benefit from justifying war, ethnic violence, or exclusionary policy, or from redirecting blame for the in-group's hardship onto a target group.
Offensive or defensive
Mainly offensive, attacking a target group's standing as people. It often doubles as a defensive cover too, recasting the in-group's own violence as self-defense against a dangerous "infestation" or "threat," rather than violence against actual people.
How it works
Setup: an out-group has already been marked as distinct from "us," usually through labeling and visible symbols. Delivery: repeated language or imagery substitutes animal, object, vermin, or disease terms for the group's humanity. Effect: empathy and moral concern toward the group erode gradually, opening space for harsher treatment that would otherwise feel unacceptable.
Preconditions
An identifiable out-group, a channel with real reach and the ability to repeat the message, and some existing distance, fear, or grievance for the framing to attach to.
When it fails
It weakens fastest against direct, ongoing personal contact with real members of the target group; decades of contact research, building on psychologist Gordon Allport's contact hypothesis, consistently find that sustained contact reduces prejudice. It can also backfire quickly if a credible, visible counter-voice objects immediately, before the framing has time to repeat and settle in.
Amplifiers
Repetition over time, official or authority sanction, crisis or wartime conditions, and visual media specifically, since images dehumanize faster and stick longer than text alone. In the social-media era, recommendation algorithms that reward high-engagement content can also amplify dehumanizing posts, since outrage and disgust both drive strong engagement.
Natural habitat
Resources needed
Low to medium. A single repeated slogan or label can start the process. What makes it durable and dangerous is sustained reach: an official broadcaster, a state institution, or a platform that keeps surfacing it.
Scale dependency
Works best at scale and with repetition over time, but a single officially sanctioned act, a law, an official broadcast, a state-published pamphlet, can do an outsized share of the work on its own.
Digital mutation
Algorithmic recommendation systems can spread and reward dehumanizing content because it tends to be highly "engaging" in the metrics platforms optimize for, generating strong reactions like outrage and disgust. Memes and short video compress dehumanizing imagery into a form that spreads faster than a written argument ever could.
First documented use
The underlying pattern, casting outsiders as a contaminant, pest, or sub-human category, recurs across many pre-modern conflicts and isn't traceable to one origin point. The most thoroughly documented, systematic modern use is Nazi Germany's "Untermensch" ("subhuman") campaign, including the SS-published photo pamphlet Der Untermensch (1942), distributed by the SS publishing house Nordland Verlag, which framed Jews, Slavic peoples, and others as biologically inferior to "Aryan" Germans.
Named and theorised by
Social psychologist Herbert Kelman gave the first influential academic account of dehumanization as a mechanism enabling mass violence, in his 1973 paper. Genocide scholar Gregory Stanton later placed it as a distinct, predictable stage in his Ten Stages of Genocide model, developed from 1987 onward and still used today by the organization Genocide Watch to monitor active conflicts.
Tradition
Spans Western academic social psychology and genocide studies (Kelman, Haslam, Stanton) and is also extensively documented as a tool of 20th-century state propaganda across multiple regimes. It isn't confined to one rhetorical school the way some IPA-named techniques are, since it shows up consistently across very different political systems.
Key evolution
Pre-mass-media: spoken rumor and religious or ethnic scapegoat narratives. Mass media era: state-run radio and print, most documented in Nazi Germany and in the RTLM radio broadcasts before and during the 1994 Rwandan genocide. Social media era: platform-amplified hate speech and meme-coded imagery, most documented in the 2017-18 violence against the Rohingya minority in Myanmar.
Cross-cultural notes
German-language scholarship has a direct cognate term, Entmenschlichung, and the post-1945 German reckoning with Nazi propaganda makes this one of the most extensively studied techniques in that tradition specifically, because of the "Untermensch" campaign's scale and documentation.
Pre-bunking note
Dedicated inoculation-style research is less developed for dehumanization specifically than for some other techniques on this site. The strongest related evidence comes from intergroup contact research: studies building on Gordon Allport's contact hypothesis consistently find that direct, sustained contact between groups reduces prejudice and undercuts dehumanizing claims. That suggests prevention here works less like a one-time lesson and more like an ongoing relationship.
Backfire risk
Yes. Repeating the exact dehumanizing label or image in order to condemn it, especially as a vivid, shocking quote, can spread and normalize the very language being criticized. Naming the pattern in general terms and pointing to its documented effects is safer than reproducing the specific slur or image itself.
Commonly paired with
Often precedes
Organized exclusion, persecution, or violence. In genocide studies it sits several stages before extermination, as part of a longer process, not as an isolated, one-off event.
Often follows
Classification and symbolization of an out-group. A group has to already be marked as distinct from "us" before dehumanizing language has anything to attach itself to.
Gets confused with
Blames a group for a problem it didn't cause. Dehumanization goes further, denying the group's basic standing as people, not just its innocence. You can scapegoat a rival nation's leaders for a hardship while still describing them in fully human terms; dehumanization is the additional step of denying they're people at all.
Attaches a single negative label to discredit one claim or person in the moment. Dehumanization is a sustained reframing of an entire group's nature, built through repetition, aimed at the group's standing as people rather than at one specific argument.
Casts a target as evil, powerful, and threatening, like a devil. Dehumanization casts a target as subhuman, weak, or contaminating, like an animal or object. Both strip moral standing, but in opposite directions: evil-and-powerful versus lesser-and-disposable.
Primary sources
Accessible reading