The following situation is fictional. It is not a report about real people and does not show how often this kind of manipulation occurs. Its purpose is to separate the technical authenticity of a recording from certainty about its meaning.
Ten seconds. That is the length of the video that has just appeared in your feed.
The vertical footage was recorded through a tram window. Two men are standing on the pavement. One suddenly shoves an older man carrying a shopping bag, knocks him to the ground and points at him. The video ends. Its caption reads: “Another brutal attack in broad daylight in the city centre!”
Brain networks that rapidly process possible threats can trigger an emotional response before we have consciously analysed the scene. The amygdala is one important part of this network, not the brain’s single “alarm”. In the comments under this model post, people are already demanding punishment, and someone writes:
Because this is a model, the following paragraphs do not add newly verified facts. Each adds a different hypothetical detail to show how broader context can alter the interpretation of the same fragment.
1. The same fragment, different possible stories
First imagine adding thirty seconds. The older man had pulled out a pocket knife and threatened a woman with a pram; the second man intervened. The original fragment has not changed, but the act may now look like a defensive intervention.
Add a different hypothetical fact: the object was not a knife but a set of keys, and the older man was severely disoriented by hypoglycaemia. The moral and legal assessment reopens. Add a call to emergency medical dispatch and a possible systemic cause appears.
New information changes the meaning of the scene, not the image data in the original ten seconds. Technical authenticity alone therefore cannot establish motive, cause or responsibility.
2. Authenticity is not completeness
Debates about disinformation often focus on synthetic voices, photomontages and fabricated reports. But there is another problem alongside technical falsification: a genuine recording may be paired with an interpretation that the recording itself does not establish.
Question A: “Is this recording authentic?” (technical authenticity)
Question B: “Does it support the interpretation attached to it?” (evidential meaning)
A positive answer to the first question is not automatically an answer to the second. A real person and real words may be recorded accurately while the account of motive, cause or sequence remains wrong.
A truthful fragment removed from context may be more persuasive than a complete fabrication because part of the message survives an authenticity check. That does not mean every short recording misleads or that nothing can be learned without complete context.
3. Why a quick story can feel persuasive
Psychological research describes several tendencies that can influence judgement under incomplete information. These are not three buttons in the brain and they are not a diagnosis of every act of sharing. Their strength depends on the person, task and situation.
Availability heuristic
A scenario that is easy to recall may receive disproportionate weight when people estimate likelihood. Classic experiments support this tendency, but they did not measure the prevalence of context manipulation on today’s platforms.
Emotional contagion
Other people’s emotions can influence our own experience. It does not follow that every angry caption automatically disables critical thinking.
Motivated reasoning
In studies of political arguments, people more readily challenged information that conflicted with their existing position. Applying that result to every other setting is a plausible hypothesis, not a universal law.
4. Manipulation by subtraction
Propaganda has long worked by adding, distorting and omitting information. Digital distribution has made it exceptionally cheap to circulate a short extract detached from its original time, place and source.
- Temporal crop: show the reaction while hiding the event immediately before it.
- Spatial crop: exclude what is outside the camera’s field of view.
- Semantic isolation: detach a sentence from the condition, negation or question to which it belonged.
- Conversation crop: display one message without the preceding exchange, timing and relationship between participants.
Manipulation does not always need to add anything. Sometimes removing two seconds before and two seconds after is enough.
— Jiný Kontext editorial principle5. When context excuses nothing
Context is not a universal solvent of truth or responsibility. It may change the meaning of an act, or merely explain its circumstances without excusing it. We therefore separate four steps:
| Level | Meaning | Example |
|---|---|---|
| 1. FINDING | What was directly recorded or observed. | At 14:02 the footage shows a hand movement and a person falling. |
| 2. EVIDENTIAL TRACE | The finding assessed with its origin, integrity and other traces. | Longer footage, metadata, a medical report and independent testimony. |
| 3. INTERPRETATION | A hypothesis about cause, motive and meaning. | Was it an attack, defence, accident or medical crisis? |
| 4. CONCLUSION | A moral, expert or legal assessment under the relevant rules. | A legal conclusion after admissible evidence has been assessed. |
The problem begins when we jump from a single finding to a certain conclusion and treat the invisible intermediate steps as facts.
6. Five questions before sharing
- What exactly can I see or hear directly? Separate observation from the caption.
- Who published the recording, and where did it come from? Look for the original file and timeline.
- What happened before and after? Check a longer version if one is available.
- What remained outside the frame? Absence from the shot is not proof of absence from the scene.
- Which other explanation fits the same fragment? What evidence would distinguish between them?
7. What this article claims — and what it does not
Technical authentication confirms properties of a file or recording within the scope of the analysis. It does not by itself establish that the event is complete or that an attached interpretation is correct.
This model scenario does not determine how often genuine recordings online are misleading. The cited psychological work does not measure the prevalence of such cases.
8. What would weaken the empirical part of the claim?
It would be materially weakened by a series of preregistered experiments across different event types in which adding verified events before and after an extract almost never changed participants’ judgements compared with the fragment alone. The logical distinction between authenticity and completeness would remain, but the claim about the practical power of contextual cropping would be weaker.
9. Conclusion
The sentence “It is obvious what happened” can end an inquiry at the moment it should begin. Immediate certainty does not always come from stronger evidence. Sometimes it comes only from fewer visible questions.
Before asking whether something is true, ask one quieter question: how much relevant information can I not see yet?
— Jiný Kontext