What you “Get”, is What You See

Seeing the Truth

More than eighty percent of Americans have now watched the video of Renée Good being shot.

They’ve seen it from multiple angles.

They’ve watched it slowed down.

They’ve replayed it frame by frame, argued over milliseconds, freeze-framed body posture and vehicle movement.

And yet, almost nothing has changed.

People’s interpretations remain remarkably stable — and remarkably predictable. If you know someone’s political orientation before they press play, you can usually predict what they will “see” by the time the video ends.

From the left, the video is read as an unjustified killing: a masked, heavily armed federal agent escalates quickly and uses lethal force during an operation already viewed as coercive and illegitimate, the shooting understood less as an aberration than as a predictable outcome of an imposed federal presence. From the right, the same sequence is parsed in reverse. Responsibility shifts to the woman who was killed, seen as having placed herself in danger by resisting or antagonising officers engaged in lawful duties; whether she intended harm matters less than whether the agent could reasonably have perceived a threat. In that frame, the death is tragic but defensible — the consequence of a situation she failed to control.

This raises an uncomfortable question:

What if the problem isn’t that people are ignoring the evidence — but that they are genuinely seeing something different?

Seeing Isn’t Recording

We tend to think of vision as mechanical. Light enters the eye, an image forms, and the brain records what happened. Like a camera.

That’s not how vision works.

The eyes don’t deliver pictures to the brain. They deliver signals — patterns of light, contrast, motion, and colour. The brain then constructs what we experience as “seeing” by combining those signals with memory, expectation, emotion, and prior belief.

In neuroscience, perception is an act of interpretation, not reception. The brain is constantly predicting what it expects to see and updating those predictions when incoming data disagrees. When the data is ambiguous — as fast, chaotic, stressful video footage almost always is — expectation fills the gaps.

This is not a flaw in a few people. It’s how human vision works.

“Looked But Failed to See”

Motorcycle safety researchers have known this for decades.

In the UK, accident investigators identified a recurring phenomenon called LBFTS — “looked but failed to see.” Drivers involved in fatal motorcycle collisions would insist, truthfully, that they had looked directly at the road. Eye-tracking studies later confirmed it: they did look.

And yet the motorcycle never registered as a meaningful object.

More unsettling still, later research identified a related failure: “saw but forgot.” Drivers looked directly at an oncoming motorcycle, then moments later behaved as if it had never existed. The image entered the eye, reached the brain — and failed to stick in working memory long enough to influence action.

These were not reckless drivers. They were ordinary people whose brains filtered reality based on expectation: I’m looking for cars. The motorcycle didn’t fit the template.

From Roads to Bodycams

The same mechanism applies here.

When people watch footage of Renée Good’s shooting, they are not neutrally decoding pixels. They are running those images through deeply ingrained mental models:

• What police encounters usually look like

• Who is assumed to be dangerous

• What “threat” feels like

• What authority is presumed to mean

Once a brain settles on an interpretation — this was self-defence or this was unjustified violence — additional information doesn’t necessarily challenge it. Often it reinforces it.

This explains why slow-motion replays so often fail to persuade. Slowing the footage doesn’t remove bias; it gives the brain more time to justify the story it has already chosen.

This Isn’t (Just) Dishonesty

It’s tempting to conclude that people who disagree with us are lying, willfully blind, or acting in bad faith.

Sometimes they are.

But much of what we’re seeing now looks closer to perceptual bias, not moral failure. The same cognitive machinery that causes a driver to miss a motorcycle at a junction can cause a viewer to misinterpret a violent encounter — especially when politics, identity, and fear are involved.

This doesn’t absolve institutions. It doesn’t resolve accountability. And it certainly doesn’t mean “everyone’s truth is equally valid.”

It does mean that video evidence alone cannot save us.

When you asked how I’ve been here without you
I’d like to say I’ve been fine and I do
But we both know the truth is hard to come by
And if I told the truth, that’s not quite true — John Denver Some Days are Diamonds

The Harder Problem

If seeing were believing, the debate would already be over.

But vision is not neutral. Perception is not passive. And truth, once filtered through belief, is far more fragile than we like to admit.

The danger isn’t that people refuse to look. The danger is that they look — and their brains quietly decide what matters before they realize a decision has been made.

That should unsettle all of us.

Because if we can’t agree on what we’re seeing — even when we’re watching the same thing — then the real crisis isn’t misinformation.

It’s perception itself

Our eyes and the brain aren’t just passive processors of light/reflected wavelengths; rather, the brain acts more like a receiver or filter of consciousness, and what we think of as images or internal “pictures” are constructs tied to all of our past experiences and beliefs.

When humans and animals accumulate evidence over time, they are often biased. The confirmation bias, in which new evidence is given more weight when it agrees with existing beliefs, is a ubiquitous yet poorly understood example of such biases

Online misinformation isn’t only a problem of bad facts. It’s a problem of bad framing, the context in which we see events.Frames don’t just emerge; they’re built and pushed. Propaganda, influence ops, and disinformation campaigns shape how people interpret whatever evidence is put in front of them. That’s always been true, but in today’s online environment those framing contests begin within minutes of any breaking event.

In the U.S., the political right has become especially adept at this. Under the second Trump administration, the White House and federal agencies aren’t just reacting to events; they’re actively shaping the narrative to suit political goals. In this case, DHS quickly framed the killing as justified, casting the victim as a threat—even a “domestic terrorist”—while amplifying selective evidence, including through Vice President JD Vance, to portray her as the aggressor. Leaders on the left, including Minneapolis Mayor Jacob Frey, offered a competing frame: that ICE’s presence was unwanted, destabilising, and making the city less safe.

Whatever your politics, this matters. These struggles over framing aren’t a side effect of the news cycle—they’re how meaning is manufactured. Anger narrows perception and boosts certainty, which is why it’s so aggressively cultivated across the political spectrum. Long before facts are weighed, conclusions have already formed. In the end, much of what we see lives in our heads before it reaches our eyes.

Leave a comment