In 1912, Alfred Wegener proposed that continents drift across the Earth's surface. His evidence included striking visual maps showing how Africa and South America's coastlines fit together like puzzle pieces. Scientists dismissed his theory for decades—not because the images were wrong, but because they seemed too persuasive, too simple, too much like seeing shapes in clouds. The visual argument felt compelling in ways that made geologists suspicious.

This tension between seeing and believing runs through all of science. We treat scientific images as transparent windows onto nature—photographs of cells, brain scans, particle tracks in cloud chambers. But every scientific image involves countless decisions about what to show, how to show it, and what to leave out. These decisions shape not just what we see, but what counts as evidence.

Understanding how scientific images persuade doesn't undermine their evidential value. It reveals something important: visual arguments work differently from verbal ones, and their power comes from sources we rarely examine. The rhetoric of scientific images operates in plain sight, which is precisely why it's so effective.

Every scientific image makes an argument, even when it appears to simply record what's there. Consider a textbook diagram of cellular mitosis. The clean, schematic representation emphasizes certain structures—chromosomes aligning, spindle fibers pulling—while abstracting away the messy, dynamic reality of an actual dividing cell. This isn't deception; it's selection, emphasis, and interpretation working together to communicate.

The sociologist Michael Lynch has shown how scientific images employ visual rhetoric through composition, color, labeling, and juxtaposition. A before-and-after comparison of tumor scans makes an argument about treatment efficacy. A graph's axis scaling determines whether a trend looks dramatic or trivial. Arrow overlays on photographs direct attention and imply causation. None of these choices are neutral.

What makes visual rhetoric particularly powerful is its apparent transparency. We've learned to scrutinize written claims—to ask about assumptions, evidence, and logical steps. But images seem to bypass this critical apparatus. They present themselves as showing rather than arguing. This creates what Latour calls immutable mobiles: representations that can travel unchanged from laboratory to journal to textbook, accumulating authority at each step.

The most persuasive scientific images often combine indexical force (they were caused by the phenomenon they represent) with iconic clarity (they look like what they depict). An electron micrograph of a virus particle has both properties—it was physically produced by electrons bouncing off that structure, and it presents a recognizable image of a sphere with surface proteins. This double authority makes questioning such images feel like questioning reality itself.

Takeaway

Scientific images argue through selection, composition, and presentation—their apparent transparency is itself a rhetorical achievement that makes their arguments harder to see and therefore harder to question.

What we see in a scientific image represents the endpoint of extensive, skilled labor that remains largely invisible in the final product. Consider an fMRI brain scan showing regions 'lighting up' during a cognitive task. Between the experimental subject lying in the scanner and that colorful brain image lies an enormous amount of work: setting statistical thresholds, correcting for multiple comparisons, choosing color scales, selecting which brain slice to display.

The historian of science Lorraine Daston distinguishes between mechanical objectivity (letting instruments record without human intervention) and trained judgment (expert interpretation that shapes what counts as a meaningful image). Scientific imaging has always required both. Early microscopists debated whether to draw exactly what they saw through the lens—including artifacts and individual variations—or to produce idealized images representing the 'true' form of a specimen.

This productive work shapes what images can show in fundamental ways. Certain phenomena become visible only when researchers know what they're looking for. The first X-ray crystallography images of DNA looked like blurry smudges; only trained eyes could extract structural information from them. The famous Photo 51 that helped reveal DNA's helical structure required Rosalind Franklin's expertise both to produce and to interpret.

Understanding image production doesn't reduce scientific images to mere social constructions. It reveals something more interesting: scientific images are achievements that require skill, judgment, and tacit knowledge. Their evidentiary power depends on this invisible labor. When we forget the work that produces scientific images, we misunderstand both their strengths and their limitations.

Takeaway

Scientific images emerge from extensive skilled labor—setting parameters, making interpretive choices, applying trained judgment—and this hidden work fundamentally shapes what the images can show and what they can prove.

Visual evidence carries a distinctive persuasive force rooted in our deep-seated intuition that seeing provides direct access to truth. This intuition isn't entirely wrong—vision is our most information-rich sense, and visual evidence often enables discoveries impossible through other means. But the power of visual persuasion can outrun its epistemic warrant.

The philosopher Ian Hacking has argued that we believe in entities we can see through microscopes not primarily because of the visual experience itself, but because we can intervene through them—using fluorescent dyes to mark structures, manipulating specimens while watching the effects. Seeing becomes believing when integrated with doing. The problem arises when visual evidence floats free of this interventionist context.

Medical imaging provides instructive examples. Brain scans showing differences between groups—patients with depression versus controls, criminals versus non-criminals—carry tremendous rhetorical force. They seem to reveal the biological basis of behavior in vivid color. Yet these images represent statistical averages, not individual diagnoses. The gap between what the image shows and what we conclude from it is a space where visual persuasion can mislead.

This doesn't mean we should distrust scientific images. It means we should understand them as what they are: sophisticated tools for producing and communicating evidence, with distinctive strengths and characteristic failure modes. The solution isn't skepticism but visual literacy—learning to read scientific images as constructed representations rather than transparent windows, while still appreciating their genuine evidential power.

Takeaway

Visual evidence persuades partly because seeing feels like direct access to truth—but this intuition can be both epistemically productive when grounded in intervention and manipulation, and misleading when images circulate detached from the practices that give them meaning.

Scientific images don't simply record reality—they construct it for particular audiences and purposes. Recognizing this doesn't diminish their value; it helps us use them better. Every powerful tool requires understanding, and scientific images are among our most powerful tools for making the invisible visible.

The next time you encounter a brain scan, an astronomical photograph, or a molecular model, consider the labor behind it. Ask what choices shaped what you're seeing. Notice how the image makes its argument. These questions don't undermine scientific evidence—they engage with it more seriously.

Visual persuasion in science isn't a problem to be solved but a phenomenon to be understood. Images will continue to play central roles in scientific communication. Understanding their rhetoric makes us better readers of science, not skeptics of it.