Expertise Is a Social Relation

The recognition problem

A puzzle that bothers epistemologists more than it should: someone can know something true, have good reasons for believing it, and still not count as an expert. Someone else can be wrong about nearly everything in their domain and keep expert status for years.

Expertise isn't only what's in your head. It's whether other people treat you as an expert.

This sounds like relativism. It isn't. The earth orbits the sun regardless of who believes it. But expertise — the social status of being someone whose claims carry weight — is relational all the way down. You can't be an expert alone. You need others to recognize you as one.

How recognition works

Think about what makes someone an expert in practice. Credentials help; institutional affiliation helps more. Publications, citations, awards, committee memberships all accumulate into a kind of epistemic credit score. When you speak, people listen. Your claims get the benefit of the doubt, and your mistakes get explained away as anomalies.

Strip away the markers and the same claims meet skepticism. The same evidence gets dismissed. The same arguments start to sound like the ravings of someone who's read too many books and understood too few.

This isn't conspiracy, just how social knowledge works. We can't verify everything ourselves, so we lean on trust networks, with credentials as proxies and institutions as amplifiers. The system mostly works — genuine expertise tends to accumulate genuine recognition over time. But the mapping is imperfect, and the imperfections are where it gets interesting.

The gatekeeping function

Every field has gatekeepers. Journal editors decide what gets published, department committees decide who gets hired, conference organizers decide who gets a platform, funding agencies decide whose research continues.

These gatekeepers do filter for quality, or try to. They also filter according to existing paradigms, methodological commitments, institutional relationships, and the personal networks and status hierarchies nobody likes to mention. Revolutionary work struggles partly because it threatens the very frameworks gatekeepers use to evaluate it.

This isn't necessarily bad. Most revolutionary claims are wrong, and the filter keeps charlatans and cranks from drowning a field in noise. But the same filter has systematic biases. It favors the central over the peripheral, the conventional over the novel, the credentialed over the outsider — and, above all, the already-recognized.

Expertise compounds. Early recognition brings resources, resources bring productivity, productivity brings more recognition. Those who start without it face a climb that has nothing to do with the quality of what they know.

The credentialing pipeline

How do you become an expert? First you need access to training, which tends to be expensive, slow, and geographically concentrated. Then you need to be selected by existing experts who control admissions, mentorship, and advancement. Then you produce work that satisfies the standards those experts set, and finally you need that work recognized by venues those same experts control.

At every stage, existing experts shape who becomes a new one. The pipeline reproduces itself. This isn't only nepotism, though nepotism happens; it's structural. The criteria for judging expertise are set by people who already have it, using frameworks inherited from the people before them.

Some fields are more closed than others. Academic philosophy is notoriously insular — try publishing in a top journal without the right pedigree and methodology. Software engineering is unusually open: working code gives relatively objective feedback, and credentials matter less than demonstrated ability. I build software for a living, and the difference is real, but even there recognition hierarchies form. Who gets invited to speak. Whose opinions shape the discourse. Whose work gets cited.

Knowledge without recognition

What happens to knowledge that lives outside recognized channels? It doesn't disappear. It circulates in different networks, dismissed by official experts and embraced by communities that don't trust those experts. Sometimes it's wrong, sometimes it's right, and either way it lacks the legitimacy that would let it enter mainstream discourse.

This is where things get dangerous. When legitimate expertise becomes associated with particular institutions, classes, or political positions, people who distrust those institutions start distrusting expertise itself. They don't grow more humble about what they know. They shift their trust to alternative hierarchies — influencers, ideologues, anyone who performs opposition to the establishment.

So the problem isn't that people trust the wrong experts. It's that expertise has become illegible. The markers that used to signal competence now read, to large parts of the population, as markers of tribal affiliation instead.

The politics of legitimate knowledge

Recognition is political — not in the narrow sense of electoral politics, but in the sense that power shapes what counts as knowledge and who counts as a knower.

Consider which research questions get funded, which methods get treated as rigorous, which journals carry prestige, which kinds of evidence get treated as compelling. None of these are neutral judgments. They come out of historical struggles, institutional path dependencies, and the interests of whoever holds power inside knowledge-producing systems.

This is not the claim that knowledge is just power, or that truth is whatever powerful people say it is. That's the lazy misreading of every thinker who ever noticed that knowledge has political dimensions. The actual point is quieter: the processes that produce and certify knowledge sit inside broader social structures, and those structures have political characteristics. Ignoring that doesn't make it go away. It only makes you naive about how knowledge works.

What follows

If expertise is a social relation, then changing who gets recognized as an expert is a political project as much as an epistemic one. Diversifying scientific institutions is partly about fairness, but it's also about widening the range of questions asked and methods used. Different people notice different things.

Building alternative institutions works the same way. It isn't only about competing with the established ones; it's about opening new pathways to recognition that don't run through gates controlled by people who might not recognize valuable knowledge when they see it.

The catch is that alternative institutions can become just as closed and captured and blind to their own biases as the establishments they set out to challenge. The trouble lives in the structure of gatekeeping itself, not in any particular gatekeeper.

Living with it

There's no clean solution here. Some filtering is necessary; we can't evaluate every claim ourselves, and most claims don't deserve the attention. But every filter introduces bias, every gatekeeping system advantages some people over others, and every expert community has blindspots shaped by who's in it and where it came from.

The best you can do is stay humble about your own certainties, pay attention to challenges from outside your recognition networks, and remember that "expert consensus" names a social fact rather than a guarantee of truth. Experts are often right, especially about well-established findings in mature fields. But the processes that produce them are human processes, with everything that implies.

Expertise is real. Knowledge is real. Who counts as an expert, and whose knowledge counts as legitimate, is a matter of social negotiation as much as epistemic achievement. Seeing that doesn't undermine expertise. It just puts it back in the social world where it was living the whole time.