Ought, context, and why epistemologists are talking past each other

01/08/2025

What if internalists and externalists about justification aren’t really disagreeing—but just speaking in different contexts? In this blog post, Professor Darren Bradley explores whether a contextualist treatment of “justified” could dissolve the debate.

One of the most entrenched disputes in epistemology is between internalists and externalists about justification. Internalists say justification depends on what's "in your head" — your evidence, your perspective, your reasons. Externalists say it also depends on the world — whether your beliefs were formed by reliable processes, whether they track truth.

The arguments are familiar. The Evil Demon thought experiment: imagine your evidence is the same, but you're being tricked by a demon. Shouldn’t your beliefs still be justified? Internalism wins. Then there's clairvoyance: someone gets a reliable gut feeling that the President is in New York. No evidence, no reasons. Just reliable clairvoyance. Should we say he's justified? Externalism cheers. Internalism groans.

And so, the dialectic goes. For decades.

But what if there's a way to explain why these sides keep clashing — and why both seem to have a point? That’s where contextualism comes in. It lets us understand the internalism vs. externalism debate as about, not who’s right, but who’s talking in which context.

Contextualism to the rescue?

In the philosophy of language, contextualism is the idea that the meaning of a term depends on the context of conversation. For example, “tall” is context-sensitive: Michael Jordan is tall for a normal person, but not for an NBA centre.

Now apply that to normative terms like “ought”. What you ought to do depends on context — your goals, the relevant possibilities, the standards in play. In semantics, this gets modelled by two parameters:

A modal base (the relevant set of possible worlds)

A goal or ordering (how to rank those worlds)

For example: “Bob ought to take an umbrella” is true if, in the highest-ranked possible worlds (given it’s raining, and his goal is staying dry), Bob takes the umbrella.

Now, here’s the leap: what if “justified” is like “ought”? What if being justified in believing something just means you epistemically ought to believe it? That’s the deontological conception of justification. It ties justification to what’s permissible, required, or rational — and opens the door to contextualism.

So, we say:

“S is justified in believing P iff S epistemically ought to believe P.”

Once we import contextualist semantics, this becomes:

“S is justified in believing P iff S believes P in the best worlds, given the relevant facts and goals.”

And here’s the twist: what counts as the relevant facts — i.e., what goes into the modal base — varies by context.

Why the debate falls apart

This explains the endless back-and-forth between internalists and externalists. They’re just assuming different modal bases in different conversational contexts.

In first-person deliberation (“what should I believe?”), we naturally only include internal facts — our evidence, experiences, reasons. That’s the internalist context.

In moral or epistemic appraisal (“should they be blamed?”), we often fall back on what was accessible to the agent — another internalist-leaning context.

In third-person evaluation (“is this person a good source of info?”), we care about truth-tracking. We include external facts like reliability. That’s the externalist context.

So, when an internalist says, “The demon victim is justified”, and the externalist says, “No they’re not”, they’re both right — given their contexts. They're not disagreeing about a fact. They're using “justified” with different contextual parameters.

Isn’t that just changing the subject?

You might worry this turns epistemology into semantics. What about the property of justification — not just the word?

Some philosophers argue that there's a normatively privileged kind of justification, one that has real authority regardless of context. Others say we should pick the best version of the concept for philosophical use. Contextualism is compatible with these moves, but it weakens the motivation for thinking there’s a single, privileged kind.

And crucially, contextualism doesn’t eliminate all disagreement. Philosophers can still argue about:

Which contexts matter most

What should go in the modal base

What kinds of norms are epistemically central.

So, this isn’t a semantic trick to shut down debate. It’s a way to clarify what’s really at issue — and to avoid talking past each other.

The takeaway

Contextualism gives us a clean way to explain why internalism and externalism both seem compelling yet endlessly talk past each other. By treating justification as a context-sensitive term (like “ought” or “tall”), we get:

A unified explanation of conflicting intuitions

A way to preserve disagreement where it matters

A new frame for moving forward.

Darren Bradley is Professor of Philosophy at the University of Leeds’ School of Philosophy, Religion and History of Science, where he has taught since 2014. He works primarily in epistemology and metaphysics. He is the author of A Critical Introduction to Formal Epistemology (Bloomsbury, 2015) and numerous papers applying semantics to epistemology e.g. Ought-Contextualism and Reasoning (2021) Synthese 199 (1-2):2977-2999 and Reasons for Belief in Context (forthcoming) Episteme.

You can find more of his work at darrenbradleyphilosophy.