How should we define propositional knowledge?

Is it possible that our classical pursuit of a singular, static definition of propositional knowledge is fundamentally misguided? For centuries, epistemology has struggled to patch the cracks exposed by Edmund Gettier in his landmark 1963 paper [Is Justified True Belief Knowledge?](https://analysis.oxfordjournals.org/content/23/6/121.abstract). Rather than viewing this as a failure, contemporary theorists suggest that knowledge may be multi-dimensional, adaptive, or even fundamentally un-analyzable. To advance beyond the basic "justified true belief" (JTB) model, we must explore five distinct, sophisticated definitions that challenge, refine, or deconstruct how we conceptualize what it means to "know." ## 1. The Causal Account (Goldman) Alvin Goldman proposed a radical shift away from internal justification toward external, naturalistic processes. In his seminal paper [A Causal Theory of Knowing](https://www.jstor.org/stable/2023568), Goldman argues that for a subject to know a proposition, there must be a reconstruction of the JTB model where a causal chain connects the fact with the belief. > "S knows that p if and only if the fact that p is causally connected in an appropriate way with S's believing p." For example, if you see a rebuilding of a volcanic eruption on a seismograph, your belief that a volcano erupted is knowledge because the physical eruption causally triggered the seismograph, which in turn caused your belief. ## 2. The Tracking Theory (Nozick) In *Philosophical Explanations*, Robert Nozick introduced a subjunctive, counterfactual account of knowledge known as "truth-tracking." Nozick's definition replaces traditional justification with two subjunctive conditionals: 1. If $p$ were not true, $S$ would not believe that $p$. 2. If $p$ were true (in slightly different circumstances), $S$ would believe that $p$. This definition beautifully handles gettierized scenarios. If you look at a highly realistic hologram of a vase, your belief that "there is a vase on the table" is true only by coincidence; if the vase were not there, you would still believe it was (violating condition 1). Thus, you do not possess knowledge. ## 3. Virtue Epistemology (Sosa and Zagzebski) Virtue epistemology shifts the focus from the properties of the proposition to the intellectual character of the cognizer. Linda Zagzebski, in her influential book [Virtues of the Mind](https://www.cambridge.org/core/books/virtues-of-the-mind/1BF77BD2B0058A2B6F0764724A9A65E1), defines knowledge as a state of cognitive contact with reality arising from acts of intellectual virtue. Under this framework, belief must be "apt"—meaning it is true *because* of the agent's intellectual competence, attentiveness, and intellectual honesty, rather than mere epistemic luck. ## 4. Knowledge-First Epistemology (Williamson) Perhaps the most counterintuitive turn in modern epistemology is Timothy Williamson’s "Knowledge-First" framework, detailed in [Knowledge and its Limits](https://academic.oup.com/book/5256). Williamson rejects the decades-long project of breaking knowledge down into simpler components like belief, truth, and justification. Instead, he argues that knowledge is a broad, prime, un-analyzable mental state. > "Knowledge is the most general factive mental state." In this view, we should not define knowledge using belief; rather, we must define belief as a botched or attempted state of knowledge. ## 5. Pragmatic Encroachment (Stanley and Fantl) Drawing unexpected connections to decision theory and ethics, the theory of Pragmatic Encroachment posits that whether a true belief counts as knowledge depends heavily on the practical stakes of the situation. As Jason Stanley argues in [Knowledge and Practical Interests](https://academic.oup.com/book/10852), if the stakes are low, a well-supported belief constitutes knowledge. However, if the stakes rise dramatically (e.g., a medical diagnosis where a mistake is fatal), the exact same level of evidence may no longer suffice to claim "knowledge." Here, the traditional boundary between epistemology and action dissolves.