Knowing-that: propositional knowledge
“Rina knows that the train leaves at six.”
This is the form that dominates analytic epistemology. It involves a proposition: the train leaves at six. In ordinary cases it is factive: if the train does not leave at six, Rina did not know that it did, however reasonable her belief may have been.
Propositional knowledge is indispensable. But treating it as the master template for all knowing can distort the landscape.
Knowing-how and procedural knowledge
A surgeon may know how to tie a knot she cannot fully describe. A cyclist can balance without solving differential equations. A jazz musician can resolve a phrase without possessing a proposition that lists every micro-movement.
Gilbert Ryle famously attacked the “intellectualist legend” that intelligent action must always be guided by prior propositional instructions. Contemporary debate is more complicated: some philosophers argue that knowing-how can still be analyzed propositionally, perhaps as knowing a way to perform an action. The empirical and philosophical point that survives is narrower:
Competent performance can outrun explicit verbal representation.
Procedural knowledge matters because human expertise is often compressed into sensorimotor and perceptual discrimination rather than accessible sentences.
Knowledge by acquaintance
Bertrand Russell distinguished knowledge by acquaintance from knowledge by description. Whatever one thinks of his exact theory, the contrast identifies something real: knowing about grief and having grieved are not interchangeable epistemic positions. You can know neurobiology, poetry, and population statistics about grief while lacking first-person familiarity with the phenomenology. Conversely, a grieving person may know the experience intimately while knowing almost nothing about its biology or epidemiology.
Acquaintance is not automatically superior. First-person immediacy can reveal aspects inaccessible to outsiders, but it can also be narrow, idiosyncratic, or conceptually confused. “Lived experience” is evidence about some questions; it is not a universal trump card.
Tacit knowledge
Michael Polanyi’s slogan that “we can know more than we can tell” points to skills and discriminations that resist exhaustive articulation. Radiologists, chess players, craftspeople, musicians, and mechanics often perceive patterns before they can explain the cues.
Tacitness creates a problem: if expertise cannot be fully stated, how can outsiders audit it? The answer cannot be “trust the expert blindly,” but neither can it be “only explicit rules count.” We often test tacit expertise through track record, controlled comparison, inter-rater performance, apprenticeship outcomes, prediction, and error analysis.
Embodied knowledge
Bodies are not merely containers for brains. Posture, sensorimotor history, pain, fatigue, tool use, perceptual habits, and environmental affordances change what can be noticed and done. Phenomenology, ecological psychology, enactivist traditions, and embodied cognitive science disagree about theory, but converge on a warning: cognition does not always decompose neatly into detached propositions manipulating neutral inputs.
Calling something “embodied knowledge,” however, should not become a fog machine. The phrase is useful only when one can say what information or competence depends on bodily coupling and how that dependence changes performance.
Social and distributed knowledge
A ship enters harbor safely because no single sailor contains the complete computational state of the navigation system. Charts, instruments, procedures, spatial arrangements, talk, trained roles, and multiple people jointly carry the task. Edwin Hutchins’s study of navigation made this distributed structure vivid:
Cognitive properties can belong to an organized activity system, not just to isolated individuals.
Modern civilization extends this principle enormously. No one person knows how to manufacture a leading-edge microprocessor from raw materials, run an international air-traffic network, or verify every component of a vaccine supply chain. Yet societies can possess operational knowledge embodied in institutions, standards, archives, artifacts, and coordinated specialists.
Does the system literally know? Philosophers differ. But even if “knowledge” ultimately requires individual knowers, the individual’s epistemic success can depend on structures no individual understands in full.
Testimony, expertise, and authority
Testimony is not merely gossip but is the mechanism through which language lets one person’s contact with evidence become another person’s reason for belief.
If every testimonial belief required complete independent verification, testimony would cease to function as testimony. C. A. J. Coady argued that the Western tradition had underestimated how fundamentally our knowledge depends on the word of others.
Expertise adds a layer: we defer not simply because someone says p, but because we believe they possess relevant competence. Yet nonexperts face a meta-expertise problem: how do I identify genuine expertise without already knowing the field well enough to judge the expert directly?
Useful proxies include:
- training and demonstrated competence;
- relevant rather than merely adjacent expertise;
- track record under conditions where error can be observed;
- independence from the claim’s incentives;
- openness to criticism and correction;
- agreement among independently trained specialists;
- quality of the institution’s error-detection mechanisms;
- ability to distinguish known, uncertain, and speculative conclusions.
None is infallible. Together they often dominate “I watched three videos and formed my own view.”
Practical wisdom
Aristotle’s phronēsis is conventionally rendered “practical wisdom,” but the English phrase does not exhaust its ethical and deliberative role in Aristotle’s system. It concerns good deliberation about action in human life, not simply possessing more true propositions.
Other traditions also distinguish liberating, transformative, moral, or practical cognition from merely possessing information. But the concepts are not interchangeable. The important general question is whether a person can know many relevant facts yet still fail to judge what matters, what to prioritize, when to stop optimizing, or how to act under tragic tradeoffs.
They plainly can.
Interim conclusion:
There is no good reason to assume that knowing-that, knowing-how, acquaintance, tacit competence, distributed knowledge, testimony, expertise, and wisdom are simply one phenomenon wearing different clothes. Nor is there good reason to declare them utterly unrelated. They share a family resemblance around successful cognitive contact, competence, orientation, or action in relation to reality.
The mistake is to force every member of the family into the shape of a sentence in a textbook.