theories of making

Works

Three statements, from sociology, design, and anthropology.

• Sennett's The Craftsman (2008) argues that the desire to do a job well for its own sake is a basic human impulse, systematically obstructed by modern work organization, and traces the craft-thinking connection through hand, tool, and resistance.
• Pye's The Nature and Art of Workmanship (1968) supplies the workmanship of risk and of certainty: whether the result is continuously at hazard during the work, or predetermined by the jig and the machine.
• Ingold's Making (2013) and The Perception of the Environment (2000) supply correspondence: the maker follows the material's grain and tendencies rather than realizing a prior form, which reverses the hylomorphic model inherited from Aristotle.
• Polanyi's tacit knowledge, documented elsewhere in this record, is the epistemological support the whole tradition leans on — we know more than we can tell.

Central ideas

Four ideas, and the second is the field's most useful distinction.

• Knowledge in the hands: skilled making involves judgments that cannot be fully stated, so instruction supplements practice and never replaces it.
• Workmanship of risk against certainty: risk means the outcome is at hazard throughout — a slip ruins it — where certainty means the process guarantees the result. Neither is better; the distinction explains what each can produce.
• Correspondence, not imposition: material has grain, tendency, and resistance, and skilled making is a conversation with them rather than the execution of a plan on passive stuff.
• Resistance is informative: difficulty in the work is where the knowledge is, which is why removing all friction from a process removes the learning with it.

Distinctive vocabulary

Five terms, guarded.

• Craft: skilled making with judgment. Not handmade, and not a hobby category.
• Workmanship of risk: work whose outcome is continuously at hazard. Not dangerous work.
• Tacit knowledge: what is known and cannot be fully stated. Polanyi's term, documented separately here.
• Correspondence: following a material's tendencies. Not communication in the metaphorical sense.
• Hylomorphism: the form-imposed-on-matter model. The thing this tradition is arguing against, not a term it uses approvingly.

Debates and disagreements

The field's quarrels are about romanticism and about who gets to make.

• Whether craft is being romanticized: Sennett's craftsman is a dignified figure, and critics note that actual craft labour has often been poorly paid, physically destructive, and organized by guilds that excluded most people.
• Whether the risk/certainty distinction ranks: Pye insisted it does not, and the tradition's readers persistently treat risk as superior — a misreading with real consequences for how manufactured things are judged.
• Whether tacit knowledge is a real category or a placeholder: critics argue it names what has not yet been articulated rather than what cannot be, which would make the whole epistemological claim provisional.
• Whether making can be studied without making: the field is written largely by people with workshop experience, and its own thesis implies that a purely textual engagement with it is limited.

Intellectual relationships

Two siblings here share its five placements.

• Poetics, documented here, studies the demands of the object; this tradition studies the maker's engagement with the material — the same five claims read from opposite ends of the workbench.
• Pragmatist aesthetics, also here, locates value in the experience; making locates it in the doing and the resistance met.
• The pedagogical-developmental lens holds the transmission question — Schön's reflection-in-action is craft thinking given a professional-education vocabulary.
• Tacit knowledge and the epistemological lens supply the account of unstatable knowing that this tradition depends on.

How it changes this reading

Five placements, read through material rather than form or experience.

• The tradition's throughline in this record is that material is an honest interlocutor: it reports failure without regard for feeling, which is why accuracy-over-affirmation reads here as a fact about feedback rather than a preference about manners.
• For clarity and beauty, it supplies fitness — the resolved form is the beautiful one because both are what a solved problem looks like.
• For practice as the unit, it supplies the epistemology: no route to this knowledge skips the doing.
• For iteration, it supplies correspondence — a maker following a material is revising continuously, so iteration is the relation rather than a discipline added to it.

Useful comparisons

Against its neighbours.

• Against design theory: making studies the encounter with material, design studies the specification that precedes it — and this tradition's central argument is that the second cannot fully govern the first.
• Against art theory: craft's standard is fitness and skill, art's is meaning and effect, and the boundary between them is a historical settlement rather than a natural one.
• Against skill acquisition research: same subject, different method — narrative and phenomenological description against measurement of expertise.

Where the ideas meet

The shared floor.

• Thinking and making are not sequential: the hands are part of the reasoning, not its output device.
• Material has agency of a kind: it constrains, suggests, and resists, and treating it as inert produces bad work.
• Time is required: no member believes skill can be compressed, and all three treat the duration as constitutive.

Where they part

Where they part.

• On the unit: the craftsman as a social figure (Sennett), the work process (Pye), the ongoing relation with material (Ingold).
• On modernity: whether industrial production destroys craft or relocates it, which decides whether the tradition is elegiac or analytic.
• On plans: Ingold nearly dissolves the design stage, where Pye's certainty category takes it entirely seriously.

Limits

What the tradition cannot do.

• It cannot transmit what it studies: the field's own thesis means reading it does not confer the knowledge it describes.
• It cannot evaluate a maker: skill is demonstrated in work, and no account of craft licenses a judgment about a person's practice from a description of it.
• It generalizes from material crafts: how far the claims extend to writing, code, or care work is asserted more often than shown.
• It has no economics: the conditions under which craft is possible are largely outside the tradition's account, which the class objection presses.

Criticisms

Standing objections, at strength.

• That it romanticizes: the dignified craftsman is a literary figure, and historical craft work was frequently exploitative and exclusionary.
• That risk-over-certainty is a taste presented as an analysis, despite Pye's own insistence otherwise.
• That tacit knowledge may be a name for the not-yet-articulated, which would make the epistemological centre softer than the tradition claims.
• That the field is written by people who could afford workshops, and it says little about making under economic pressure.

Common misreadings

Four.

• Craft as authenticity marketing: the vocabulary attached to products to signal value the process did not involve.
• Handmade as superior: Pye's distinction is analytic, and using it to rank outputs is the misreading he anticipated.
• Tacit knowledge as a conversation-stopper: 'you can't explain it' is a claim about a limit, not a licence to stop trying.
• Extending craft metaphors without argument: calling any careful work craft borrows a tradition's authority for a claim about material engagement that may not hold.

What remains outside this idea

The boundary.

• That a maker possesses skill: the tradition describes what skill is and cannot certify anyone's.
• That a well-made thing was made well: outcome underdetermines process, and the workmanship of certainty exists precisely because good results can be guaranteed without judgment.
• That craft knowledge transfers to non-material work: the extension is plausible, widely assumed, and not demonstrated.