Humanities research journal — L04
L04 — The Epistemology of Craft
Tacit Knowing, Skilled Judgment, and the Making of an Instrument
2026-11-05 · v1.0 · Literature/conceptual review edition / craft-observation validation pending
| Code | L04 — The Epistemology of Craft |
|---|---|
| Focus | tacit knowing · subsidiary / focal awareness · craft judgment |
| Evidence | Polanyi [1–2] · Sennett [3–4] |
| RavenForge link | Craft Judgment Record |
| Upstream | L02 · L03 · L13 |
| Downstream | L05 — Craft, CNC and the Location of Judgment |
| Status | Literature/conceptual review edition / craft-observation validation pending |
| Date / revision | 2026-11-05 / v1.0 |
1. Research question — when does making become knowing?
A drawing describes a target geometry; a tolerance defines an acceptable range. The maker must still read the current situation and choose an intervention. L04 asks how specifications are used in judgment. The same correction request may require different actions when material condition, support, tools or production stage change. Until observed, this possibility is not evidence of any particular maker’s competence.
2. Polanyi — subsidiary and focal awareness
Polanyi describes knowing a focal object through the integration of contributing clues [1]. Subsidiary awareness names a functional relationship to that object, not simply unconsciousness. Focusing on a clue itself can alter the meaning it has within performance. This from–to relationship is more useful here than treating tacit and explicit knowledge as separate stores. The Tacit Dimension provides background; its entire text was not checked for this edition [2].
3. Tacit knowing is not secret knowledge
An incomplete explanation does not make explanation or testing impossible. Conversely, a longer manual does not necessarily transfer the integration required in performance. L04 records observable cues where possible without claiming to reproduce the whole judgment process. Unexplained aspects remain explicit gaps with conditions for further review. This recording principle is a RavenForge methodological choice.
4. Indwelling, skill and apprenticeship
Polanyi’s discussion of tool use concerns attending to a task through reliance on body and tool [1]. An educational application can therefore propose records of demonstration, performance, feedback and renewed attempts. Repeating an instructor’s words differs from judging appropriately under changed conditions. L04 specifies no universal time threshold for acquiring skill and does not treat repetition counts as guarantees of competence.
5. Sennett — reading making and thinking together
Sennett extends craftsmanship beyond manual occupations to the development of skill and a commitment to doing work well for its own sake [3]. The official contents place workshops, machines, material consciousness, the hand, resistance and ambiguity, and quality-driven work within the same inquiry [4]. L04 uses that scope to approach making as an activity involving judgment and learning. It does not claim a full chapter-by-chapter verification or infer the superiority of hand tools from bibliographic evidence.
6. Material resistance and standards of quality
Unexpected cutting resistance or contact marks may prompt further inspection; they do not establish a cause on their own. Tool condition, grain direction, jig support and operator movement remain alternative explanations. Quality is not reduced to a single sensory score. When geometry, structure, playing requirements, serviceability and appearance conflict, the priority chosen must be recorded. This is a proposed instrument-making application of Sennett’s material and quality themes.
7. Specification and judgment — preserve disagreement
If a measurement meets tolerance but a player reports discomfort, retain conformity and the use evaluation as separate findings. A positive impression does not cancel a structural nonconformity either. Version the criteria and reasons for changing them so that standards are not silently adjusted after seeing results. Distinguish a fluent explanation from a judgment supported by repeated observation.
8. Making examples — applications, not observations
In fretwork, straightedge responses, surface traces and resistance changes may prompt further checks. In neck shaping, section dimensions, side-to-side comparisons and grip experience belong in distinct records. Fit assessment can record gaps and contact traces; setup records should include strings, tuning, support and environment. These examples prescribe neither a procedure nor acceptance values, and are not reported observations of RavenForge instruments.
9. Craft Judgment Record — preserve the basis of a decision
Record one judgment using the fields below. Attach the workpiece ID, date, operator, drawing/tool/jig versions, environment and paths to photographs or measurement files. Separate prior observation from interpretation after intervention; do not fill unobserved cues with guesses. The form is a RavenForge proposal, not a standard devised by Polanyi or Sennett.
| Field | Record |
|---|---|
| Explicit specification | Drawing, tolerance, inspection conditions and criterion version |
| Observable cue | What was actually seen, heard, felt or measured |
| Judgment | Chosen interpretation, alternatives and grounds |
| Intervention | Location, method, sequence and scope for reversal |
| Outcome | Post-intervention measurements and observations; unknowns stay unknown |
| Confidence / uncertainty | Grounds for repeatable, tentative or unresolved status |
| Review condition | Triggers such as string tension, environmental change or player feedback |
10. Articulability and testability are separate axes
A criterion that is easy to state has not thereby been shown to be valid. A judgment whose reasons are difficult to articulate may still be assessed against observable outcomes under defined conditions. The matrix is a recording distinction, not a scoring scale for people or expertise.
| Outcome test conditions defined | Outcome test conditions undefined | |
|---|---|---|
| Easier to articulate | Fix tolerance and measurement conditions; check actual conformity | Documented preferences or rules still require validation |
| Harder to articulate | Check outcomes and repeatability even if cues are only partly explained | Keep judgment provisional; design observation conditions first |
11. Boundaries with L02, L03 and L13
L02 addresses bodily incorporation, L03 the actions available to a player, and L13 technological mediation. L04 concerns how makers judge from cues and revise those judgments. Earlier studies provide continuity of questions. Their existence does not independently establish the accuracy of a maker’s judgment or the quality of a product.
12. Myth audit — claims not adopted
This study does not conclude that craft intuition outranks measurement, that tacit knowing is entirely inexpressible, that only handwork is authentic craft, that accumulated hours guarantee expertise, or that skilled judgment is always objective. Nor does it assume that documentation replaces all skill or that CNC removes judgment.
13. Future validation — a protocol not yet executed
First narrow the task and judgment question, then fix the criteria in advance. Collect cues before intervention, the intervention chosen, and subsequent measurements and use evaluations separately. Do not confuse repeated measurement of one workpiece with comparisons between different workpieces. Retrospective explanations are limited by memory and hindsight; concurrent speech may alter performance, so record when explanations were collected. Preserve failures, disagreements and incomplete cases. Where comparison is feasible, report differences between independent reviewers. Sample sizes and accuracy targets should follow the research question and a pilot, not arbitrary values assigned in this article.
14. Current conclusion and the handoff to L05
Making knowledge can be approached as the practice of integrating explicit resources with sensory cues, experience and quality standards to judge a particular situation. Records should connect that judgment with observable outcomes and make revision possible. The next regular study, L05 — Craft, CNC and the Location of Judgment, will examine how judgment is distributed across design, toolpaths, fixtures, intervention during machining and inspection. L05 is a future study, not part of this publication.
References and verification scope
[1] Polanyi, M. (1965). The Structure of Consciousness. Brain, 88, 799–810. Authorized web text follows Knowing and Being (1969), 211–224.
https://www.polanyisociety.org/mp-structure.htm
[2] Polanyi, M. The Tacit Dimension. University of Chicago Press, 2009 edition; foreword by Amartya Sen. Bibliographic record and publisher description.
https://press.uchicago.edu/ucp/books/book/chicago/T/bo6035368
[3] Sennett, R. The Craftsman. Yale University Press, 2009 paperback, ISBN 9780300151190. Bibliographic record and publisher description.
https://yalebooks.yale.edu/book/9780300151190/the-craftsman/
[4] Sennett, R. The Craftsman. Official table of contents, Yale University Press.
https://drupal.yalebooks.yale.edu/sites/default/files/files/TOC/sennett_craftsman.pdf
Scope checked on 2026-11-05: [1] is the authorized public text hosted by the Polanyi Society. Its notice states that it follows the 1969 Knowing and Being version, with a modified first paragraph relative to the 1965 Brain publication. For [2], the Chicago Press 2009 edition’s bibliographic record and description were checked; for [3], Yale’s 2009 paperback record and description; for [4], the official contents. This is not a claim to have verified the complete books or chapter-specific wording and page quotations. Additional secondary literature suggested in the research report is not used as direct evidence in this edition. All making examples, matrices and record forms are RavenForge methodological proposals.