The Diagnostic Workshop
Test a reading task, explain the failure, then revise the artifact
By Evelyna Bellamy
“A workshop can make the gap between intention and interpretation inspectable through a defined task and a recorded revision. This proposed exercise uses invented responses to demonstrate diagnosis without presenting them as research. Reader involvement supplies evidence about particular encounters; broader claims need an appropriate evaluation design.”
Editorial essay by Evelyna Bellamy; not externally peer reviewed. 628 words. Estimated at 200 words per minute for the abstract and main prose; references and interactive studies take additional time. Publication details revised 10 October 2026.
Revised 2026-10-11. Complete bounded draft with identified evidence and declared examples. Earlier universal claims and unsupported causal assertions withdrawn.
An intention is not yet a test
“I want this paragraph to be clear” is a useful wish and a poor evaluation instruction. Clear to whom, for what action, with what background? A diagnostic workshop begins by giving the artifact a task. For an explanatory passage, that might be locating a condition, reconstructing a sequence or distinguishing a proposal from an established result.
The process I propose is to record an intention, expose a version, observe a response, formulate a possible mechanism and revise one relevant feature. These steps are editorial practice, not an experimentally validated sequence. The author can reject a reader’s proposed solution while still taking the observed difficulty seriously. Diagnosis is an argument about a response, rather than a vote on taste.
A miniature workshop, explicitly invented
Use the fictional instruction: “Submit the request after approval unless the item is urgent.” The intended policy is that all requests require approval; urgency affects their place in the queue. The sentence permits another reading: urgent requests may be submitted before approval. The prompt asks a reader whether an urgent request can be submitted before approval and where the sentence supports the answer.
Invented Reader A answers yes and points to “unless.” Invented Reader B answers no but cannot locate the priority rule. These responses are illustrative, not collected observations. They show why an aggregate satisfaction score would be unhelpful here: two people might both like the sentence while inferring incompatible actions. The useful artifact is the answer paired with its reason.
Revise the dependency, not merely the surface
A revised version could read: “Obtain approval before submitting any request. Mark an approved urgent request for priority handling.” The edit separates the approval dependency from the scheduling modifier. It changes the syntax and the information structure. A subsequent test would use the same decision task, with a new version identifier, rather than asking whether the rewrite sounds better.
If the next reader still infers that urgency bypasses approval, several explanations remain possible: an established workplace convention, another contradictory instruction, unfamiliar vocabulary or an issue in the test prompt. The edit does not establish which mechanism is responsible. A workshop should retain competing explanations until its observations can distinguish them.
Choose readers and claims together
W3C WAI’s guidance on involving users in accessibility evaluation supports participation alongside standards evaluation and explains why particular evaluations have limits. Applied here as a methodological caution, it suggests recording the participant context and the task rather than treating a few responses as a universal accessibility verdict. The workshop is not a substitute for the relevant technical checks.
Flower and Hayes’s 1981 paper is included as an identified work in writing-process research. Its publisher record was checked; the full article was not reviewed for this revision. The steps above therefore do not borrow an unverified finding from it. If this exercise were presented as a study, participant selection, materials, consent, coding rules and uncertainty would require their own account.
Preserve the useful disagreement
A revision log should connect each change to the response that prompted it, its proposed mechanism and the result of the next defined task. Keep the original version available. Do not merge an observed difficulty, the facilitator’s explanation and an improvement claim into one field. A failed revision can be informative when that distinction survives.
The earlier draft extended the workshop into guaranteed leadership and empathetic reverse-engineering. Those outcomes were not established. My narrower judgment is that a workshop becomes productive when it lets another person interrupt the author’s theory of the artifact. Its achievement is not that intention wins; it is that a specific revision becomes accountable to a particular encounter.
APA 7th Edition Citations & Primary Sources
WAI guidance recommends combining user involvement with standards evaluation and describes limits of individual evaluations. It does not validate the constructed workshop below.
Primary source textPublisher title, authors and journal identity checked. Full article was not reviewed here; the workshop’s steps and results are an editorial proposal, not findings attributed to this paper.
Primary source text