Background and Context
This maxim was collected October 26, 2018, the last-filed of three numbered "Rules of Good Instructional Design," created roughly twenty-seven minutes after the Second Rule and twenty-six after the First — the one member of the set where the creation order actually matches the numbering, unlike its two siblings, whose timestamps land in the reverse of their labels. Read as the close of a single sitting's worth of accumulated professional irritation, it also reads as the widest-angle of the three: where the First Rule polices the individual sentence and the Second polices the document's internal structure, the Third steps outside the document entirely, to the situation the reader is actually standing in when they try to make sense of it.
Interpretation
The maxim has the same two-clause architecture as its siblings — a positive instruction ("good context is key") followed by the stakes clause that gives the set its recurring vocabulary ("makes you look like an idiot") — but it names two distinct failure modes inside a single word, context, where the other two rules named one. "Bad or missing" is doing real work: missing context is an absence, a gap the reader has to fill in on their own, usually incorrectly; bad context is worse than absence, because it actively points the reader somewhere wrong while looking like guidance. An instruction that says "revise section 3" with no note about which draft, whose feedback, or what changed since the last version is missing context — the reader has to guess. An instruction that says "revise section 3 per the attached notes" when the attached notes describe a different section is bad context — the reader isn't guessing, they're confidently wrong, which is a harder error to catch and a more embarrassing one to have caused.
What "context" means here is the surrounding information that lets a fact, instruction, or claim be correctly situated: what came before it, what it assumes the reader already knows, what changed and when, why this version exists rather than an earlier one. Instructional-design theory has a name for supplying it deliberately — the advance organizer, David Ausubel's term for a short framing passage given before new material that tells the learner how what's coming connects to what they already know. Cognitive load research gives the same idea a mechanical explanation: working memory is a narrow channel, and a learner who has to simultaneously parse new content and reconstruct the missing background it assumes is doing two jobs where good design would have done one. Context, on this reading, isn't decoration around the instruction — it's part of the instruction's actual payload; strip it out and the remaining words often stop meaning what the writer intended, even though every individual sentence may still be grammatically sound.
That distinction is what separates this rule from its siblings rather than simply restating them. The First Rule assumes a document is otherwise sound and asks only that the words be correct. The Second assumes the words are correct and asks whether they'll stay correct once duplicated. The Third assumes the words are correct and undisturbed and asks a different question again: does the reader, standing where they're actually standing — without the meeting the writer sat in, without the previous version they never saw, without the tacit knowledge the writer forgot wasn't universal — have what they need to correctly interpret them at all? A perfectly proofread, perfectly non-duplicated instruction can still fail this test completely. In that sense the Third Rule is the least forgiving of the three, because it can't be satisfied by care applied to the document alone; it requires the writer to imagine a reader who wasn't present for anything the writer already knows, and to build that absence back in.
There's a specific instructional-design failure this maxim is clearly reaching for: an assignment, job aid, or slide deck written by someone deeply immersed in a project, for an audience that is not. The writer's context is so total and so automatic that it becomes invisible to them — they no longer notice which facts they're assuming, because those facts stopped feeling like facts and started feeling like common sense. The reader has no such immersion. They meet the instruction cold, and if the instruction was written for someone who already knew the backstory, the reader either fills the gap with a wrong guess or gives up trying to fill it at all. Either way, the writer looks like they didn't think about who would actually be reading this — which is precisely the maxim's charge.
Current Relevance
The maxim's underlying risk has grown more, not less, acute since 2018, largely because so much instructional and professional content now circulates stripped from whatever produced it. A slide pulled from a deck and forwarded without the talk that accompanied it; a screenshot of a policy passed around a chat channel without the thread that explained why it changed; a short-form video or microlearning clip that has to hit under ninety seconds and so cuts the framing first, because framing is what's easiest to cut and hardest to notice is missing. Each is a small, deliberate act of context removal, made for good production reasons, that quietly reproduces the exact failure this maxim warns against at scale.
Generative AI tooling sharpens the problem in both directions at once. A prompt handed to an AI system without the background a human collaborator would have absorbed over weeks — the project's history, the audience's actual skill level, the reason this draft exists rather than the version before it — routinely gets back fluent, confident output built on the wrong assumptions, because the system had no way to ask for what wasn't given. And an AI-drafted summary handed to a human reader carries the same risk from the other side: it can read as complete and well-organized while quietly having dropped the caveats, the "as of" dates, or the "this only applies if" clauses that gave the original its actual meaning, because a summary's job is to compress, and context is usually the first thing compression sacrifices. The maxim's warning — that missing or bad context looks like a small omission but produces a large, credibility-costing error — describes both failure modes with equal precision, even though neither existed in its current form in 2018.
The same dynamic recurs constantly outside instructional design proper: a policy quoted out of the paragraph that qualified it, a data point cited without the base rate that would have changed its meaning, a decision announced without the reasoning that made it defensible, each one technically accurate and functionally misleading. Remote and asynchronous work compounds it further, the same way it compounds the Second Rule's drift problem — a colleague reading a decision three time zones and two weeks removed from the meeting where it was made has no chance to ask the clarifying question a person in the room would have asked in the moment, so the document has to supply, in writing, everything that meeting's presence used to supply for free.
Impact and Legacy
As with its two siblings, the Third Rule's legacy is entirely local — a private working standard rather than a public argument, legible only inside the small numbered set and the wider maxim folder it belongs to. Read as the set's closing member, it functions as something closer to the animating principle behind the other two than a third, unrelated item on a checklist. A grammatically flawless, non-duplicated instruction can still fail the reader if it assumes knowledge the reader doesn't have; read that way, the First and Second Rules are really specific instances of a single, broader discipline the Third Rule names directly — make sure the reader has what they need to understand you, whether that means correct words, a single authoritative copy, or the surrounding facts that give the words their meaning. That the Third Rule was filed last, and is the only one of the three whose creation order matches its number, gives it a plausible claim to being the summary insight of the sitting — the one arrived at after the other two had already been written down and named, rather than the first irritation reached for.
Its closest public analogue is the instructional-design literature's long-running argument for advance organizers and schema activation — Ausubel's framing device, and the broader constructivist claim that new information is only retained in proportion to how well it's anchored to what the learner already knows. Mike's maxim arrives at a compressed, practitioner's version of the identical insight, without the theoretical apparatus: not "activate prior schema" but the plainer, harsher "don't make me look like an idiot by leaving out what I needed to know." Its influence, like its siblings', never left the private folder it was written into, but the convergence with a well-established strand of learning-science research is itself a kind of legacy — evidence that the rule names something instructional designers keep rediscovering independently of the theory, from the practitioner's side rather than the researcher's.
Contrasting Views or Controversies
The maxim's clearest weakness is that "context" is elastic almost to the point of being unfalsifiable — it can be invoked to justify nearly any amount of additional framing, background, or caveat, and instructional design has its own well-supported case against over-explaining. Cognitive load theory, the same body of research that explains why missing context overloads a learner, also explains why excessive context does: an advance organizer that turns into three paragraphs of throat-clearing before the actual instruction arrives spends the reader's attention on background they didn't need, at the cost of the working memory left over for the task itself. The maxim gives no guidance on where "enough" context ends and padding begins, which leaves it more useful as a disposition to check against than a rule that specifies its own stopping point.
A second, sharper objection is that "enough context" is not a fixed property of a document — it depends entirely on who's reading it, and a designer who over-corrects toward exhaustive context for a novice audience will bury an expert audience in explanation of things they already know. Differentiated instruction exists precisely because the same material can be under-contextualized for one learner and over-contextualized for another simultaneously; the maxim's "you," addressed to the single hypothetical reader who "looks like an idiot" without enough context, has no obvious answer for the case where two different real readers need two different amounts of it.
A third, more structural objection concerns pedagogy that withholds context on purpose. Problem-based learning, discovery learning, and what cognitive scientists call desirable difficulty — Robert Bjork's term for a deliberately introduced obstacle that produces worse short-term performance but stronger long-term retention — all depend on a learner grappling with genuine ambiguity rather than being handed the full frame up front. A designer who read this maxim too literally, and treated every gap in context as a defect to be closed, would be eliminating a class of intentional pedagogical friction that the research supports as effective, in service of a rule aimed at accidental, not deliberate, absence. The maxim, read carefully, is a warning about context the writer forgot to supply, not an argument against context a designer chose to withhold as part of the lesson — but its short form doesn't make that distinction, and a rigid application of it could flatten a genuinely useful teaching technique into a design flaw.
Practical Application
- Individual (designer/writer): Before publishing an instruction, job aid, or slide, ask what a reader who wasn't in the room — didn't hear the meeting, didn't see the earlier draft, doesn't share your assumptions — would need to correctly interpret it, and write that in rather than trusting it to be obvious.
- Individual (reviewer): Read the material cold, deliberately setting aside what you already know about the project, and flag every place where understanding depends on knowledge the document itself doesn't supply.
- Organizational (learning teams): Build a standard context-setting element — an advance organizer, a "why this matters" framing, a changelog note — into the template for instructional materials, so supplying context becomes a default step rather than something each writer has to remember to add.
- Organizational (AI-assisted workflows): When prompting or delegating to an AI system, supply the same background a competent new hire would need — project history, audience level, why this version exists — rather than assuming the system will infer what a human collaborator would have absorbed over time; treat missing context in a prompt as the direct cause of a plausible-sounding but wrong output, not a mysterious AI failure.
- Civic/professional (any public communication): Before quoting a policy, data point, or decision out of its source, check whether the surrounding paragraph, caveat, or "as of" date changes what the excerpt actually means — a technically accurate quotation stripped of its frame is exactly the "bad context" this maxim warns against, not a neutral shortening.
- Organizational (onboarding and handoffs): When a new hire, contractor, or successor takes over a project, hand over the why alongside the what — the abandoned earlier approach, the constraint that shaped the current design, the decision that looks arbitrary until you know what it was chosen over — since the person inheriting the work has none of the context that made those choices obvious to whoever made them, and will otherwise either repeat a mistake already ruled out or waste time re-deriving a decision that was already settled.
Background on the Author
Mike Hower's collected maxims read as field notes from a career built on watching instructional material succeed or fail in front of real learners, and the Third Rule's target — context the writer forgot wasn't universal — is the kind of failure that becomes visible only after enough repetitions of watching a learner misread something the writer thought was self-evident. That it closes the numbered set, and is the one of the three whose filing order actually matches its label, is consistent with its character as the broadest and most abstract of the three: a rule reached for after the more concrete irritations behind the First and Second Rules had already been named, when what remained to say was the thing underneath both of them. Someone who had only ever written instructions for an audience that shared their full context — a note to a colleague already in the loop, say — would have little reason to notice this failure mode; it shows up specifically in instructional design, where the writer is almost always more informed about the material than the learner they're writing for, and where that gap in context is the ordinary condition of the job rather than an occasional accident.