Interlude

Theory: From Theory to Practice | Template: The Historical Essay | Words: ~1,720

# Theory, Craft, and the Distance Between Them

There is a joke among instructional designers that every theory survives the journey from lab to classroom the way a snowball survives the journey from Alaska to Lagos. The joke is unfair to theorists and unfair to teachers. But it points at something real. The distance between a learning theory in its original form and the version of it that shows up in a course template or an LMS setting is sometimes so great that the two are related only by ancestry.

Act II is about that journey.

What We Used to Believe

For most of the twentieth century, the assumption was that good theory would simply diffuse into good practice. Dewey wrote, teachers read, and classrooms changed. In reality, the diffusion was slow, selective, and heavily filtered by whatever the dominant institutional pressure of the decade happened to be. Tyler's rationale for curriculum (Tyler, 1949) was widely adopted not because it was the deepest model on offer but because it fit the post-war appetite for measurable objectives. Bloom's taxonomy (Bloom, 1956) spread because administrators could put it on a form. Gagne's nine events (Gagne, 1985) spread because they mapped cleanly onto a slide deck.

None of this is a criticism of the theorists. It is a description of what institutions do with ideas: they keep the parts that fit the paperwork and quietly drop the rest.

The Research That Changed Everything

Three lines of work forced the field to take the translation problem seriously.

First, research on expert teaching showed that what distinguishes strong teachers is not command of a single theory but a working repertoire of them, selected on the fly according to the learner and the moment (Shulman, 1987; Hattie, 2009). Expert teaching is plural. Most instructional design templates are singular.

Second, research on implementation science showed that even when a method is proven to work, the fidelity with which it is implemented in the field usually determines whether it works for this cohort in this institution this year (Durlak & DuPre, 2008). A method is not a result. A method plus its implementation conditions is a result.

Third, research on design-based methods argued that the way to close the gap between theory and practice is not to wait for theory to trickle down but to iterate theory and practice together, in real settings, with real learners, producing both local improvement and generalisable claims (Brown, 1992; Design-Based Research Collective, 2003). This is the intellectual ancestor of the modern instructional design profession, and the reason a serious ID team looks more like a research lab than a production line.

Together, these three strands make it clear that instructional design is a second-order discipline. It is not applied psychology. It is not applied data science. It is the craft of reconciling a theoretical claim, a real population of learners, an institutional reality, and a budget, in a single artefact that has to actually work on Monday morning.

The Framework

Act II walks through the models that emerged to do that reconciling: ADDIE, SAM, Dick and Carey, Merrill's First Principles, Keller's ARCS, Kirkpatrick's four levels, the Community of Inquiry framework, and a long list of assessment and evaluation models that try, with varying success, to close the loop.

Read in isolation, these models look like competitors. Read in sequence, they look like the field's slow, messy, cumulative answer to the same underlying question: how do you build something that teaches, and how do you know whether it worked?

Some of the models in Act II are still fit for purpose. Some are kept alive mostly by habit and certification economies. Part of the work of the act is to name the difference.

The Invitation

As you read Act II, resist the urge to rank the models. Ask instead which problem each one was trying to solve, and whether that problem is still the problem you are being paid to solve today. The ones that survive the question honestly are the ones worth keeping in your toolkit.

Which of the models you use every week are you using because they still work, and which because they still fit the form?