Introduction To The Series
Theory: Series Introduction | Template: The Framing Essay | Words: ~1,850
# Why Content Delivery Is Not the Same as Learning
For most of the last two decades, the conversation about educational technology has assumed a simple equation: if we can deliver the right content to the right person at the right time, learning will happen. That assumption is the reason every EdTech pitch deck eventually lands on the word "personalisation," and the reason adaptive platforms measure themselves by the sophistication of their recommender systems. It is also, quietly, the reason so many of those platforms have disappointed the teachers and learners who depend on them.
This series is written against that assumption. Not against technology, and not against personalisation. Against the equation itself.
What We Used to Believe
The content-delivery view of learning is older than EdTech. It descends from a long line of thinkers who, with good intentions, treated the learner as a container to be filled. Freire named this "the banking model" and argued that it confuses instruction with deposit-making (Freire, 1970). Skinner's teaching machines, decades earlier, had already encoded the same assumption into hardware: small steps, correct response, reward, repeat (Skinner, 1958). When the internet arrived, we ported the banking model to the browser. When algorithms arrived, we ported it to the recommender.
The appeal is understandable. Content is easy to count. Delivery is easy to automate. Engagement is easy to measure in clicks. The harder things — transfer, identity, belonging, the willingness to try again after failure — resist measurement, so they tend to fall out of the product spec. What is measured is what gets built.
The Research That Changed Everything
Three bodies of evidence, none of them new, make the content-delivery view untenable.
First, research on transfer and situated cognition shows that knowledge learned in one context rarely moves to another without deliberate design (Lave & Wenger, 1991; Bransford et al., 2000). Delivering a well-sequenced module does not guarantee the learner can use it when the context changes. Transfer is an outcome of how the learner was engaged, not of what was delivered.
Second, research on motivation and self-determination shows that the same content produces radically different outcomes depending on whether the learner experiences autonomy, competence, and relatedness (Deci & Ryan, 2000). A recommender that optimises for engagement metrics often erodes autonomy; a system that grades everything erodes competence; a system with no human presence erodes relatedness. The content can be perfect and the learning can still fail.
Third, research on implementation and translation — the so-called "17-year gap" between a finding entering the literature and the same finding entering classroom practice (Morris et al., 2011) — shows that the bottleneck is almost never the content itself. It is the architecture around the content: the professional development, the assessment, the governance, the incentives, the equipment, the time.
Put together, these three bodies of work say something the industry has been slow to admit: the problem of learning is not a content problem. It is a whole-person, whole-system problem, and content is one input among many.
The Framework
This series walks that argument through 76 theories, models and frameworks, organised into three acts.
Act I traces how we came to think about learning in the first place — the philosophical, psychological and political assumptions that still shape what a lesson plan looks like today.
Act II covers how we tried to build, measure and study that thinking — the instructional design methods, assessment models, and research frameworks that turned theory into practice, with all the compromises that entailed.
Act III is about where we are now, and what is going wrong: adaptive systems, AI-driven personalisation, governance gaps, and the quiet failures that accumulate when we pretend the content-delivery equation is enough.
The companion article you are reading now is the first of five meta articles that sit between the acts. They do not analyse a single theory. They step back and ask the questions the rest of the series takes for granted: where did these ideas come from, why do they persist, and what should we do with them now.
The Invitation
If you work in education, instructional design, EdTech, or AI policy, the stakes of this distinction are practical. A system built on the content-delivery equation will keep producing dashboards that look healthy while the learners on the other side of the screen quietly disengage. A system built on a whole-person view will look messier in the short term and more humane in the long term.
The point of this series is not to convince you that adaptive learning is bad. It is to give you a vocabulary — 76 of them, in fact — for telling the difference between a system that is actually helping someone learn and a system that is simply delivering things.
What would change, in the next platform you help design or procure, if you stopped treating content delivery and learning as the same thing?
# ACT I — HOW WE CAME TO THINK ABOUT LEARNING