You finish a book and take plenty of notes. Two months later you open them again. Every entry still makes sense on its own, yet together they no longer tell you what the book was about. Asked to explain it to someone, you would not know which note to start from.
Did you not take enough notes? Probably not. Most of what mattered is there. What is missing is something else: how those points were organised in the first place.
What the notes lost
In Organising knowledge with multi-level content, Francis Miller traces the trouble we have understanding knowledge, remembering it, and explaining it to others back to two problems.
Problem one: the structure is buried in the prose
Miller starts from the observation that all knowledge has structure. A description, a concept, a line of argument is made of parts, and the author deliberately arranges those parts into some structure in order to get the meaning across. Sometimes it is cause and effect, sometimes a classification, sometimes an argument that moves step by step towards a conclusion.
But books and notes can only be written one line after another. The structure has no place of its own. It hides between the lines, and the reader has to rebuild it from the linear text.
While reading, you did build a framework in your head: which question this chapter was answering, which claim this experiment supported. The notes, though, were taken in reading order, and the framework never made it onto the page. After a while, even you can no longer reconstruct it.
Problem two: the overview and the details share one level
In a good explanation, the upper level is the summary that pulls the threads together, and the lower level holds the details and secondary information. Miller argues that when these levels are not separated, or when one of them is simply missing, understanding suffers. He describes two cases: the summary is there but unmarked, drowned in detail; or there is no summary at all, only details.
A great many notes look exactly like this. A conclusion and an example sit side by side, with nothing to show which governs which. Or the pages fill up with examples and quotations without a single sentence saying what they add up to.
Put the two problems together and what the notes leave behind is a flat list of points. You cannot see how they are organised, and you cannot tell which ones are the big picture and which are the details.
What experts have more of is the connections between clusters
Miller’s thinking comes from the physicist Frederick Reif. In the second half of his career Reif turned to cognitive science and education research. In Applying Cognitive Science to Education he asks: given how limited our capacity to process information is, how can anyone cope with large bodies of knowledge? His answer is organisation. As Miller quotes him (p. 44), the organisation of knowledge “determines how readily the knowledge can be remembered, appropriately retrieved, kept consistent, communicated to others, and extended beyond its original scope.” So he argues that teaching should attend to the organisation of knowledge as carefully as it attends to the content.
Reif’s analogy is a map. Nobody draws a single map of the world with every street and every building on it. Maps come in levels: the world, countries, regions, cities, streets. Each level is made of a set of related content, which Reif calls a knowledge cluster, and a single point on one level can open into a whole map on the level below.
Miller adds one move to the analogy: zooming in and out, the way you do on a digital map. Zoom in for the details; zoom out for the whole; or pan across the same level to look at a neighbouring area. Learning a subject or reading a book, he says, is precisely this.
Miller puts it this way on page 49:
Experts are able to zoom in and zoom out fast because they know all the relevant clusters and how they are connected.
He goes on to explain that experts can locate the part of their knowledge that a problem calls for and choose a useful level of detail.
The point is not that experts know more. It is that they know which level each piece of knowledge sits on and what it connects to. So when a problem comes up, they can go straight to the level where it lives.
Miller describes learning as the same process: becoming familiar with more and more clusters; working out how those clusters connect, between levels and within a level; sorting out what is central and what is subordinate. Learning keeps moving between whole and part. Once you know the details of a part, the whole means more; once you know how a part relates to the whole, the part means more.
Back to those notes. They kept the pieces, but not the way the pieces group into clusters, and not the relations between the levels. Without that, there is no zooming in or out, only reading from the first page to the last. New material has nowhere to hang. The notes keep growing; the structure does not grow with them, and so understanding never accumulates.
Write the structure down with the content
If structure is part of understanding, it is worth recording alongside the content, rather than recording only the content and hoping the structure stays in your head.
That is the idea behind ZoomMap: a topic becomes a multi-level knowledge map.
- Organise the levels with nested cards. A card states one conclusion. Its elaboration, examples and evidence go into cards on the level below. The summary sits above the details instead of sharing a level with them.
- Organise the structure like a canvas. Cards on the same level can be placed freely, with lines marking cause, contrast and support, and a note on why they are connected.
- Move between levels without losing the structure. The whole map expands or collapses one level at a time. For the big picture, look at the top levels only; for the details, open the levels below. This is Miller’s zooming in and out.
- Back every card with sources. A conclusion on a card can link back to the passage it came from. Return after a while and you can still check what you understood at the time.
Try a small test first. Without opening your notes, explain a topic you have studied in three sentences. Those three sentences are the first level of the map. Wherever the words will not come, the structure is not yet clear.
See a real map
We turned Miller’s paper itself into a ZoomMap map. Start reading at the top level, pick one place to zoom into, then step back out and see where it sits in the whole.
Open the Organising knowledge with multi-level content knowledge map
References
- Francis Miller. Organising knowledge with multi-level content: Making knowledge easier to understand, remember and communicate. June 2018. See section 2.1 (pp. 43–46) for the discussion of Reif and section 2.2.2 (p. 49) for the passage on experts.
- Frederick Reif. Applying Cognitive Science to Education: Thinking and Learning in Scientific and Other Complex Domains. MIT Press, 2010 paperback edition. Reif’s ideas in this article are discussed through Miller’s paper; Miller cites pp. 142–143 and 161 of the book.

