Lost between two lectures
When I was a student there were no slides and no recordings. There was a lecturer, a pen, and whatever you managed to get down before the next sentence arrived.
What I did with those notes afterwards took admittedly a fair amount of time, but it proved effective for me. I typed my notes up on a PC. I had not used a PC at school, so owning one and being able to use it for uni was a novelty that I enjoyed. I was also able to type reasonably quickly. But what made the method unusually effective was that I could add material. As part of revision I would consult textbooks, sometimes several of them, adding material that was missing, correcting things I had misunderstood and filling in additional information. Over the years this turned into something I had not quite planned: my own textbooks, written in my own words, which I went on using long after the exams were over.
However, very soon I found a problem with the approach.
The sentence I could not place
The degree was not as structured as current courses, so topics did not appear once and were done with. The same ground came back, months or years apart, in a different context taught with a different emphasis. Which is how I ended up with two versions of the same statement in my notes, one I typed up earlier either from lecture notes or a book, the other freshly taught. The problem was: one statement excluded the other. So which one was correct and which one was wrong? I tried my older lecture notes, but could not find the statement in the form I had written it down. I tried the textbooks I commonly used – also with no success. It bugged me, because both versions had wider implications for understanding the topic, and one clearly was wrong.

© RudolphLAB, 2026
Checking this was not as simple as it would be today. The internet was still in its infancy, so drilling down to the original facts took considerably more effort. But the indignation about having to spend even more time on typing up notes turned into a very useful habit: I started noting down, meticulously, where each section had come from. Lecture, and which one, or a specific textbook. Nothing grander than that. It was bookkeeping, done for entirely selfish reasons, so that I could go back to where I had acquired the information and check it for accuracy.
The sources that did not matter
What I find interesting, looking back, is which sources they were. Lecture notes. General textbooks. These are precisely the sources we now discourage students from citing in their assignments. And yet it did not matter, in that context, what the source was. What mattered enormously was that I could go back and find where a claim had come from, which allowed me to verify it independently when necessary.
What the habit was actually doing
I have only recently realised that the bookkeeping had a second effect, one I certainly did not intend. Because the same topics kept reappearing across different lectures and textbooks, maintaining and revising these notes forced me to revisit related material in different contexts, over and over, rather than treating each course as a self-contained block.
To put this into context: like many students, I struggled with biochemistry. The enormous amounts of substrates, products, enzyme names – I never saw the point of learning all this material by heart. But for the exams we had to. This changed entirely when I significantly advanced my organic chemistry knowledge: by learning reaction mechanisms, I realised I did not need to learn all the structures. I needed to learn exactly one, the starting point. Then I would learn the precise chemical reactions that would take place, which allowed me to develop the entire pathway, rather than learning it by heart. One topic informing the other. For me it made things a lot easier.
Before this becomes too neat a moral: it caught me out in the final oral examinations. The examiner wanted me to jot down a pathway rather quickly to talk about something related to it. He took the fact that I had to think through each step as weakness, and biochemistry turned out to be my worst exam result.
Still, the point stands that I was, without knowing the word for it, revisiting related material from different blocks rather than studying each one in isolation. There is a substantial cognitive-psychology literature showing that interleaving related material can improve learning and later discrimination compared with studying each type in a separate block, although the benefit depends on the material and the task. I did not design this. I stumbled into it as a side effect, realising that some of the topics only briefly mentioned in one content block were indeed the same as some of the topics in other content blocks, just with more detail. And the computer made it easy to join these up, replacing shorter and more superficial sections with the deeper and more detailed content from more specialised lectures.
When I later started working with primary papers, proper referencing turned out to be easy, too. Not because I had been taught the conventions particularly well, but because a reference was just a more formal version of something I was already doing. "Molecular Genetics, Lecture 7" became a journal, a volume and a page. The reflex was the same. Only the standard of the source, and the stakes, had changed.
The rule without the reflex
Today we do it the other way round. Students are told, from early on and quite rightly, to use peer-reviewed literature, and they do. What most of them have never had is a reason of their own to develop the underlying reflex – the habit of actually going back and checking where a claim came from. The rule arrives without the disorientation that would have made it make sense.
The results are often quite amusing. A citation appears in the right place, in the right format, from a properly peer-reviewed journal, and has nothing whatever to do with the sentence it is attached to. Nobody is cheating. The student has done exactly what they were asked: attach a good source. Retracing a claim back to where the underlying result was established was never quite part of the request. The citation is a token that proves compliance, rather than a route back to the actual experimental evidence.
Showing them does not help
I have tried to address this directly. I run a seminar in which I show students that something they have learned rests on data that do not exist. It is a memorable session. As far as I can tell, it does not really change anything.
I think I know why. I did the digging, and I hand over the punchline. They understand my finding, but they do not get disoriented by their own. What worked for me was not being shown a problem. It was running into one, on material I cared about, with nobody watching, and with a real cost attached to not knowing where something had come from.
Which is the awkward part. That cannot be assigned. The moment it becomes an exercise, it is a hoop, and hoops teach you to jump, not to ask why the hoop is there.
An open question
Of course one person's route is not evidence, and it is entirely possible that I am describing a habit that would have formed for me under almost any circumstances. But I do notice one thing. My habit grew on forgiving ground. Lecture notes and general textbooks tolerate being got wrong. You can lose track of them, get confused, suffer a little, and correct course, and none of it counts towards anything. Insisting on primary literature from the very start is rigorous, and defensible. It may also be removing the low-stakes confusion in which the underlying reflex – and, as a bonus nobody designed for, some genuinely useful cross-topic revision – used to have room to form.
I do not have a fix. I am not sure one can be designed; the moment it is designed, it stops working, for the same reason my seminar does not quite work. Perhaps the most a teacher can do is arrange for the right kind of accident to be a little more likely, and then get out of the way.
How do you let someone get usefully lost, in a system that has been built so that nobody does?