Recommends a zettelkasten system for note-taking.
About Luhmann and his slip box. Ahrens asserts, but does not here show, that the slip box is different in some important way from a wiki.
How do you actually use a zettelkasten to produce a book (or paper, blog post, etc.)? Let's work backward from the final product:
So, how does this work in the forward direction?
Any tool must be used properly.
In school, writing is the final step–the culmination of planning and research, after the problem has been chosen and an angle of attack determined.
This idea should be not merely inverted, but totally changed: writing should not be the last step, or even the first step, but the only step. When studying, you take notes; when thinking, you arrange your thoughts deliberately and record interesting ideas; you select a problem area that reveals itself from your notes; and in the end, indeed, you produce a piece of finished writing from all this. Writing pervades the process, because writing is both the ultimate goal and the means to accomplish that goal.
So, take this attitude: nothing counts other than writing.
The slip-box must have not only many notes, but good quality notes properly integrated. Three types of notes should be distinguished:
Do not attempt to capture all interesting-seeming information indiscriminately; the volume will be overwhelming and the majority useless.
Do not take notes tightly bound to a single, pre-determined project; they will be less useful in the future, and too many good avenues of inquiry will be avoided for the sake of maintaining focus. Attempting to mitigate this by maintaining many such projects in parallel will simply create an explosion of overhead, preventing any work from getting done.
Finally, do not treat all notes as fleeting notes; that way lies chaos.
These mistakes all lead to one conclusion: the more notes you take, the worse off you are. A terrible result!
You can't pick a topic to write about without first knowing something in the subject area, so the idea that you pick a topic first and only later begin doing research is obviously false.
At this point in the book, I feel like there's been a lot of "you should really trust that this method is the best" but not so much concrete explanation of exactly what the method is or how to apply it.
What, for example, should a permanent note look like? Is a permanent note a miniature essay of a paragraph or two? Where in this scheme do the kind of notes taken as memory aids–lecture notes, etc.–fall? Are these fleeting notes? Literature notes?
The details of this scheme need to be made more explicit.
A good feedback loop is essential to learning. Writing zettels provides immediate feedback on the state of one's understanding, and a very specific and low-stakes way.
Even though this study was never published, makes no claims about intelligence and is statistically irrelevant, it does seem to confirm what most of us believe anyway…
There are three limited resources involved in writing: attention, short-term memory, and willpower.
Writing is a process made up of many individual tasks: proofreading, organizing the outline of the document, etc. We cannot focus on more than one thing at a time, so it is a waste of effort to attempt to do all of these at once. Rather, it is better to focus on each stage, one at a time, to give each the amount, as well as the kind, of attention it requires.
Our short term memory holds only a small number of items–seven is the traditional figure, though modern estimates are lower, perhaps four (this was also discussed in Human-Computer Interaction I: Fundamentals & Design Principles). Writing things down so that they can be dealt with later gets them out of short term memory, freeing it for the things we are currently working on.
Willpower is spent making decisions of all kinds–both trivial and momentous (this is referred to as ego depletion), so it is beneficial to arrange the workflow to eliminate needless decisions. Thus, the same note-taking method is used when reading, there is only one place to store notes, and notes are all retrieved in the same way. Furthermore, by keeping notes of all important things (not relying on short-term memory), it is possible to take breaks, which restores willpower, as well as allowing the brain time to process what is being learnt. Taking breaks when needed makes learning more efficient.
Ahrens is very insistent that "the only thing that matters is […] the writing of the actual slip-box notes", and that literature notes should be taken with this as the sole end. But this doesn't seem like the kind of notes you would take–or would want to take–when you're learning something new.
If the book is fairly easy reading, then my notes look like Ahrens suggests: they will usually be a simple list of page numbers and keywords, which serve as either an index of interesting things in the book, or a note to do further research. However, if I'm reading a textbook on a subject on which I'm a beginner, I might take much more extensive notes, do the exercises, etc. Ahrens allows that when reading, you should take whatever kinds of notes work for you, but it's clear that note-taking for any purpose other than later writing slip-box notes is disrecommended.
The problem here is that at this stage I am simply not prepared to make useful, permanent, slip-box notes. During the early stages of learning, any original observations that I am equipped to make will be shallow, at best. Of course, I want to form connections with my existing knowledge–that's how learning works–but writing an interesting, original, standalone note? That's an unreasonable goal, when you're starting out in a new area.
This is closely related to the other major issue with zettelkasten: bootstrapping the system is hard. "Only write a note when you can connect it to other notes," they say, so how do I begin writing the first note? Or the hundredth, for that matter? Unless your area of interest is very narrowly defined, you are not going to be able to actually link together notes in this way until you have a lot of them. When learning something novel, your notes are also likely to form a subgraph disjoint from the rest, or one with few bridges into the rest of the notes, so the system won't work as intended. How do you address this, in practice?
Context is required for understanding. The slip-box enables us to learn by encoding knowledge in a way that is optimized for retrieval.
Ahrens develops the idea that we should optimize notes for retrieval: links (and similarly keywords, incoming links from indices) should not merely categorize the note, they should present the note at a time when we're likely to want it.
Having concrete notes can help us to spot ideas we've had before, or ideas that belong to someone else, when they later come to mind. They help to spot the differences and similarities between ideas.
This is another chapter that basically just says "the slip-box does all the hard work for you!" and gives many varied examples of vaguely-related stuff the convince you that there's some chance of it being true.
Ahrens instructs that you should narrow ideas into a range suitable for the writing project, allow the structure to change as needed, however seems fruitful, and to write multiple things at once, switching projects when you feel stuck. I think that's about all the practical advice given.
Ahrens writes: "It is not so difficult to get used to thinking within an external memory of notes, as the advantages become obvious quite quickly." But Luhmann said that it requires a number of years for the slip-box to be so useful (KuehnLuhmann2015a).