The zettelkasten method of note-taking has received a lot of interest in the last couple years. This interest has sparked a number of debates among adherents to the method about particular aspects of its implementation. I wrote about one of those debates a few months ago–i.e. the debate about the role of so-called folgezettel.
Another such debate revolves around whether or not we should take and store quotations in our zettelkasten. It’s common for advocates of the zettelkasten method to answer in the negative. In doing so, they often point to the fact that Niklas Luhmann did not tend to take a lot of quotations, instead opting to write paraphrases of what he read in his own words. In this post, I’d like to argue in favor of taking quotations. To do so, I will draw on my own experience, as well as the advice and reasoning of some folks who have been rather influential with respect to research methods.
Before we proceed, however, I will offer a caveat: As always, I am not arguing that the way I do things is the one true way. As the kids say these days, "you do you." Or, as Sascha Fast says, "don’t dehorsify the horse." Indeed, part of my argument against the advice to not take quotations is that it too often relies on an orthodox attempt to follow Luhmann’s particular implementation. But Luhmann is neither the inventor nor the sole historical user of the zettelkasten method. Nor was he perfect in his work (a point I’ll return to below). His project is not yours or mine, just as my needs are likely not the same as yours. Do what is best for your work. That said, I do think there is a good argument to be made for taking quotations and, as you’ll see, you don’t just need to take my word for it.
Why I (mostly) take quotations for my zettelkasten
My history with the method began in high school when I was competing in policy debate. Taking quotations from your sources was required for creating evidence "cards" to be used in support of your arguments. That’s still the best practice even today, though there is much less scissor, tape, and paper action involved. Likewise, in my AP/Honors English and history classes we learned and were required to use the notecard method. In that case, the taking of quotations was not required. Paraphrases were allowed and since we were hand writing the notes on 4×6 cards, it was much more labor intensive to take quotations.
Nonetheless, if given a choice, I would default to the method I learned in debate. Even though we weren’t writing the quotations out by hand, but rather, photocopying, marking, cutting with scissors, taping on paper, writing summary lines, and then re-cutting and re-taping as necessary to build outlines of our arguments, this method was still considerably more work than the notecard method. I prefered it, however, and still do (minus all the tape and paper, that is). The reason for that preference remains the same too.
Taking a quote or writing a paraphrase in one’s own words are often presented as an either/or situation. But, as we would’ve said in debate, these two alternatives are not mutually exclusive, so "perm, do both." In fact, that’s what we did and what I still do in most cases to this day. And it’s considerably easier, almost trivial, with computers and digital content to do both.
But, of course, just because I’ve always done it that way and just because it’s relatively easy to do doesn’t mean it’s a good idea. The other reason that I default to taking a quote if possible is because it keeps me closer to the original material from which I am drawing. I try my best to represent fathfully in my own words the point that the author was trying to make. But there are times when I re-read my summary later against the original quote and realize that I was putting my own thoughts and words into the author’s mouth or, even worse, misprepresenting the author’s point entirely. Without having that quote, I wouldn’t have caught my error.
And it’s not just me. None of us is perfect. That inlcuded Niklas Luhmann. In fact, he was criticized at times for misrepresenting other people’s ideas and arguments in his own work. Johannes Schmidt, in his essay about Luhmann’s zettelkasten in Alberto Cevolini’s edited collection, Forgetting Machines: Knowledge Management Evolution in Early Modern Europe, writes:
Luhmann jotted down only a few keywords in the course of his reading along with the respective page numbers, some of which he also wrote on the back of the cards containing bibliographical information in the second collection. These notes are extremely brief and are not really excerpts in the strict sense – for instance, the notes from reading an entire book frequently fit onto one of these cards… Instead of giving an exact account of what he had read, Luhmann made notes on what came to his mind in the process of reading, with an eye to the notes already contained in his file.
Schmidt notes, however, that
Perhaps, this circumstance also explains the phenomenon familiar to many Luhmann readers that it is often difficult to find the argument in the publications by other authors that Luhmann cited them for.
Taking quotations with our notes is a check against this kind of error. And for all that we might admire Luhmann’s work as I do–I often use and cite him for his distinction between first- and second-order observation–misrepresenting other people’s work is an error.
This is a point that was driven home for me once again when I took a seminar on qualitative research methods in graduate school (a topic I now teach at the undergraduate and graduate level at the University of Utah). One of the most well known qualitative research methods is grounded theory. I won’t go into detail on grounded theory except to say that the notion of groundedness refers to taking care that one’s insights remain grounded in one’s data. We should be able to trace our findings back to the memos we wrote along the way, which should themselves be informed by the data (e.g. field notes, interview transcripts, documents, etc.) that we carefully read, re-read, and coded. And those coded passages from our data should be traceable back to the original context from which they were extracted. Though most grounded theorists would likley balk at this metaphor, I tend to think of it in terms of maintaining as best we can a "chain of custody" for evidence in a criminal investigation.
So, to put it in zettelkasten terminology, my view is that our so-called "permanent notes," which I call "memos" (a la grounded theory) in my own system, should point back to our "literature notes," which in turn should point back to the original context by way of page numbers (when available) and a quotation. If I can pull it off without too much hassle–and having done this process for years with paper, scissors, and tape, my bar is pretty high for what counts as too much–this is my prefered way to work.
That said, there are times when I will make an exception. When have I sometimes not taken quotations? One instance was preparation for my comprehensive exams. I was up against a hard deadline and the process of OCRing text from books and articles was taking too long. Instead, for a significant number of sources on my reading lists, I wrote my own paraphrases with page references and then marked the orginal passage in the margin of the book or article. I then had all of them physically arranged in my office for quick reference if I needed to go back to the original text, which I did in many cases. Similarly, in putting together the literature review for my dissertation proposal, I took quotations from most sources, but not all of them, especially if it was a source that seemed fairly tangential and which was unlikely to get more than a passing mention and quick citation.
Again, when feasible, I prefer to do both. I take a quotation and also write a summary in my own words. With the summary, I try to be as faithful to the author’s point as possible. I try not to interject commentary or analysis of my own, instead just repeating in my own words the point of the passage. If I have commentary, I will add that after the summary and the quotation and I will clearly call out the commentary with "COMMENT: " followed by whatever my comment is. Sometimes it’s as little as one sentence, other times an entire paragraph. In the end, this gives me "atomic" notes consisting in many (but not all) cases of a summary, a quotation with page number/location if appropriate, and (sometimes) a comment.
In my writing, I often just make use of the summaries. But, that quotation is always there for context and to make sure I’m representing the author’s point correctly. Sometimes, if I have made a comment, that will end up informing what I write or becoming part of the writing itself. If the summary and comment are particularly salient, I might use both. For example, "John Smith has argued [SUMMARY] (Citation w/page number). Specifically he writes, [pared down excerpts from the QUOTATION as appropriate]. However, [COMMENT]." This is, of course, sort of an ideal case. It doesn’t always or even often work out this way in real life. Nonetheless, having each note "card" include a summary, quotation, and comment is extremely helpful.
Taking quotes for a zettelkasten has a long history
But you don’t have to take my word for it. There are plenty of other folks who have argued in favor of the kind of approach that I have just described.
A recent example can be found in writing from Professor Joel Chan from the College of Information Studies at the University of Maryland, College Park. He describes in detail his process for taking notes and how it maps onto concepts from the zettelkasten method. Most relevant for this discussion is his description of the relationship between what he calls "synthesis notes" (i.e. permanent notes in zettelkasten parlance), "observation notes" (i.e. literature notes), and "context snippets" (i.e. quotations). As in my approach, Prof. Chan’s observatrion notes are typically "grounded in at least one context snippet note" and his systhesis notes are supported by more than one observation note often drawn from more than one source. The context snippets grounding his observation notes can include "specific figures, data items, tables, or quotes that are the basis for an observation, as well as metadata (e.g., authors, year, publication) and methodological details that are important for understanding and evaluating an observation note." These context snippets serve to help substantiate and contextualize the observations recorded from our reading.
But the idea that taking quotations from our reading is important to grounding our observations is not new with Prof. Chan either. Indeed, this has been a common practice for centuries. Two examples serve to make the point. First, Mortimer Adler and Charles Van Doren in their classic, How to Read a Book, encourage the taking of quotations as a way of keeping ourselves grounded in the material, making sure we’re not misrepresenting what we’ve read. This is key, they say, to trying to be as objective as possible when doing what they call "syntopical reading," that is, reading on a particular topic across a number of sources and then synthesizing what we have read.
Following Adler and Van Doren’s advice, I will let them make the case in their own words and do so by sharing three of the "cards," "literature notes," or "observaiton notes" with "context snippets" that I took from my reading of their work and which I keep in my zettelkasten:
Always include a quotation in effort to be objective
Dialectical detachment or objectivity should, ideally, be maintained throughout. One way to insure this is always to accompany an interpretation of an author’s views on an issue with an actual quotation from his text.
Charles Van Doren Und Mortimer J. Adler, How to Read a Book (A Touchstone Book), Page 327 [Kindle ebook]
Must quote and summarize in effort to avoid bias
The syntopical reader must resist certain temptations and know his own mind. Perfect dialectical objectivity is not guaranteed by avoiding explicit judgments on the truth of conflicting opinions. Partiality can intrude in a variety of subtle ways—by the manner in which arguments are summarized, by shades of emphasis and neglect, by the tone of a question or the color of a passing remark, and by the order in which the various different answers to key questions are presented. In order to avoid some of these dangers, the conscientious syntopical reader may resort to one obvious device and use it as much as possible. That is, he must constantly refer back to the actual text of his authors, reading the relevant passages over and over; and, in presenting the results of his work to a wider audience, he must quote the opinion or argument of an author in the writer’s own language. Although it may appear to do so, this does not contradict what we said earlier about the necessity of finding a neutral terminology in which to analyze the problem. That necessity remains, and when summaries of an author’s argument are presented, they must be presented in that language and not the author’s. But the author’s own words, carefully quoted so as not to wrench them out of context, must accompany the summary, so that the reader can judge for himself whether the interpretation of the author is correct.
Charles Van Doren Und Mortimer J. Adler, How to Read a Book (A Touchstone Book), Page 316 [Kindle ebook]
COMMENT: Compare to critique of Luhmann’s sometimes partial readings.
Reading across passages from multiple texts tagged with same topic is instructive
Second, the collection of a number of passages on the same topic, but from different works and different authors, serves to sharpen the reader’s interpretation of each passage read. Sometimes, when passages from the same book are read in sequence and in the context of one another, each becomes clearer. Sometimes the meaning of each of a series of contrasting or conflicting passages from different books is accentuated when they are read against one another. And sometimes the passages from one author, by amplifying or commenting on the passages from another, materially help the reader’s understanding of the second author. Third, if syntopical reading is done on a number of different subjects, the fact that the same passage will often be found cited in the Syntopicon under two or more subjects will have its instructive effect. The passage has an amplitude of meaning that the reader will come to perceive as he interprets it somewhat differently in relation to different topics. Such multiple interpretation not only is a basic exercise in the art of reading but also tends to make the mind habitually alert to the many strains of meaning that any rich or complex passage can contain.
Charles Van Doren Und Mortimer J. Adler, How to Read a Book (A Touchstone Book), Page 323 [Kindle ebook]
Finally, there’s a good case to be made that the zettelkasten method is not only a scientific approach to observing and intepreting the world around us, but that taking quotations from our reading–which we’ll recall Prof. Chan describes as a part of our observation–is a key component of that process. No less than Francis Bacon, often credited as being the father of modern scientific method, advocated for the use of the method of commonplacing, a forerunner to the zettelkasten method. As Prof. Manfred Kuehn as argued,
Bacon felt that commonplace books might be useful instruments of the mind, especially if the commonplaces are collected from the nature of the world, and not from the concepts of Aristotelian philosophy. In fact, for Bacon the commonplace book is not a means of storing stock phrases and ideas in accordance with traditional "commonplaces," but a tool in the discovery of a new "interpretation of nature." It is for him not a "top-down," but a "bottom-up" approach that "arrives at the most general axioms last of all" (New Organon, 43). This is "the true way," which remained "untried," but which he recommended.
That is, Bacon advocated a slightly different approach to the standard method of commonplacing. Instead of using a list of “common” “places”—i.e. standardized categories for one’s index—he said that those categories should instead emerge from one’s observations of the world. This is sort of like the difference between deductive and inductive coding in qualitative research methods. Bacon was, in essence, arguing in favor of an inductive method of coding one’s notes on the world as a foundation for scientific innovation and discovery.
Those notes, Bacon said, were better when they included an excerpt, a quotation, from our reading. As Prof. Anne Blair explains in her excellent book, Too Much to Know: Managing Scholarly Information before the Modern Age,
Francis Bacon explained succinctly that notes could be made either “by epitome or abridgement” (that is, by summarizing the source) or “by heads or commonplaces” (that is, by copying a passage verbatim or nearly so and storing it in a notebook under a commonplace heading for later retrieval and use). Bacon considered the latter method “of far more profit and use,” and most note-taking advice focused on this practice of excerpting. Early modern pedagogues taught their pupils to copy choice passages they encountered in their reading into notebooks, sorted under topical headings called commonplaces (loci or topoi), and enjoined them to continue the practice as adults.
In this post, I have argued in favor of taking quotations as part of an implementation of the zettelkasten method. In my own work, I have found it a good practice to aid understanding and avoid as best I can misrepresenting others’ ideas. Sure, Luhmann mostly didn’t take quotations. But we do not have to copy him exactly. Besides, he has been criticized for precisely that which we seek to avoid by taking quotations–i.e. at times misrepresenting the ideas of others. Unlike in Luhmann’s day, it is no longer difficult to take quotations. In fact, in some cases like taking notes from a Kindle book, it’s easier to take quotations than not. So, the cost-benefit, I believe, is in favor of taking quotations. This is why it has been a recommended practice for so long in different fields, from Francis Bacon to grounded theorists to high school and college debate teams and many more besides. Again, however, you should do what works best for your particular situation. But do consider the value of taking and keep quotations.