The Archive

I have more or less given up on the Mac. The Mac Mini will probably not be updated again, and running ConnectedText on Parallels became more and more difficult. (More precisely Parallels became more difficult. Every time my Internet connection broke, Parallels would not start. I eventually got it to work, but I never understood what it was that made it work. My best guess, it had to do with the keychain function of OSX). But whatever may have been the reason, I now use a NUC7i5BNK with 8 GB of memory and Windows 10 (64 bits) and have no problems.

I can still run the Mac mini I have (It's hooked up to the same monitor as the NUC), but I rarely do. Therefore, I have not tried The Archive and I cannot review it myself. But there is a very informative review on Welcome to Iherwood. As far as I can see, the most important feature of The Archive is

The Archive will automatically set a unique number to each note you create in the format of yearmonthdayhourminute that the note was first made. Call that the note ID. You can append a note title to provide a clue as to the content of the note. Together those will make up the file name of the note — each note is saved as a separate plain text file in the designated folder.

As people who regularly read this blog know, I think the unique number is superfluous in modern database systems. I use direct links to entries (in ConnectedText). Luhmann who used unique numbers in his paper-based system, thought the numbers were in themselves beneficial because they allowed for branching and continuation of notes.[1] He may have been right, but the simple unique numbering of the Archive does not seem to allow for this—unless there is something I miss.

It is, by the way, fairly easy to furnish a program like ConnectedText with this capability through AtoHotkey. I think that some users have adopted "yearmonthdayhourminutesecond for this purpose (to allow for notes that are only seconds apart).

As I said, I don't find I need it, but I understand that different people have different needs.

I recommend the review.[2] And, on the base of it, the program (as long as you understand what you get into.[3]


1. See also Different Kinds of Links on this blog.
2. One nitpick: Luhmann did not "invent" or "develop" the slipbox. He only developed a numbering system for the notes (that is most certainly intriguing.
3. But see also the newest post (on the three layer structure of notes in: The Archive. "My archive became opaque like the sea: You can see a couple inches into the deep but you know there is much more that you can’t access. You can dive deep, but still you just see a couple of inches at any time. Therefore, I thought of it in terms of unexplored territory for which I need mapping methods and such." It would be my claim that direct links to topic names would be less opaque, even though it also benefits from structural notes.

Luhmann's Alternative to the Zettelkasten

I recently bought to volumes of late Luhmann interviews, namely Niklas Luhmann, Archimedes und wir (Berlin: Merve Verlag, 1987) and Wolfgang Hagen, ed. Warum haben Sie keinen Fernseher, Herr Lumann? Letzte Gespräche mit Niklas Luhmann (Berlin: Kulturverlag Kadmos, 2004). In the latter volume, in the interview with Wolfgang Hagen, entitled "Die Ralität der Massenmedien" (pp. 79-107), Luhmann reveals that he has an alternative to the Zettelkasten. Hagen notess: "[The Zettelkasten] is, so to speak, the basis of your work. If it were taken from you, then it [your work, that is] would be difficult. Luhmann answers:

Yes, it would be difficult. But I now have an alternative, namely half-finished book manuscripts which are found in boxes under the desk, and then I can of course ... When I discover something somehow interesting for "sovereignty," I can write it directly into the manuscript where "state" and "soverignty" is discusse.

It is interesting to ask how different this alternative really is, since you could also look at the Zettelkasten as a collection consisting of manuscripts that are in varying degrees unfinished.

Luhmann would not be the only one with such Zettelmanuskripts. Nabokov and Arno Scmidt used such a method as well.

In this incarnation, Luhmann technique is perhaps closer to working with Scrivener or Ulysses.

Different Kinds of Links

My own implementation of a Zettelkasten does not make a distinction between continuing topics (or weiterführende Zettel) and branching ones (also called ergänzende Zettel). I did not think that this was important for two reasons
1. an electronic topic or Zettel can be as long as I want it to be, and
2. it does not seem to be really important for the praxis of taking note to differentiate between supplementation or branching and linear continuation. For one thing, you can always add new stuff to any particular topic as you need to. No acrobatics in numbering seem necessary. And linear sequence is in some sense just as much a link as is branching off. [[Kladde|Heft]]

But what if I wanted to make such a distinction between different sequences? It is, of course possible. Linear sequences can be indicated by either linking the end of one topic to the topic that follows. (You could use something like [[following topic|blah, blah, blah]]. You could, of course, also use the name of the topic followed by a number from 1 to n. And you could reserve textual links for branching topics.

There are other ways of doing this, as with properties, for instance.

The only question is whether it would be worth my while. I am not sure it is, but I am willing to be convinced that it is.[1]


1. Luhmann's numbering system represented, as it were, the exoskeleton of the Zettelkasten. A wiki database does not need it, as any entry in it represents a fixed location in a database. Textual implementations, like nvALT and other do not have this advantage. If you rename either the link or the linked item, the connection is broken. They might need an exoskeleton as well.

Zettelkasten Nonsense

Just read this, but won't give you the link because that would defeat the purpose. You can search for it, if you really want to see it. In any case, I found a blatant piece of distortion that MUST be based on complete ignorance about Luhmann's Zettelkasten:

Luhmann war ein sammelnder Denker und denkender Sammler. Er begann in den 1950er Jahren einen Zettelkasten aufzubauen, der zu einem rasant anwachsenden, dynamischen Katalog mit einem Register wurde. Mithilfe manövrierbarer Zettel setzte er unterschiedliche Begriffe, Theorien und historische Dokumente in Beziehung, verschob sie und arrangierte sie wieder neu. Am Ende umfasste der Kasten insgesamt 90.000 Zettel.

Luhmann's Kasten was not about moving and arranging topics. It was about giving them a fixed positions. Nor did Luhmann understand himself as a "mere" collector. Rather than being an unsystematic "collecting thinker and a thinking collector", he was a systems thinker—to say nothing about the "lightning fast" increase of the number of Zettel and the "historic documents" found among them. These fundamental mistake may make no difference to the advertisers of office equipment, but even a little research might have made a difference to accuracy.

An Ordinary Zettlkasten

This is an ordinary commercial product for a Zettelksten (by Herlitz). It's nothing magical.

Nor is Luhmann's Zettelkasten magical. He wrote on Zettel 9/8,3: "Menschen kommen. Sie bekommen alles zu sehen, und nichts als das—wie beim Pornofilm. Und entsprechend ist die Enttäuschung." People come [to look at the Zettlkasten]. They get to see everything, and no more than that-just as in a pornographic movie. And they are equally disappointed.

On Zettel 9/8a2 he called the Zettelkasten "eine Klärgrube" or a "septic tank;" (perhaps even "cesspool"). Waste goes in, and gets separated from the clearer stuff.

Also Zettel 9/8i "Zettelkasten mit dem komplizierten Verdauungssystem eines Wiederkäuers. Alle arbiträren Einfälle, die Zufälle der Lektüre können eingebracht werden. Es entscheidet dann die interne Anschlussfähigkeit." Or: "The Zettelkasten [is] like the complicated digestive system of a ruminant animal. All arbitrary ideas, the accidents of reading can be brought in. What afterwards decides is their internal capability of connecting up."

See Luhmann Handbuch.

The Card Index System

The card index system; its principles, uses, operation, and component parts is a 1911 book by one R. B. Byles on the advantages of card indexes for businesses. It's an interesting read, if you are interested in the history of this technology.

The Introduction starts with the claim that "[r]oughly speaking, the world is divided into two classes: those who use the Card Index System and those who do not." The short book is meant to convince those who do not use the system to change their minds.

Gershom Scholem's Zettelkasten

See also Guide to Using the Card Index and Gershom Scholem's Milon HaZohar Card Index – Yehuda Liebes. We are told that "[t]he notes are in the form of white cards, which Scholem stored in impeccable order in a long narrow wooden drawer that fit them perfectly, in his impressive writing desk (which today serves as the Scholem collections librarian's desk at the National Library)." It would be nice to know the format of the "cards" or, perhaps better, "slips."

It is more than likely that he learned this technique before he changed "Gerhard" to "Gershom."

Lesage's Notes on Playing Cards

Georges-Louis Lesage (1724-1803), like some of his contemporaries, took notes on play cards.[1] Here is one card:



Two other persons who used playing cards are Gibbon and Rousseau.



1. See here

Luhmann's Zettelproblem

This photograph illustrates one of the limitations of a paper-based Zettelkasten. You could say it illustrates only the problems created by a cheapskate like Luhmann. For more on this, see Luhmann's Frugality.

Hermann Burger, Lokalbericht

Hermann Burger (1942–1989) wrote a novel in 1970 called Lokalbericht (1970). It appeared in October 2016. I will buy it and read it. The protagonist of the novel Günter Frischknecht (a composite name made up of Günter Grass and Max Frisch) is writing somewhere in the county of Ticino at the same time a dissertation and a novel by means of two different Zettelkästen.[1] They get mixed up and the slips intermingle. What to do in this situation? Reality and irreality apparently can no longer be distinguished. The novels conceit reminds me of Sterne's Tristram Shandy (and apparently similar situations abound).

I must read it! Ordered it already!


1. "Knecht" in the name may refer to the magister ludi in Hesse's Glass Bead Game.

"Zettelkasten" in Grimm's Wörterbuch

Grimm's Wörterbuch is one of the most interesting early dictionaries of German. Its origins go back to the first half of the nineteenth century and gives a good indication of how German words were used in the eighteenth's century. Here is the entry for "Zettelkasten:"

Zettelkasten, m., kasten zur aufnahme von zetteln, z. b. von theaterzetteln: S. Hensel familie Mendelssohn 3, 22, zumeist aber von alphabetisch geordneten zetteln mit notierungen oder auszügen aus literar. oder wissenschaftl. werken od. ä.: leben des Quintus Fixlein, aus funfzehn -kästen gezogen Jean Paul w. 3, 3 H.; die hier angedeutete schaffensart wird vielfach getadelt, so von Immermann 20, 36 B., von Fr. Th. Vischer altes u. neues 3, 388; Hebbel III 7, 397 W.

Meaning:

box for keeping of slips, like playbills, see Hensel, The Mendelssohn Family 3, 22; but usually [it means] alphabetically ordered slips with notes or excerpts from literary, scientific, or similar works. The Life of Quintus Fixlein, Drawn from fifteen Zettelkästen. The way of working here indicated is often criticized, like in Immermann 29, 36, in Fischer's Old and New 3, 388, in Hebbel, iii, 7 397 W

So "Zettelkasten" already has negative connotations in the early nineteenth century.

What can we learn from Grimm about "Zettel"? They were originally called "Zeddel" which simply meant small pieces of paper (slips). The word derives from the Italian "cedola," which came from the Middle Latin "cedula" and ultimately from the older "schedula." The Latin "scheda" or "scida" means torn off strips of paper [must mean "papyrus, I think}. It all originally come from the Greek "σχίδη."

Nottingham, Again

Nottingham 3 is an update of a clone of Notational Velocity. It's explicit about its debt: "In the spirit of Notational Velocity, Nottingham is an elegant notepad for macOS. Designed to be lightweight and easy to use, it's a breath of fresh air perfect for storing whatever is on your mind." But the claim is that "Version 3.0 is completely rewritten from the ground up in Swift 3 to take full advantage of macOS's newest features and to lay a solid foundation for future improvements." And it promises to be fast: "Lightning-fast search across even the largest library of notes." I wonder what the "largest library" is. 100,000 notes, 1,000,000? It would be nice to know.

It now also claims to be able to do "wikilinks." But this is at best a half-truth. It can link to existing notes, and it seem to be able to keep track of the name-changes of those notes, but it cannot create new notes from a typed link.[1] Notational Velocity and nvALT can't do that either (and in addition they cannot keep track of name-changes), but they are free.[2] Nottingham costs $14.95 (though you can try it out for free).

I dislike the name. Searching for "Nottingham" gets you a lot about a city in in the East Midlands of England, getting to the application is harder. It's a complaint about some applications I have for a long time. Try "Brainstorm," for instance!


1. It does not do so reliably, however. There seems to be a problem with the program. I have deleted it again from my machine!
2. Actually nVALT can create a new link from an expression enclosed in double square backets. Clicking on an expression enclosed in double brackets, puts it into the search window, pressing enter creates the topic. nvALT does not, however, keep track of name changes.

Luhmann's Zettelkasten Described

Johannes Schmidt: Der Zettelkasten als Zweitgedächtnis Niklas Luhmanns is a video of a lecture given by Johannes Schmidt that provides a very detailed description of Luhmann's Zettelkastn. I learned many new details (without having to revise my earlier views).

Highly recommnded!

Warning: The lecture is in German.

CintaNotes Limits, Again

I had not tested CintaNotes but relied on someone else's testimony when I wrote the previous post on it. So, I thought I would actually test it with my approximately 10,600 text files from ConnectedText.

The short of it: I couldn't. CintaNotes uses its own file format (*.db). You can import db, and xml files, no txt files. It has a selection for "All files", but it does not recognize txt files. So I could not import the notes I used in ResophNotes and nVALT. This means that I could not test CintaNotes due to its limitations in import.

For what it's worth I used an SSD drive for all the tests. My notes are small, most of them 500 words or less, and I did not even test search in ResophNotes. Just moving around between different files and trying to change a word or two was too slow already.

I was going to try xml import, but then I found in the help file: "You can import notes using the File/Import command. You can import notes from previously exported XML files and from other notebook (.db) files." So, there does not seem to be a way to import any notes taken in any other application before. This makes the application useless to me (and, I suppose, to many other people who would like to migrate to CintNotes from any other note-taking application.) I have, accordingly, deleted CintNotes from my machine again.

Let me just say that the inability to import text files seems to me a more serious limitation than ReophNotes limited ability to handle a large number of text files.

ResophNotes Limits

I tried ResophNotes yesterday on the more than ten thousand notes exported from ConnectedText. It worked, but not very well. It took a while for notes to open. Clicking on a link eventually opened the target note, but it took unacceptably long (even though I have a quad-core processor). Christian Tietze reports that this behavior starts with about 1500 notes (see his comment on the previous post). I would suppose that this slow-down depends at least to some extent on the processor and memory specification of your system, but I also think that it is unacceptable for a large body of notes. More than 20,000 notes are not unheard of for a life-time of note-taking, even if you do not just copy web pages but take careful notes. Thus, Jules Verne amassed 20,000, Luhmann seems to have ended up with about 90,000. Roland Barthes with 12,0000, and Hans Blumenberg with 30,000.

In any case, it does not seem unreasonable to expect a robust note-taking or Zettelkasten-system to handle a large number of notes (with headroom to spare).

My Translations of Luhmann

“Communicating with Slip Boxes” (which is about Luhmann's own Zettelkasten) and “Learning how to Read” still can be found online here. See also this post.

One Way I Organize My Notes

As everyone who reads this blog knows, I use ConnectedText to organize my notes. ConnectedText is a "personal hypertext." Accordingly, links between notes play a large role. They are very important to me. However, they are not sufficient. I also use categories or tags to classify them. In my main project that has more than 10,000 topics, I have 223 categories or tags (ConnectedText does not sharply differentiate between the two). There are also properties and attributes which I use very sparingly.

One fundamental organizational device I use is what is called a "smart topic" in ConnectedText. It may be thought of as the "reverse" of a wiki category, and it is just a topic with an embedded search. So, if I would like to have an easy way of knowing how many topics I have on the medieval university and its institutions, I could create a topic called "medieval university" and put the following inline query into it: "[[$ASK:medieval and university|Index|]]" and it will list every topic on the subject. (If you are curious, there are 24).

Over the years, I have found three types of smart topics particularly useful, one belongs to the category "Person," another to the category "Concept," and the final one to the category "Theory." So, if I want to know how many and what topics I have, say on A. J. Ayer, I just put "[[$ASK:Ayer|Index|]]" into my Ayer page. If I want to know how many and what topics I have on the concept of "order," I do the equivalent for it, and if I want to know about "string theory" the same. (I assure you that there are not many notes on the last topic [I checked and found exactly 2]).

Some people, including Luhmann, think that classifying one's notes by persons is a deficient way of achieving order. As a Systems theorist, he is somewhat disdainful of this approach:

One possibility is to remember names: Marx, Freud, Giddens, Bourdieu, etc. Obviously most knowledge can also be ordered by names, eventually also by names of theories such as social phenomenology, theory of reception in the literary disciplines, etc. Even introductions to sociology and basic texts are conceived in this way. What one cannot learn from such works, however, are conceptual connections and especially the nature of the problems that these texts try to solve. Still, even candidates in exams at the end of their studies want to be examined on Max Weber or, if that is too much, on Humberto Maturana, and they are prepared to report on what they know about these authors.

Luhmann clearly thinks this is sub-par for the course. I don't disagree. That is why I also sort things according to concepts.

One might argue that this is still deficient. As Karl Popper claimed, "Theories may be true or false. Concepts can at best be adequate and at worst be misleading. Concepts are unimportant, in comparison with theories." And this is why I also have a category named "theory" under which I can find smart topics with searches for particular theories.

This is one way I organize my note-taking. It is by no means the only one, as I suggested at the very beginning.

History of the Index Card?

See here. I am rather doubtful about this history. Thus, it is not really true that Linné (Linnaeus) "invented" the index card. For one thing, he used paper slips, and for another, playing cards and paper slips were used already for the purpose of cataloging and taking notes long before him. What about Aldrovandi or Gessner as the "inventors" of the index card in the sixteenth century, or about 100 years earlier?[1]



This criticism of the "history of the index card" does, of course, not diminish Linnaeus's other achievements in any way. The article has other problems as well.

A propos Linnaeus, Weinberger (2007) claims that "Linnaeus's organization took the shape it did in part because he constructed it out of paper." "he used paper - atoms - to think through the order of the natural world." Whether this kind of reductionism can be maintained is also highly questionable - or so I would claim. This seems to be at best an over-simplifiction. It may well be that paper cards were a necessary condition in the eighteenth century, but that does not mean that they were also sufficient conditions. To his credit, Weinberger says "in part." Others have made such reductionist claims without any qualification.


1. Let me be clear: I don't claim that they invented the index card either. I am just claiming that they have just as much right as Linnaeus. Index cards were invented much later.

Peirce's Note Cards

Charles Sanders Peirce was an early adopter of the index card system. He used 3 x 5 and 2 X 5 cards between 1886 and 1891 for his drafts of definitions for the Century Dictionary and Cyclopedia. Many have been lost, but there are four large boxes extant with several thousands of cards.

Info-Base 4.8

A new version of Info-Base with minor improvements has been released. From the Website:

• very fast (also with very large amount of data) and effectively for any kind of data ...
• the ideal "Zettelkasten" ...

I say: Info-Base is in many ways what the newer versions of InfoSelect should have been. It preserves the spirit of InfoSelect 1.0.