"History Workshop movement" here (https://hypothes.is/a/pzSbkDSWEe-GVmsfDdhvrg) is another good example of the serendipity of autocomplete functionality in Hypothes.is helping to link together disparate examples of ideas which I'd long since forgotten. In this case to a tangential idea I'd read about a year prior (https://hypothes.is/a/bxMX5MKJEe2Wkq_zinG3iw) and been interested in, but completely forgotten about.

Now I've got a link from that to the founder of the movement in 1967.

These always read as hyperbole to me, but it's difficult to explain the surprise and serendipity of re-finding things in one's notes on a regular basis. It's akin to the sort of cognitive dissonance that religious people have when encountering the levels of complexity formed by living systems through evolution. Not having better words for describing the experience, they may resort to descriptions of magic or religion to frame their experiences.

The area of serendipity becomes much more powerful when one has ideas both directly interlinked, ideas categorized with subject headings or tags, or when one can have affordances like auto-complete.

The method Forte suggests and outlines allows for some serendipity, but not as much as other methods with additional refinements. Serendipity in Forte's method isn't as strong as in others.

In this section he's talking about some of the true "magic of note taking" which is discussed by Luhmann and others.

link to:<br /> Luhmann's writings on serendipity and surprise when using his zettelkasten (Communication with the Slipbox...)<br /> Ahrens mentions of this effect

There's some magical thinking involved here. While the system has some arbitrary internal branching, the surprises come from the system's perfect memory that the human user doesn't have. This makes it appear that the system creates its own creativity, but it is really the combinatorics of the perfect memory system with use over time.

Link to: serendipity of systems based on auto-complete

The addressing system that many digital note taking systems offer is reminiscent of Luhmann's paper system where it served a particular use. Many might ask themselves if they really need this functionality in digital contexts where text search and other affordances can be more directly useful.

Frequently missed by many, perhaps because they're befuddled by the complex branching numbering system which gets more publicity, Luhmann's paper-based system had a highly useful and simple subject heading index (see: https://niklas-luhmann-archiv.de/bestand/zettelkasten/zettel/ZK_2_SW1_001_V, for example) which can be replicated using either #tags or [[wikilinks]] within tools like Obsidian. Of course having an index doesn't preclude the incredible usefulness of directly linking one idea to potentially multiple others in some branching tree-like or network structure.

Note that one highly valuable feature of Luhmann's paper version was that the totality of cards were linked to a minimum of at least one other card by the default that they were placed into the file itself. Those putting notes into Obsidian often place them into their system as singlet, un-linked notes as a default, and this can lead to problems down the road. However this can be mitigated by utilizing topical or subject headings on individual cards which allows for searching on a heading and then cross-linking individual ideas as appropriate.

As an example, because two cards may be tagged with "archaeology" doesn't necessarily mean they're closely related as ideas. This tends to decrease in likelihood if one is an archaeologist and a large proportion of cards might contain that tag, but will simultaneously create more value over time as generic tags increase in number but the specific ideas cross link in small numbers. Similarly as one delves more deeply into archaeology, one will also come up with more granular and useful sub-tags (like Zooarcheology, Paleobotany, Archeopedology, Forensic Archeology, Archeoastronomy, Geoarcheology, etc.) as their knowledge in sub areas increases.

Concretely, one might expect that the subject heading "sociology" would be nearly useless to Luhmann as that was the overarching topic of both of his zettelkästen (I & II), whereas "Autonomie" was much more specific and useful for cross linking a smaller handful of potentially related ideas in the future.

Looking beyond Luhmann can be highly helpful in designing and using one's own system. I'd recommend taking a look at John Locke's work on indexing (1685) (https://publicdomainreview.org/collection/john-lockes-method-for-common-place-books-1685 is an interesting source, though you're obviously applying it to (digital) cards and not a notebook) or Ross Ashby's hybrid notebook/index card system which is also available online (http://www.rossashby.info/journal/index.html) as an example.

Another helpful tip some are sure to appreciate in systems that have an auto-complete function is simply starting to write a wikilink with various related subject heading words that may appear within your system. You'll then be presented with potential options of things to link to serendipitously that you may not have otherwise considered. Within a digital zettelkasten, the popularly used DYAC (Damn You Auto Complete) may turn into Bless You Auto Complete.

Nat's conclusion is correct, but his reason for arriving at that conclusion is wrong.

You're not faced with the question of where to put things with Roam because you can do the following:

(1) You can tag a new entry on the fly, in-line, CLI style. (2) If the tag exists, it will autocomplete, if it doesn't you can create it with no extra effort (3) Any tags you add are links to their respective pages, which allows you to (a) navigate their as soon as you've typed the tag/page name and (b) it creates a backlink on those pages so your new entry is automatically linked to from there.

https://www.w3.org/TR/wai-aria-practices-1.1/#combobox

Specific href: https://developer.microsoft.com/en-us/fabric/#/controls/web/combobox