Zettelkasten Forum


Planning for serendipity

I am currently reading "Serendipity - It doesn't happen by accident" by David Cleevely
(https://www.conductingserendipity.com).
D. Cleevely was and is a driving force behind Cambridge's innovation power. He claims that while we cannot plan serendipity, we certainly can set up structures and habits to make it very likely.
In this forum, there has already been quite some discussion around about this: a code that calls up random cards from the Zettelkasten for example.
Just like Richard Feynman and his 12 favorite problems,
https://fortelabs.com/blog/12-favorite-problems-how-to-spark-genius-with-the-power-of-open-questions/
one could have a set of topics needing input. Each new zettel is checked if it clicks with one of these topics.
In times of AI, one could experiment with asking one's favorite AI to find surprising combinations of zettels or the like.
What other strategies to foster serendipity have already been discussed in this vast forum, or are brand-new?

How to zettel for stellar revelations? What a fantastic journey!

Comments

  • edited October 2

    Oh, this is a really broad topic for me, and over time I have identified many factors that enable serendipity.

    Overall, I think the entire Zettelkasten practive acts as a major laboratory that fosters serendipity, as it already includes or can easily incorporate many of these factors:

    • Atomicity, through which many small, recombineable units are created.

    • The inclination to connect, and therefore bring together and compare different ideas.

    • The inclination not to just copy content as it is, but to transform it by internalizing it personally (simply put, I often make interesting "discoveries" just by generalizing or abstracting a broader, more universal principle from a specific instance).

    • The Zettelkasten's ability to thrive in disorder (it's essential to read Luhmann's paper on this).

    • The possibility (or even tendency) within the Zettelkasten to use emergent constructs to guarantee order and structure, rather than relying on planned, rigid structures that create silos.

    • The opportunity within the Zettelkasten to develop not just one, but all of our viewpoints/contexts. In my practice, for example, I interleave thought of my work, studies, passions, hobbies, and everyday life, and ofter they connect together or they find a new life in a different context

    • The system's capacity to "surprise" us (again, reading Luhmann's paper is crucial here) thanks to its internal dynamics.

    • The "forgetting machine" effect (I can't recall who coined this).

    • The system's predisposition to be browsed and re-explored (via structure notes, Folgezettel, or other forms of train of thought representations built over time in an emergent way) rather than depending solely on targeted searches.

    In general, I think serendipity develops best when our approach to the Zettelkasten is largely free from ahead predesigned structures, favoring bottom-up growth. However, over time I've also noticed a few dynamics that work in the opposite direction.

  • @Marcel_Liauw said:
    I am currently reading "Serendipity - It doesn't happen by accident" by David Cleevely
    (https://www.conductingserendipity.com).
    D. Cleevely was and is a driving force behind Cambridge's innovation power. He claims that while we cannot plan serendipity, we certainly can set up structures and habits to make it very likely.
    In this forum, there has already been quite some discussion around about this: a code that calls up random cards from the Zettelkasten for example.
    Just like Richard Feynman and his 12 favorite problems,
    https://fortelabs.com/blog/12-favorite-problems-how-to-spark-genius-with-the-power-of-open-questions/
    one could have a set of topics needing input. Each new zettel is checked if it clicks with one of these topics.
    In times of AI, one could experiment with asking one's favorite AI to find surprising combinations of zettels or the like.

    I have a vision of a LLM that can find semantic similarities to what I am working on among other notes. The LLM would have digested the current state of my Zettelkasten, of course. In my view this system couldn't be a chatbot as we know them because they usually adapt too easily but I would need it to stay unadapted so as to maximize surprise and the unexpected.

    What other strategies to foster serendipity have already been discussed in this vast forum, or are brand-new?

    Being receptive: e.g., noticing potential word play, which can draw one into a productive detour; associations that arise without planning.

    I ran a search for serendip and found this note:

    The Zettelkasten Triad (Serendipity, Synergy, Syzygy)
    ------------------------------------------------------
    
    :id: tom.20260317232754.1
    :created: 2026-03-17 23:28
    :seealso: tom.20260317173731.1 How Zettelkasten Work
    :seealso: tom.20260311232848.1 Zettelkasten and the Five Levels Of Thinking Support
    :seealso: tom.20260317233456.1 Zettelkasten And Surprise
    :seealso: tom.20260318003715.1 Zettelkasten As A Game Of Trivial Pursuit
    
    The holy triad of Zettelkasten creative support:
    
    Serendipity - Find the unexpected.
    Synergy - Feel the reinforcement.
    Syzygy - Ride the resonance.
    

    Notice the next to last :seealso: line. Navigating to that note gives me this:

    Zettelkasten And Surprise
    --------------------------
    
    :id: tom.20260317233456.1
    :created: 2026-03-17 23:35
    :seealso: tom.20260317232754.1 The Zettelkasten Triad
    
    As Shannon wrote and Luhmann knew, a predictable message communicates nothing.
    
    If every search were so precise that it only returned what you expected, then you could not gain anything new. 
    
    So you don't want to make searches too precise when you hope to learn something.
    

    This note feeds right into my next suggestion:

    Making searches on more general terms than you are looking for can lead you to stumble on forgotten or unexpected notes and semantic regions. Using a word stem instead of a complete word can do the thing. An example is searching for cogni instead of cognitive or cognition (I never remember which of these two forms I have used so I search for both using the stem). Here's part of the results I get for this search - if you notice a few duplicates under different headings, they are clones not mistakes:

    Cognition /Plants
        Plant Cognition - Intelligent Behavior? 
        Plant Cognition - Anticipation 
    
    Cognition /Vannebar Bush And The Memex 
        Memex Cognitive Trails 
    
    Ideas /Mine /Emotional vs Cognitive Memories /Emotionally-driven Memories 
        Infant Pairs Emotional Trace With Cognitive One 
        Infant''s Cognitive Memories May Have Mistakes 
        Emotional Associations Probably Influence Cognitive Storage 
    
    Ideas /Mine /Emotional vs Cognitive Memories /Emotionally-driven Memories /Emotional Associations Probably Influence Cognitive Storage 
        Need Cognitive Development To Store Reliable Memories 
    
        Ideas /Mine /Emotional vs Cognitive Memories /Emotionally-driven Memories /Emotional Vs Episodic Recall Varies Across People 
        Can Person Intentionally Alter Balance Of Emotional vs Cognitive Recall? 
    
    Ideas /Mine /Human Language /Language Origins 
        For a vocalization to express a cognitive state ... 
    
    Knowledge /Concepts /Standard Definitions of Concept 
        Wikipedia on Concepts of Cognitive Science 
        Concept (Open Encyclopedia Of Cognitive Science) 
    
    Mind Mapping 
        Mind Maps As Cognitive Scenes 
    
    Playground 
        Explanation as a Compound Cognitive Process 
    
    Playground /Local Neighborhood - Thought Trails 
        Memex Cognitive Trails 
    
    Playground /Structure Card Experiments /Clone Structure Outline for Purpose of a Zettelkasten, What it is 
        Me: Cognitive Instrument 
    
    Zettelkasten /Purpose of a Zettelkasten, What it is 
        Me: Cognitive Instrument
    

    Depending on one's mood and interest at the moment of search, one might get intrigued by (or remember) one or another of these hits.

  • edited 12:36AM

    Thank you both for fast and dense feedback. When reading it, I realized that I started filling my Zettelkasten one year ago, at age 61, and do not even have as many as 400 zettels. From a "keep it simple"-perspective (@Edmund!), this is fine. However, I would need to transfer decades of material on my hard disk and beyond into my Zettelkasten to tap the potential of all that I created and digested in the past. There is plenty of discussion about the critical mass of zettels for the Zettelkasten magic.
    Using the word stems in searching is certainly an excellent method not to miss too much.

    How to zettel for stellar revelations? What a fantastic journey!

  • If you are interested in serendipity, one concept that might be interesting for you is adjacent possible, which is defined as

    a limit on change where systems can only grow or invent things that are right next to their current state

    It explains innovators and thinkers who are beyond their Zeitgeist and why some work becomes valuable after a long time.

    You cannot plan serendipity as the precise point in time, but plan for its emergence by extending and collecting notes in certain direction.

    I first came across the concept of adjacent possible in the book "Why Greatness Cannot be Planned". One other reference might be "Where Good Ideas Come From". It embraces a networked understanding of thought which is very similar to the Zettelkasten Method.

    One other reference came to my mind is Csikszentmihalyi's book on creativity. He surveys original thinkers and lays out his findings and one conclusion is the need for domain knowledge for creativity to occur.

    My reply became a "see also" list, but these were my two cents. :smile:

    Selen

Sign In or Register to comment.