Context Before Knowledge?
I've been reading about Zettelkasten, PKM, note-taking workflows, and more recently the interaction between personal knowledge systems and AI.
The more I explore these topics, the more I find myself wondering whether we sometimes start one layer too high.
Most discussions focus on:
knowledge capture
note-making
linking ideas
writing
retrieval
thinking
All of that is valuable.
But what if there is something underneath all of it?
Context.
Not knowledge itself, but the network of experiences, unfinished thoughts, personal references, habits, projects, memories, and associations that make a piece of knowledge meaningful to a specific person.
A new note-taking system begins empty.
A new AI chat begins empty.
A new workspace often begins empty.
Yet people rarely arrive empty.
They usually carry years of accumulated fragments:
scattered notes
notebooks
OneNote pages
bookmarks
half-developed ideas
personal terminology
recurring questions
abandoned projects
What interests me is not how to organize those things.
What interests me is whether there is a meaningful distinction between:
managing knowledge
and
reconstructing personal context.
In other words:
Is knowledge management primarily about storing and linking information, or is part of the challenge helping a person rediscover the context that already exists around that information?
I am not proposing a product.
I am not promoting a method.
I am mainly interested in how people here think about the relationship between context, knowledge, and meaning.
For those who have worked with a Zettelkasten for years:
Do you see a difference between knowledge and context?
Does your Zettelkasten help you remember information, or does it help you reconstruct a way of thinking?
Have AI tools changed how you think about this distinction?
I would be very interested in your perspectives.
Kind regards,
Willy Wurzel 🥕
Howdy, Stranger!
Categories
- 3K All Categories
- 153 Research & Reading
- 702 The Zettelkasten Method
- 12 Knowledge Work
- 101 Writing
- 473 Software & Gadgets
- 157 Workflows
- 733 The Archive
- 15 Plug-In Showcase
- 88 Resolved Issues
- 225 Projects Logs and Journals
- 86 Project: Zettelkasten.de
- 54 Critique my Zettel
- 178 Random
- 375 Introduce Yourselves!
Comments
I see Zettelkasten as a way to augment cognition. You have your existing set of configurations of neurons in your head already when you start, organized in a semantic network. It's up to you whether you want to externalize them or attach new knowledge to them. Either way, Zettelkasten enriches the network of connections. There's a reason why some people prefer the term Second Brain even though it doesn't encapsulate what the method really does.
I think seeing Zettelkasten only as a learning tool restricts its capabilities. You can start with whatever you want. I am trying to construct my Zettelkasten out of core beliefs and principles. I have a worldview, for example I believe that we are born with an understanding of intuitive physics. This contradicts with the blank slate opinion and leads me to a different political worldview. What I mean is knowledge builds on top of itself. And we have context at the bottom. I don't see it something that shouldn't be in my Zettelkasten. I wouldn't mix atomic notes, though with different kinds of knowledge, episodic or semantic.
I am reading "Meno & Other Dialogues" right now. When the humanity didn't have the accumulated Wikipedia, it was all context.
Selen
Small correction:
I accidentally posted and deleted an earlier draft of my reply.
Nothing profound. Just me tripping over my own workflow.
The current version is the intended one.
🥕
Thank you,
Your distinction between context and accumulated knowledge is very close to what led me to ask the question in the first place.
What makes this interesting to me is that I am not approaching it only as a theoretical question.
Over the last months I have accumulated roughly 400 MB of logs, conversations, notes, project fragments, and unfinished thoughts. Much of it was never intended to become a Zettelkasten, yet it clearly contains context: recurring questions, assumptions, interests, forgotten projects, and connections that I no longer remember consciously.
I am currently experimenting with a database and an AI-assisted retrieval layer on top of that material. It is far from finished. It creaks, it breaks, and nothing is proven.
But I am already using it.
In fact, I used it while writing this reply.
To be transparent about my process:
I retrieve potentially relevant material from my own notes and logs.
I read the retrieved material myself.
I compare it with the current discussion.
I write, rewrite, and often remove more than I keep.
I review and correct everything before posting.
I do not post AI-generated text without reading and editing it first. The responsibility for what gets posted remains mine.
One thing I keep noticing is that there is a difference between:
information I know;
information I once knew;
context that shaped how I think.
The last category is the most difficult to retrieve.
I find your distinction between episodic and semantic knowledge particularly interesting because much of what I am retrieving seems to sit somewhere between the two. Not just facts, and not just memories, but traces of how certain ideas were formed over time.
My current working assumption is:
Empty does not necessarily mean empty. Sometimes it simply means: not looked at yet.
That may be true for old notes, abandoned projects, and even parts of our own thinking.
Whether a Zettelkasten can serve as a bridge back to that context is one of the things I am trying to understand.
Thank you for the thoughtful reply.
🥕
I think that context is very important, especially when one's thinking is incomplete. IMO, a note works better when it helps one to restore the context of one' thinking when it was written. After all, a note can only be flattened representation of a complex thought process. When an idea has been refined down to a clean, separable core, if that ever happens, then the context may be less important. Until then, it is.
Writing notes that help restore context is part of an art I sometimes call "cardcraft", sometimes as "caching a scene". It may even take several cards working together to restore the context.
Here is part of one of my own notes about cardcraft:
Cardcraft is a learned skill:
- Craft cards according to their purpose.
- Cards may have different purposes.
Subject: a card's "aboutness". The title should help with:
- search
- recognition
- refreshing the scene
Scene cache:
- refresh, re-light the scene, rejoin the groove. Yay!
I care a lot about context. When I find a claim that triggers my curiosity, my first impulse is to research and/or reconstruct context. I like to look up terms. I like to verify quotes. I like to check references.
My Zettelkasten doesn't collect decontextualized knowledge atoms. I use it to make context explicit. That's why I care about people, sources, timelines and questions in my notes.
I think that knowledge needs context in order to be meaningful.
Both.
Yes. LLMs make it easier to reconstruct context. Their training data contains much more context than any human could ever process. The challenge is to leverage their training data.
In this recent example you can see how research results went into refining a prompt. When you scroll back in that thread, you see how the prompt evolved. The prompt itself now contains a lot of knowledge. It references key texts, summarizes key questions and lists relevant names. And this is just the current iteration. Developing such prompts is a lot of work. The benefit is that with each iteration the results become more useful. The clearer the context becomes, the clearer the knowledge becomes. Working with LLMs helped me discover names and technical terms that I wouldn't have found otherwise. Those findings provide context for the next iteration. Those findings help me ask better questions.
LLMs are useful for finding context. However, identifying meaningful knowledge remains a human task. I don't consider LLM generated material as knowledge, but as hints for potential knowledge. Surprisingly, Zettelkasten becomes more important than ever. It helps ask more focused questions and document the learnings.
One thing that stands out to me from the replies is that nobody seems to begin with knowledge.
People begin with curiosities.
Questions.
Fragments.
Observations.
Interests that refuse to go away.
Knowledge appears later.
Perhaps context is not beneath knowledge or above it.
Perhaps context is the path we walk while knowledge is still forming.
Play.
Stumble.
Log.
Follow the interest.
🥕
I'm not sure if I understood the issue correctly, but in my case, part of what you called "context", when it was useful, was assimilated into my Zettelkasten, after some kind of processing.
I didn't start with the idea of having to include all of it anyway; that's just unfeasible in my case.
An old note taken in a file of some kind, for example, once retrieved and reread, was treated as a source and processed to produce zettels. Sometimes fully reformulated, other times simply split and pasted it.
So, in my experience, I wouldn't necessarily describe this as "reconstructing personal context". It is more like selectively processing and assimilating parts of my previous context into an evolving knowledge network.
But if another person finds it useful to recover a much larger part of their previous context, including old experiences, unfinished thoughts, terminology, projects, and so on: why not? It may be exactly what they need.
I think the important thing is ultimately that the process leads to something useful, rather than doing it simply for the sake of doing it (or avoiding it simply because some tutorial, theory or methodology doesn't prescribe it, because they conventionally assume that we are starting from an empty slate).
For example, in my Zettelkasten section about running, I didn't try to reconstruct all of my previous "context." I had journal notes and data from my runs, observations, questions, discussions with other runners, articles, and ideas I had accumulated over time.
Whenever something seemed relevant, I processed it and turned it into notes, connecting it with the other knowledge I was developing.
Through this process, for example, I gradually developed my own understanding of the relationship between heart rate, intensity, and the ability to sustain an effort for a long time. It didn't start from already structured knowledge; it emerged progressively from the combination of personal data, experience, reflection, discussions with other people, and external sources. Integrating today's material, but also retrieving and re-analyzing some of of the past.
So the point wasn't to "reconstruct my whole context around running," but rather to select and assimilate from that context whatever could contribute to building knowledge that was useful to me as a runner.
And in my case, much of what I eventually integrated into the Zettelkasten didn't actually come from books or articles. A significant part came from analysing my own runs: looking at what happened during a workout, comparing different sessions, noticing patterns, questioning my interpretations, to revisit old unresolved doubts, and connecting those observations with what I already knew. What I could call "my running context", indeed.