Roam is a new medium, so it's not about adapting zettels (card-based) to Roam, but developing a new native method within Roam (block-based).

"The medium is the message." - McLuhan

Zettels exist traditionally as indexed cards (1a2be) which show horizontal and vertical expansion. Roam allows blocks to float non-spatially on just a block ID, allowing multiple dimensional access.

Zettel means "note" and kasten means "box". Its main purpose isn't entirely to store notes or process them, it is to provide a conversation of ideas with serendipity and surprise. Its power is in accommodating conflicting ideas and generating a path of interconnected ideas.

Roam has the same power (or more) but does it frictionlessly. It goes beyond that, because even if we do not consciously see a link, Unlinked References open up a treasure trove of previously unseen connections.

The issue I have about using a system (zettelkasten) overlaying a system (roam) that is already built to do that, is that it brings a feeling of redundancy. This feeling of unease makes me think there is a way to simplify all of this.

Isn't every note we put into the daily notes page a "fleeting note"? To not make it get lost, we [[link]] or #tag it - this is a form of making it "not fleeting", of getting it to resurface. But this also makes me rethink... should we only link/tag things a few levels deep?

My current behaviour is fearing the loss of my notes on the daily page. Hence, I frantically tag things in an attempt to make sure I can locate them in future. But actually, maybe this is unwarranted - because we can certainly access them through Unlinked References in future.

Literature notes are "lightweight synthesis of observations collected while reading" (Andy Matuschak). Traditionally they come with citations pointing to the original source. The question is, are they really necessary?

If we have a system that automatically shows us all notes related to a source, why should we hold onto this limitation found in physical zettelkasten? Would it be insufficient to just have metadata?

The only benefit I can think of is that it is paraphrased in our own words. Not just that - it leverages on [[smaller ideas]]. These "light", low-fidelity notes introduces ideas within which can show connections between other notes.

So in a way, it's no longer just about just getting to the "big idea" (which forms a Permanent Note) but along the way picking up many "smaller ideas" that can be shared among similar topics.

It's a spectrum of fidelity - as it increases, it becomes a "permanent note" or a prominent insight. This is the distilled, atomic representation of the "literature notes" we picked up earlier. Because of its atomicity, it can be easily used to interact with other prominent ideas

The purpose really isn't just to retain memory of old ideas, but to leapfrog from prior ideas (insight leap). We need a system that allows our subconscious to brew these atomic ideas. It's not just about output, but to transcend mere synthesis.

Which also brings us to the importance of spaced repetition - imo it's not about remembering things. It's the revisiting (programmed attention) that allows insight to emerge as we wrestle with ideas over time.

https://twitter.com/syncretizm/status/1343818539220680706?s=19 on Google mentality in Roam