How notes are classified and linked. The moat isn’t “re-read what Namjoon read”; it’s grow my own tree from where he pointed. The seed is his; the tree is mine. These rules keep that line clean.
The three note types
Seed: wherever Namjoon’s finger landed (any medium: a book, an artist, a work, a show, a place, not just people).
- Holds: Provenance (where/when he pointed, with a real primary source) + “Why it matters: the door” (my own testimony voice, the shareable gold).
- Short. A hub, not an essay.
- Status:
tending. A seed is never “done.”
Branch: my own growth, such as a book I read, a work I see, an encounter I have. Grows from a Seed or a Concept.
- Holds: notes, edition info.
- Status:
queued/reading/paused/encountered. (readingis book-specific; the in-progress state.pausedis also book-specific; started, set aside, might return.) - No fixed felt sense heading. Retroactive ingests (cataloging a visit after the fact) can’t reconstruct the in-person response, so the empty stub was dropped. When a note is written fresh off an encounter, let the felt sense live as a free paragraph in the body, not a standing section.
Concept: a recurring idea (留白, the sublime, Mono-ha-as-an-idea).
- A connective hub, not a leaf. Not part of any one tree, but the root network between them.
- Status:
tending.
The decision flow
Run any candidate through these three questions, in order.
-
Did Namjoon point at it, with a real primary source?
- Yes, and the source is the thing itself → Seed. (Then check the Provenance rules below.)
- No, I grew it off something he pointed at → go to Q2.
-
Can I consume it? (read / see / visit)
- Yes → Branch.
- No → go to Q3.
-
Is it a recurring idea with no single instance I can point to?
- Yes → Concept.
The two tests, restated:
- Seed vs. branch: did Namjoon point at this, or did I grow it off something he pointed at? Pointed-at → seed; grew-off → branch.
- Branch vs. concept: can I consume it? Read / see / visit → branch. An idea with no single instance → concept.
Provenance rules (the part that self-polices)
A Seed lives or dies on its Provenance. Be strict here; it’s what keeps this from being every other ARMY’s re-reading list.
- A Seed must cite a real primary source: Namjoon’s own Live, IG post, Weverse, interview, liner note. No real source → it’s a Branch (my own discovery), not a Seed.
- The finger must land on the thing itself. Not the room he was in (standing in a space ≠ pointing at it), not the lineage (admiring Mono-ha ≠ pointing at one artist), not a fan’s inference.
- Source quality ladder: Namjoon’s own primary source > a news report of that primary source > a fan aggregator. Aggregators (Namjoon’s Library, Goodreads lists) are signposts, not sources; follow them to the underlying primary and cite that.
- Ephemeral exception (stories & Lives). IG/Weverse stories die in 24h and Lives often aren’t archived, so the underlying primary is usually gone for good. A fan-preserved screenshot of the original post is acceptable provenance - the screenshot is the primary (Namjoon’s own content, captured); the archive is only the vehicle. Cite as: @rkive story, [date] - screenshot preserved via [archive].
- Trust the screenshot, not the caption. The thing must be legible in the shot itself (Namjoon’s handle / story UI showing, the work or book clearly identifiable). If the aggregator is naming a half-hidden cover I can’t actually make out, that’s still a fan’s inference with a photo stapled on → candidate, not a seed, until I can see it for myself. (Worked through on What Are You Looking At: it read at first like a half-hidden inference, but the Korean edition’s distinctive yellow cover is identifiable in the story - match the cover design to the edition, and it’s confirmed.)
- Don’t reverse-engineer provenance to force a seed. If I find myself wanting it to be a seed more than the evidence wants it to be, that’s the tell. Let the source decide; demote to branch without guilt.
The relationship layer (how notes link)
Three link directions, sorted by the type of note they point to.
- Grows into → points up to a Concept, the abstract idea this note feeds (留白, Mono-ha). Not consumable. (Seeds precipitate concepts: 留白 crystallized once Martin, Kapoor, and Lee Ufan all pointed at the same emptiness.)
- Branches → points down to consumables: a book, work, show, place. My own growth. Carries
queued/reading/paused/encountered. - Related seeds → lateral, to other Seeds on the same register.
Flow of the forest:
Namjoon → Seed → (grows into) → Concept
\ /
→ Branch (grows from) ←
A Branch can grow from a Seed (余白的艺术 grows from Lee Ufan, since it’s his own book) or from a Concept (陰翳禮讚 grows from 留白, not from any one seed). Use the [[links]] to carry both when both are true.
Edge cases
- Same name, two types. Mono-ha the idea = Concept (the hub linking Lee Ufan, Sekine, Suga). A Mono-ha show (Glenstone, MOMAT) = Branch, since I can walk into it. Split by the consume test.
- Person vs. work. If Namjoon pointed at an artist, the artist is the Seed; specific works/books are Branches. (Agnes Martin = seed; the Glimcher monograph = branch.)
- The room, not the subject. Namjoon in the Lee Ufan Space photographing a Boltanski show → that seeds Christian Boltanski, not Lee Ufan. The venue is not the pointer.
- …unless the room is the work. The mirror case. Namjoon in the Rothko Chapel photographing the chapel → that seeds Mark Rothko, because the room is fourteen Rothko panels as one environment, with no other artist’s show inside it. Room = subject → the venue is the pointer. (Test: is there a separate show installed in the space? Yes → seeds the show. No, the space itself is the work → seeds its maker.)
- …or when the architecture is constitutive. The third room-test. Sometimes the building is not itself the work (so the “room IS the work” exception doesn’t apply) but the architecture is still constitutive of the encounter, not interchangeable. Yun Hyong-keun’s show at Palazzo Fortuny is the worked example: Yun’s monochromes are the work, but the 15th-century Venetian palazzo around them is not a white-cube. The architecture participates. In these cases, the visit can seed both the subject (Yun, via the work) AND the venue (Palazzo Fortuny). Multi-parent seeding. Test: would the encounter be the same in a generic white-cube? If yes, venue is incidental (default rule applies). If no, if the architecture is the reason for “in awe” or the reason Namjoon chose to document it, the venue earns its own seed. Guardrail: not a license to seed every museum. The architecture must be by a named or significant maker (Ando, Scarpa, the historic palazzo itself, Phifer at Glenstone), and Namjoon’s pointer must engage with the destination’s character (building visible in his frames, venue named explicitly in his words).
- Doorway vs. convergence (optional
origin:field). Most seeds are doorways: Namjoon opened a room I hadn’t been in (Agnes Martin, Lee Ufan). A few are convergences: I was already there, and his pointing confirmed it (Mark Rothko: known pre-Namjoon, stood in Houston pre-BTS). Both are seeds; the type is set by Namjoon’s pointer, not by who arrived first. If I want to track the difference, one frontmatter line (origin: doorway/origin: convergence), and the door carries the rest. Not a new note type. - Non-bloom (optional
bloom: none). A seed Namjoon pointed at that didn’t take - I’ve met the work and felt nothing. Still a seed (his pointer sets the type, not my response); it just grows no branches, and its door testifies to why it didn’t bloom. Worth a note only when the non-bloom is illuminating - a foil that sharpens what I do love (Donald Judd against Lee Ufan). Not a license to catalog everything I dislike (that’s the completeness trap). What won’t bloom maps the soil as much as what does. Applies to books I started and abandoned too, not just artists:bloom: none+ a door explaining the divergence is the right shape. - Multi-parent branches are fine. Let a branch link to both its seed and its concept; the graph is the point.
- Seed-as-wishlist (the
queuedtag on a seed). Sometimes Namjoon pointed at a place I haven’t yet visited (Museum SAN, Dia Beacon), or a work I haven’t yet seen. The seed itself staysstatus: tending(a seed is never “done”), but thequeuedtag overlays it to surface “Namjoon pointed AND I haven’t engaged yet.” Drop the tag once engaged. If the encounter warrants its own note (multiple visits, time-bound shows, specific works that take), spin off a Branch from the seed. Status reflects type; tags reflect navigation. They can disagree by design. - Namjoon’s creative works as Seeds (the new-yield test). A creative work Namjoon made (a video, an exhibition, a performance, a book he wrote) is a Seed only when engaging with it introduces something genuinely new to my garden: a concept, an artist, a place, or an idea I didn’t have words for. If it only reinforces an existing seed, it’s provenance for that seed (cite inline) or a Branch (substantive consumable I engage with). The test: after engaging, what new node would precipitate? If something new, Seed. If only existing nodes get reinforced, Branch or provenance. This justifies BTS The Return chronos & kairos as a Seed (it precipitated 無常) while keeping Indigo Live at Dia Beacon as a Branch (Dia Beacon was already seeded via diaart.org).
Worked examples (battle-tested)
| Note | Type | Why |
|---|---|---|
| Agnes Martin | Seed | @rkive IG, Jan 2022, Glimcher monograph, clean primary source |
| Lee Ufan | Seed | @rkive IG, a Lee Ufan work dedicated “R M”, artist-signed, Namjoon tagged @ufanlee. The finger landed directly. |
| Mark Rothko | Seed (origin: convergence) | @rkive IG, Dec 2021, the Rothko Chapel, room is the work. Known pre-Namjoon; he confirmed, didn’t open. |
| 留白 | Concept | An idea, not consumable; the hub three seeds point at |
| Mono-ha | Concept | The relational idea; individual shows are branches |
| 余白的艺术 | Branch | Lee’s own book, I can read it; grows from Lee Ufan + 留白 |
| Busan / Lee Ufan Space visit | not a seed | Namjoon was in the room but pointing at Boltanski; venue ≠ pointer |
| JooniesLibrary screenshot | source if legible | Ephemeral exception: a screenshot of Namjoon’s own story is the primary. The fan’s caption/ID still isn’t - verify the thing against the image. |
| What Are You Looking At | Seed (confirmed) | Read like an inference at first - then the Korean edition’s distinctive yellow cover turned out identifiable in the story. Matched the design, not the caption: a real sighting. |
| Museum SAN | Seed + queued tag | Namjoon @rkive Aug 2019 (preserved via @choi_bts2). Destination is the pointer (Ando architecture as the work). status: tending because it’s a seed; tagged queued because I haven’t visited. |
| BTS The Return chronos & kairos | Seed | Namjoon’s own thinking on camera precipitated 無常 - a new concept. Passed the new-yield test. |
| Indigo Live at Dia Beacon | Branch | Namjoon’s own performance video filmed at Dia. Dia Beacon was already seeded; the video reinforces but doesn’t precipitate. Failed the new-yield test. Substantive consumable, so kept as a Branch. |
| Yun Hyong-keun | Seed (confirmed via ARTnews 2022) | Namjoon said his works “left me in awe” at the Palazzo Fortuny and at Chinati. Provisional state resolved by the Namjoon ARTnews 2022 interview. |
| A book I set aside | Branch, status: paused | Started, life intervened, the book is shelved but the door isn’t closed. Pick up later; switch to encountered (and write the door) when finished. |
| A book I won’t finish | Branch + bloom: none | Tried it, decided this isn’t my register. The non-bloom door explains the divergence (same shape as Donald Judd for an artist). Status stays whatever it last was. |
| Palazzo Fortuny | Seed (constitutive architecture) | The Yun show Namjoon saw there left him “in awe” (ARTnews 2022). Strict rule would seed only Yun and treat the palazzo as incidental venue. Constitutive-architecture exception applies because the 15th-century Gothic palazzo materially shapes how Yun’s monochromes land. Multi-parent seeded alongside Yun Hyong-keun. |
Two standing reminders
- The door is testimony. “Why it matters” is publishable: what moved me, what I read and saw. The raw inward inquiry (why does the void pull me) stays private in Obsidian; published, it curdles into performance.
- Ship rough. Every note is
tending. Plant a node, drop[[links]]to stubs that don’t exist yet, fill them later. The forest grows from links dropped while writing; I never sit down to “design” it. Designing the taxonomy up front is the deliberation trap; this doc exists so I don’t have to re-litigate it each time.