How to Build a Research Vault

A research vault is a folder of plain-text files holding every paper you read, every concept you build, every method you compare, and every chapter you draft — across the four to ten years between your first paper and your dissertation defense. The citation manager keeps the metadata. The vault keeps the thinking. Every artefact stays yours.

NSF's Survey of Earned Doctorates recorded 58,131 research doctorates awarded by US universities in 2024, with a 91.5% response rate to the official census.^{1 2} The Council of Graduate Schools' multi-cohort PhD Completion Project, drawing on roughly 49,000 students at 30 institutions across 54 disciplines, set the ten-year completion rate at 56.6%.^{3 4} The American Academy of Arts and Sciences placed median humanities completion at 42% — and median time-to-degree at "approximately 10" years for those who finish.⁵ None of those numbers describe a sprint. They describe a decade-long project that needs an instrument designed for the duration.

Why most PhD research vaults collapse

Most PhD research vaults collapse because the student treats the vault as storage and writes the dissertation from memory. Papers pile up in Zotero. Highlights freeze inside PDFs. The chapter draft begins in year four with a folder no one can interpret. The cure: write the source-note, concept node, and methodology card the week the paper enters the vault.

Sönke Ahrens names the bottleneck plainly in How to Take Smart Notes (2017): the issue is not reading speed but the missing writing step between reading and the eventual chapter, and literature notes are "selection and filtering," not collection.⁶ Niklas Luhmann produced "more than 70 books and nearly 400 scholarly articles" — the higher count of "roughly 50 books and 550 articles" is also reported — using a Zettelkasten of approximately 90,000 cards built between 1952 and 1997, described in his essay Kommunikation mit Zettelkästen around 1980–81.^{7 8} Andy Matuschak distils the lineage for the digital era: "Evergreen notes are written and organized to evolve, contribute, and accumulate over time, across projects."⁹ The vault that collapses is the one that never made the writing step a weekly habit.

Pat Thomson, emerita professor of education at the University of Nottingham who has supervised "well over 60 PhDs," argues that the long-term reading discipline is itself the chapter: "Keeping up with what's being published takes time and effort... A partial counter to not keeping up... is a reading routine. A reading routine that works for you," she writes on her Patter blog.^{10 11} The vault is the place that routine lives.

The vault structure that survives a five-year project

The vault is one project directory with eight folders, each holding one kind of artefact. Sources, concepts, methods, synthesis, chapters, fieldwork, supervisor meetings, templates. The structure scales from year one (a dozen files) to year five (a few thousand) without restructuring. Every artefact has a folder. Every folder has a template. Nothing important sits in the inbox forever.

phd-vault/
├── 00_inbox/                # 24-hour holding pen; processed weekly
├── 10_sources/              # one .md per paper, book, talk, dataset
│   ├── ahrens-2017-smart-notes.md
│   ├── luhmann-1992-kommunikation.md
│   └── ...
├── 20_concepts/             # one .md per atomic concept
│   ├── atomicity.md
│   ├── synthesis-vs-capture.md
│   ├── temporal-organization.md
│   └── ...
├── 30_methods/              # one .md per methodological precedent
│   ├── ethnography-12-students-smith-2019.md
│   ├── prisma-2020-page.md
│   └── ...
├── 40_synthesis/            # the chapter spine — one file per matrix
│   ├── synthesis-matrix.md
│   ├── concept-map.md
│   └── gaps-and-questions.md
├── 50_chapters/             # the dissertation itself
│   ├── 01-introduction.md
│   ├── 02-literature-review.md
│   ├── 03-methodology.md
│   ├── 04-findings.md
│   └── ...
├── 60_fieldwork/            # data, interviews, observation logs
│   ├── interview-001-2026-04-12.md
│   └── ...
├── 70_meetings/             # supervisor + committee notes
│   └── 2026-05-16-supervisor.md
└── 90_templates/            # the four canonical templates below
    ├── source.md
    ├── concept.md
    ├── methodology-card.md
    └── matrix-row.md

The numeric prefixes keep the folder order stable across editors — a small discipline that pays compounding returns when the supervisor visits the vault. This eight-folder topology aligns with the consensus structure in the academic Obsidian community, where practitioners such as Elizabeth Case, who runs a multi-year PhD vault, frame the principle bluntly: "if I can't find information I've taken notes on, it makes taking future notes even harder: putting in that time feels futile."¹² The folder is not Obsidian-specific — the same files render identically in MNMNOTE, Logseq, VS Code, or a plain ls listing.^{13 14}

The source-note: one paper, one file

The source-note is one markdown file per paper, written within seven days of reading. Filename equals the BibTeX citation key. Body is fixed — frontmatter, citation, summary, methodological frame, claims with page numbers, three verbatim quotes, the gap, links to related notes, your response. Twenty to thirty minutes per paper. Written once, mined every chapter for the next five years.

---
citekey: ahrens-2017-smart-notes
authors: [Ahrens, Sönke]
year: 2017
type: book
field: knowledge-work
read-on: 2026-05-12
status: read | partial | skim | abandoned
tags: [zettelkasten, knowledge-work, smart-notes]
---

# Ahrens (2017) — How to Take Smart Notes

**Citation:** Ahrens, S. (2017). *How to Take Smart Notes: One Simple
Technique to Boost Writing, Learning and Thinking — for Students,
Academics and Nonfiction Book Writers.* CreateSpace.

## One-sentence summary
Adapts Luhmann's Zettelkasten into a three-tier note system (fleeting,
literature, permanent) and locates the bottleneck of academic work at
the missing writing step between reading and the eventual paper.

## Methodological frame
Practitioner-philosophical synthesis. Not empirical. Reasons from
Luhmann's biographical productivity (~90,000 cards) and German
educational philosophy.

## Claims (with page numbers)
- Literature notes are "selection and filtering," not collection.
- Permanent notes are written in your own words. (p. 87)
- Linking notes is the value-creating step, not capturing them.
  (p. 96)

## Verbatim quotes
- "Writing is not what follows research, learning or studying, it is
  the medium of all this work." (p. 5)
- "We have to do the work that turns information into knowledge."
  (p. 88)

## Gap this opens for me
Ahrens does not operationalise corpus velocity — what to do when the
field publishes 2,000 papers per year against a candidate who reads
three a week. My field is one of those. (Forward to
[[20_concepts/corpus-velocity]].)

## Links
- [[10_sources/luhmann-1992-kommunikation]] — origin of the system
- [[20_concepts/synthesis-vs-capture]] — central distinction
- [[20_concepts/atomicity]] — Matuschak's reframing

## My response (drafted, not for chapter)
Holds for slow-moving fields. In mine the lit note has to be lossier:
three quotes, not five. The "permanent note" tier collapses into the
synthesis matrix in practice.

The template is deliberately fixed. The same nine sections, in the same order, every time. Inger Mewburn — Professor and Director of Research Training at the Australian National University, who launched the Thesis Whisperer blog in 2010 and reports approximately 68,600 email subscribers and 12.8 million site visits — argues for fixed structures from year one: "the time to start organising and writing your academic story is right now, while you are a PhD student."^{15 16} By the thirtieth source-note, the template is the only thing that scales. By the three-hundredth, it is what lets the chapter draft itself.

Ahrens is explicit on what does not belong in this file: copy-paste passages and verbatim sentences taken without thought. "Be selective with quotes," he writes, "and don't copy verbatim, as rewriting in your own words is key to understanding."⁶ Three quotes is enough to anchor every later paraphrase to the source. Five is procrastination dressed as diligence.

The concept node: one idea, one file

The concept node is one markdown file per atomic idea. The file holds the canonical definition, the paper that introduced it, the one or two that refined it, and a paragraph in your own words on how the concept appears in your dissertation's argument. Twenty to forty load-bearing concepts carry a typical thesis. They each fit.

---
concept: atomicity
parent: knowledge-work
status: stable | drafting | unsure | retired
last-touched: 2026-05-15
---

# Atomicity

## Canonical definition
"It's best to create notes which are only about one thing — but which,
as much as possible, capture the entirety of that thing."
— Andy Matuschak, *Evergreen notes should be atomic*[[10_sources/matuschak-evergreen-notes]]

## Origin
Matuschak's web-published evergreen-notes system, derived in turn from
Luhmann's Zettelkasten and the software-engineering principle of
separation of concerns.

## In my use
The property that lets a research vault scale from year one to year
five without restructuring. One concept-node per concept, one
source-note per paper, one matrix-row per claim. Each artefact links
out; none subsumes another.

## Adjacent concepts
- [[20_concepts/synthesis-vs-capture]] — what atomicity makes possible
- [[20_concepts/concept-orientation]] — Matuschak's companion principle
- [[20_concepts/portability]] — the file-level consequence

## Where this concept appears in the chapter draft
- §2.2 (the structural property of a long-term vault)
- §3.1 (the methodological argument for the matrix)

Matuschak's atomicity principle is the discipline that lets the vault survive the dissertation. "If your notes are too broad, you might not notice when you encounter some new idea about one of the notions contained within, and links to that note will be muddied; if your notes are too fragmented, you'll also fragment your link network," he writes.^{17 18} One file per concept is the working balance. The "in my use" paragraph is the most under-rated field: it forces the candidate to translate from someone else's idea into their own argument every time the concept is touched. Concept-drift dies on the day the paragraph is written.

The methodology card: one method, one decision

The methodology card is one file per methodological precedent — every paper that informs how you chose your own method. Each card records method, sample, time horizon, analytic frame, contribution, and the disqualifier ("why not this method for my thesis"). The methods chapter drafts itself in year three from these cards. The viva asks the question they answer.

---
citekey: smith-2019-ethnography
method: participant-ethnography
sample: 12 PhD students, 1 dept
time-horizon: 18 months
analytic-frame: grounded theory
status: precedent | adopted | rejected
last-reviewed: 2026-05-15
---

# Smith (2019) — Ethnography of doctoral practice

## Why this is a precedent for my thesis
Smith ethnographically documents the same population (PhD students)
my thesis studies. The thick description supplies a vocabulary for
practice that my retrospective document analysis cannot generate.

## What the method gave the field
A vocabulary for the quotidian (highlight, scratch-pad, draft chapter)
that quantitative surveys had missed. The "three failure modes"
typology is now cited across the field.

## Why not this method for my thesis
1. Single-department sample is too narrow for my research question
   on field-level reading routines.
2. Eighteen-month observation window misses the multi-year arc that
   my dissertation argues is the load-bearing unit.
3. IRB cost, in time, is incompatible with my submission timeline.

## What I am borrowing
- The "three failure modes" framing as a thematic spine for §4.
- The fortnightly memo cadence as a fieldwork discipline.

## Where this card appears in the dissertation
- §3.1 (precedent paragraph)
- §3.2 (rejection paragraph)
- §3.3 (borrowed-element paragraph)

Patrick Dunleavy, emeritus professor of political science at the LSE who taught the Drafting and Writing a PhD course at LSE for twelve years and supervised "more than 30 doctoral students over two decades," is unambiguous in Authoring a PhD (2003): the methods chapter must answer the negative case, and the candidate must rehearse the answer at every supervisory meeting.^{19 20} The methodology card is where that rehearsal lives. The "Why not this method for my thesis" section is the field every committee member reaches for, and the field that gets the candidate through the viva.

The synthesis matrix: claims and gaps in one view

The synthesis matrix is one file with one row per source and one column per theme, claim, or methodological choice. Reading a row tells you what a paper said. Reading a column tells you which papers cluster on a theme — and the column becomes the literature-review chapter. A 200-source review drafts from the matrix in eight weeks, not nine months.

# Synthesis matrix — [working title of literature review]

Last updated: 2026-05-16
Themes: T1 = capture-vs-synthesis | T2 = atomicity | T3 = portability |
T4 = corpus-velocity | T5 = collaboration

| Source | T1 | T2 | T3 | T4 | T5 | Verdict |
|--------|----|----|----|----|----|---------|
| [[10_sources/ahrens-2017-smart-notes]] | "writing is the medium" (p.5) | atomic = permanent | filesystem | not addressed | absent | foundational; thin on velocity |
| [[10_sources/luhmann-1992-kommunikation]] | slip-box as second mind | one card = one idea | paper, not portable | ~90k cards / 30 yrs | absent | origin; pre-digital |
| [[10_sources/matuschak-evergreen-notes]] | concept-orientation | atomic + concept | filesystem + web | not addressed | "personal perspective only" | digital reframing |
| [[10_sources/page-2021-prisma-statement]] | reporting-only frame | n/a | flowchart | not addressed | inter-rater | systematic-review scope |
| [[10_sources/thomson-2023-keeping-up]] | reading routine over speed | n/a | n/a | "often and little" | n/a | rhetorical posture |

## Vertical reads (the chapter writes itself from these)

**T1 — capture-vs-synthesis (§2.1):** Ahrens, Matuschak, Luhmann
locate the bottleneck at synthesis, not capture. PRISMA presumes
synthesis downstream of inclusion. Thomson reframes as a routine
problem.

**T3 — portability (§2.3):** Ahrens (filesystem), Matuschak
(filesystem + web), Luhmann (paper). The field has not addressed
cross-tool portability as a research question. Possible
contribution.

**T4 — corpus-velocity (§2.4 — gap chapter):** Nobody addresses corpus
velocity. Every author writes as if the literature is stable. My
field's literature is not. This is the dissertation's contribution.

The synthesis matrix has been the standard tool in academic writing centres for two decades. The UNC Writing Center frames the principle directly: a literature review must be "organized around ideas, not the sources themselves," and writers must "identify connecting themes across sources and present information through those conceptual threads rather than source-by-source summaries."²¹ The PRISMA 2020 statement, "Preferred Reporting Items for Systematic reviews and Meta-Analyses," published by Page and colleagues in BMJ (2021;372:n71), formalises the same discipline at the systematic-review end: every included paper coded against an explicit set of criteria.^{22 23} Even a narrative review borrows PRISMA-style transparency in private.

The verdict column is the one that pays. One sentence per source captures what it is for in your argument: foundational, derivative, methodologically interesting, an outlier, a polemic. Tagging early — the moment the row enters the matrix — saves chapter-writing months later, when "what was this paper actually doing here?" takes ninety minutes to answer instead of fifteen seconds.

The five-year lifecycle of a research vault

A research vault is a five-year project that looks different each year. Year one is the inbox. Year two, the corpus. Year three, the synthesis. Year four, the chapter draft. Year five, the defense. The topology stays constant; the centre of gravity moves through the eight folders in a predictable arc you can plan against.

Stanford's TAPS programme requires a five-to-eight page dissertation prospectus minimally covering "the research question and context, the methodology for research, and a complete chapter by chapter plan," approved "in or before the autumn quarter of the fourth year."²⁴ Cornell's Graduate School requires the dissertation be submitted "within 60 days of the final exam," with three fixed annual conferral deadlines.²⁵ The five-year arc is not a flourish — it is the institutional schedule the vault has to absorb.

Helen Sword, former director of the Centre for Learning and Research in Higher Education at the University of Auckland, surveyed 1,223 academic writers and interviewed 100 successful authors for Air & Light & Time & Space (Harvard University Press, 2017). Her finding refutes the cliché that successful academics write every day: "behavioral habits govern how you carve out time and space for your writing. Successful academics don't necessarily write every day, but they're constantly strategizing about how and when they'll get their writing done."^{26 27} Pat Thomson reaches the same conclusion: "many successful writers emphasise consistency over daily frequency — having regular writing periods rather than forcing daily output... it's the advice that needs fixing, not you."²⁸ The vault is the structure that makes the rhythm possible.

Eviatar Zerubavel, Distinguished Professor Emeritus at Rutgers University, frames the five-year arc as a problem of temporal organization in The Clockwork Muse (Harvard University Press, 1999). The book argues that "the dreaded 'writer's block' often turns out to be simply a need for a better grasp of the temporal organization of work."²⁹ The five-year vault makes that organization visible. Files in year-one folders should not still be the centre of gravity in year four; if they are, the candidate is reading too much and writing too little.

The weekly rhythm that makes the vault work

The week runs on a single rhythm: read three to five papers, write a source-note for each within seven days, update the synthesis matrix the same week, and reserve Friday afternoon for the concept map, methodology cards, and a chapter delta. Twelve hours a week, sustainable for five years. The dissertation writes itself in the background.

Five hundred words a week for two hundred weeks is 100,000 words — a full thesis, written in increments small enough to fit between teaching, supervision, and life. Pat Thomson is explicit on the principle: "what matters is how often you do it. And how little time you can put into doing something useful. That's the key, often and little."¹⁰ The Friday-afternoon write is the discipline most PhDs skip. It is also the discipline that turns the dissertation from a year-five sprint into a four-year drip.

How citation managers, Obsidian, and Notion fit

Citation managers store metadata and PDFs. Markdown vaults store thinking. Web outliners hold project tasks. The three roles are distinct, and the friction in a research vault almost always comes from collapsing two of them into one tool. Two-system or three-system architectures consistently outperform monolithic ones in published academic-Obsidian writeups across the past five years.^{30 31}

Zotero is the strongest free citation manager in 2026: open-source, originally created at the Roy Rosenzweig Center for History and New Media at George Mason University in October 2006, now maintained by the non-profit Corporation for Digital Scholarship.³² Zotero 7.0 (August 2024) shipped a major redesign with an integrated PDF and EPUB reader, dark mode, and native Apple Silicon support; current stable is 9.0.2.³² Emiliano Heyns's Better BibTeX plugin generates "stable citation keys, without key clashes" — the citation key becomes both the BibTeX reference and the source-note filename, so every lit note can be cited from the chapter draft in one keystroke.³³ Mendeley (founded London 2007; acquired by Elsevier in early 2013 for €50 million; Desktop discontinuation reversed July 2025) remains usable, but the plugin ecosystem decides for a markdown vault.³⁴

Obsidian, Logseq, MNMNOTE, and VS Code all read the same .md vault. Inger Mewburn frames the editor's role plainly: "Obsidian is a database program, but that's a bit like saying a library is merely a book storage facility... Obsidian is an elaborate knowledge management system for writers."³⁵ Notion and Roam render the same files only partially — both use proprietary formats for the structures that matter most (Notion's database blocks; Roam's bullet hierarchy) — and that asymmetry is the structural argument for plain text. Discovery tools (Connected Papers, Research Rabbit, Litmaps, Semantic Scholar) sit upstream of the vault, finding the papers the citation manager missed; Crossref's registry exceeds 183 million scholarly works as of mid-2026, so triage is permanent.³⁶

Pitfalls common to dissertation vaults

Five failure modes account for almost every research-vault collapse: highlighting without note-taking, synthesis-by-procrastination, concept-drift across chapters, citation-manager bloat, and tool-switching mid-degree. Each is preventable. Each has a single counter-discipline that fits inside the weekly rhythm above, and each is named below with the one habit that defuses it before year three.

1. Highlighting without note-taking. PDFs accumulate in Zotero. The vault stays empty. Counter: the seven-day rule — a source read but not noted within seven days is treated as unread. Move the citation back to 00_inbox/ and re-shelve when the source-note exists.
2. Synthesis-by-procrastination. "I will synthesise everything in year three." Year three arrives and the synthesis is two years overdue. Counter: the Friday-afternoon matrix update. Synthesise this week's sources this week.
3. Concept-drift across chapters. The candidate uses "synthesis" two different ways across five chapters. Counter: the concept node's "in my use" paragraph, updated whenever the chapter draft changes. One file per concept.
4. Citation-manager bloat. Two thousand papers in Zotero, of which the candidate has read forty. Counter: the status: read | partial | skim | abandoned field in source-note frontmatter. If the source-note does not exist, the paper is not read.
5. Tool-switching mid-degree. Notion to Roam to Obsidian to Logseq, losing structure each time. Counter: the markdown vault on the filesystem. Steph Ango, CEO of Obsidian, makes the principle plain in his 2023 essay File over app: "In the fullness of time, the files you create are more important than the tools you use to create them."³⁷

A sixth, less common failure deserves naming: the over-engineered vault. Some PhDs build the vault as if it were the dissertation itself — twenty plugins, five Dataview queries, custom CSS for every callout. By month three the maintenance burden exceeds the writing rate. Counter: plain .md files plus four templates plus a citation manager. Anything else is yak-shaving.

How wiki-links carry the chapter, and where MNMNOTE fits

Wiki-links — the double-bracket syntax [[ahrens-2017-smart-notes]], originated in Ward Cunningham's WikiWikiWeb in 1995 — are the mechanism that turns a folder of source-notes into a chapter.³⁸ When the draft cites a source, the writer types [[citekey]] and gets a clickable backlink exposing every other place the source was used. The graph collects itself while the candidate writes.

The pipeline runs end to end from one library. Obsidian's Zotero Integration plugin pulls the BibTeX key, abstract, and colour-coded annotations into per-paper notes whose filenames are the citation keys. At submission, the bibliography is generated by Pandoc or a Word integration from the same Zotero library that fed the vault.³⁹

MNMNOTE is one of several editors that read this vault — plain markdown, double-bracket wiki-links, fenced templates. Obsidian, Logseq, VS Code, and MNMNOTE parse the same files; Notion, Roam, and Mendeley do not, and that asymmetry is the structural argument for plain text. MNMNOTE runs in a browser tab, stores every note locally on your own device, and requires no install or account. A first-year PhD on a borrowed laptop can have the vault running in the time it takes to make coffee; a fifth-year in submission can export every file as .md and reopen it in any editor. When the vault narrows to the literature-review chapter, the same files carry the PhD literature-review system unchanged. Zotero, Obsidian, Roam, Notion, Connected Papers, Research Rabbit, Litmaps, Mendeley, EndNote, Semantic Scholar, and MNMNOTE are all real products built by real people. Each is the right call for some candidates. Pick the citation manager you can keep using for five years; pick the editor whose wiki-link rendering you trust; keep the files yours.

Frequently asked questions

Eight questions surface in every PhD's first month with this workflow — covering scope, citation managers, AI tools, supervisor sharing, mid-degree tool switches, fieldwork data, public-vault risk, and abandoned papers. Each answer carries the working consensus from the practitioners and standards above.

How is this different from a literature-review system?

A literature review is one chapter; the research vault is the workspace for the whole degree. The literature-review system needs the source-note, synthesis matrix, and quote bank. The vault adds methodology cards, fieldwork data, chapter drafts, and supervisor meetings — eight folders, four templates, five years. The lit-review folder is one of eight. The vault outlives the chapter.

Should I use Zotero, Mendeley, or EndNote?

Zotero, in 2026, for almost every PhD. Free, open-source, maintained by the non-profit Corporation for Digital Scholarship.³² Mendeley is owned by Elsevier — acquired in 2013 for €50 million — so its plugin path is thinner.³⁴ Better BibTeX and the Obsidian Zotero Integration plugin decide it for a markdown vault. EndNote (Clarivate) is the choice where an institution mandates it.

Can AI tools build my research vault?

No, and the question is the wrong frame. AI can triage discovery, draft summaries, and surface candidate papers. The contribution is the candidate's own synthesis: the verdict column, the "in my use" paragraph, the "why not this method" disqualifier, the "my response" field. Outsourcing those makes the dissertation unsupervisable. Use AI for the edges; do the centre by hand.

How do I share the vault with my supervisor without exposing fieldwork?

Two patterns. One: keep the vault in a Git repository and share only 50_chapters/ and 70_meetings/ via a separate branch or generated PDFs. Two: keep it in a synced folder and symlink only the chapter and meeting folders. Most supervisors want the draft and agenda, not your source-notes. Encrypt 60_fieldwork/ where IRB or REC permits.

What if I want to switch markdown editors mid-PhD?

You can, and the vault is designed for it. The files are plain markdown with double-bracket wiki-links — a syntax that has survived since 1995 and that Obsidian, Logseq, VS Code, MNMNOTE, and Roam all parse identically.³⁸ Switching editors changes the reader, not the data. Most PhDs settle on one editor by year two; the optionality is the safety margin.

Where do interview transcripts and survey data go?

In 60_fieldwork/, one file per interview, with anonymised filenames and a separate encrypted lookup mapping IDs to participants. Survey data lives as CSV beside the analysis script that produced the codebook. Raw audio and consent forms stay in institutional storage; the markdown holds the transcript and codes. Cite each interview in the source-note, with type: interview in frontmatter.

Should I make my research vault public?

Almost never during the degree. Mewburn warns against publishing in-progress academic work without institutional cover, and the IRB or REC consent under which fieldwork was collected almost always forbids public hosting of identifiable data. Publish each chapter when it is published — the journal article, the book chapter, the preprint. The vault stays private; the dissertation enters through official channels.

What about papers I started and abandoned?

Keep the source-note. Mark status: abandoned in frontmatter. Write one sentence in the "my response" field explaining why. The abandonment record is itself useful — committee questions about "did you consider X?" are answered in five seconds by an honest abandoned-paper file. The paper is acknowledged, then released. Knowing what you ruled out is part of the methodology.

The discipline is the dissertation

A research vault built once and curated weekly — sources, concepts, methods, synthesis, chapters, fieldwork, meetings, templates — is the instrument that turns a five-year project into a two-hundred-week drip. MNMNOTE (mnmnote.com) is one editor that reads this vault — opens in your browser, no account, no install — and parses the same files Obsidian, Logseq, and VS Code do. The discipline is the dissertation. The files are yours.

References

All URLs verified live as of 2026-05-16.

---)