Building a System That Actually Sticks
Most people overcomplicate this. I watched a grad student spend three months customizing Notion databases with five separate properties per source, then abandon the whole thing because it was easier to just reread PDFs than navigate the system. The point of a DIY academic journal is not to build something impressive. It is to build something you will actually use when you are exhausted at 11pm before a deadline. Here is the setup I ended up sticking with after burning through two failed attempts. It is straightforward by design.
Ultimate Academic Journal Diy
You need three things: a reference manager, a note store, and a lightweight output space. That is it. Everything else is decoration that adds friction. For the reference manager, Zotero is the baseline choice. It handles PDFs, auto-generates bibliographies, and integrates with Word and Google Docs. I used Mendeley for a while and switched because its group library feature broke twice during active projects, and I lost citation keys that took weeks to reconstruct. Zotero's item types are annoyingly strict with certain journal formats, but it is stable enough that you can set it and largely forget it. For notes, I use Obsidian. The reason is simple: the files are plain Markdown and live in a local folder. If Obsidian ever dies or you decide to move on, your entire archive moves with you. ResearchGate-style platforms lock you in. When the subscription model shifts, your notes are trapped. Local files are the only insurance against that.
For output, keep a separate folder for draft documents. Do not mix drafts into your note vault. I learned this the hard way when a plugin update corrupted my reading notes because I had stored everything in one directory. Separation is not elegant. It prevented me from losing six months of annotation work.
Get the Full Details
The Folder Structure
Keep it flat enough to navigate, deep enough to avoid collision: 01_inbox for raw PDFs and links you have not yet processed.
02_references for Zotero-synced materials once they are tagged.
03_notes for your actual writing and synthesis.
04_projects for active papers or proposals.
05_archived for completed work. I started with a more granular structure that mirrored my department's taxonomy. It collapsed under its own weight within a month. You are not running a library. You are running your brain.
Tagging Is Where People Fail
The biggest mistake I see is treating tags as a filing system instead of a retrieval system. Tags should answer the question "what do I need to find when I am writing?" not "what category does this belong to?" Use methods as tags. Use findings as tags. Use debate positions as tags. Do not tag by author, journal name, or publication year. Zotero handles that. Tags are for connections your future self will need to make across documents. My rule: if a tag has more than fifteen items attached to it, it is too broad and is probably not helping you. Reclassify or delete it.
A Problem I Faced and How I Fixed It
About two years in, I hit a wall where my note-to-citation mapping broke down during a literature review. I had five hundred annotated PDFs in Zotero but could not trace a specific claim back to its origin in a reasonable time. The problem was that I had been pasting summaries into Obsidian without maintaining a living link between the note and the Zotero item. The workaround was not a new tool. It was a change to my import routine. I started using the Zotero-Obsidian connector plugin, which generates a proper citation key in your notes that links back to the original item. Every note now has a metadata line at the top with the Zotero link, date added, and a short tag list. It takes twenty seconds longer per import, but it eliminated the entire crisis. The extra time pays for itself the first time you need to verify a source while drafting.
Capture Versus Synthesis
Most DIY academic journal guides conflate these two activities. They are different. Capture is saving something for later. Synthesis is forcing yourself to write what you actually think about it. I used to capture everything and synthesize nothing. My vault became a graveyard of highlights with zero connective tissue. The shift happened when I started requiring a one-paragraph synthesis note for every three captured sources. Not a summary. A synthesis note forces you to articulate why the material matters to your own work, which is the only thing that survives into a paper. This rough cut usually takes an hour or two, but it is more efficient than trying to reconstruct your thinking from raw highlights later.
What This System Does Not Handle
It does not handle collaboration well. If you are working with a team on shared annotations, Zotero's shared libraries exist but the sync can lag by several hours, and conflict resolution is manual. For small groups, it is tolerable. For larger projects, consider a dedicated platform like LabArchives even if you keep your personal vault separate. It does not auto-generate literature reviews. You still have to write them. The system organizes your thoughts but does not think for you. Anyone promising automation in this space is selling something you do not need. Mobile access is limited. Obsidian has a mobile app but syncing requires either a paid service or a self-hosted solution like Syncthing. If you need read-anywhere access, keep your inbox folder synced to a cloud provider and treat mobile as a capture-only device. Writing on a phone is slow and error-prone.
What to Do Instead If This Feels Like Too Much
If the structure above feels overwhelming, start smaller. Use Zotero alone for six months. Tag liberally. Write summaries inside Zotero's note field. That is a complete academic journal for many people, especially at the early stages. Add Obsidian only when you feel the pull of cross-document synthesis pulling against the limits of Zotero's interface. Building a system is easy. Keeping it alive is the harder part. The best system is the one you maintain, not the one with the most features.
