Graduate coursework is a reading load with no paper attached. Two hundred pages a week across three seminars, discussed on Tuesday, replaced by two hundred more. The notes go into a document per seminar, or a notebook, or the margins, and by the end of the semester there are six hundred pages of notes that nobody will ever open again — because they are organized by week and by seminar, and no future question will arrive asking what did I read in week nine of the theory seminar.
The question that will arrive, two years later, writing the second chapter, is who was it that argued X — and the answer is in week nine, unfindable. The reading was done. The reading was lost.
A card system fixes this if it is built to survive, and most are not. This guide is about the reading practice itself — how to make cards from coursework reading that will still be findable and useful in a year, for a paper you have not yet imagined — and about the small number of disciplines that make the difference between a system that survives a semester and one that does not.
Why most systems die by November
They die of one of three things.
Organized by occasion. Notes filed by seminar and week are filed by the least useful key: nobody ever searches for a reading by when it was assigned. A system survives if it is organized by what the reading says, which is the only thing you will ever come back for.
Cards too big. A page of notes on a source is a page you have to reread to find the thing you wanted. A card that holds one thing can be found by its one-line title, and a hundred of them can be scanned in a minute.
No citation on the card. Six months later a card that says the state is a protection racket with no source is a sentence you cannot use, because you cannot cite it, and cannot trace, and are no longer sure you did not make it up. Every card carries its source and page from the moment it is written. This is the discipline that makes a card worth keeping, and it is the one that erodes first under a two-hundred-page week.
The cards, and how to keep them alive
The source card, filled once. Every reading gets one — citation, filled from the DOI or ISBN in a minute so that the filling never becomes the reason not to. Add the three lines from the literature-review guide: what question the piece asks, its method, its material. And one more, for coursework: which seminar, and the week — not as the filing key, but as a fact you may want later (that was the one we argued about).
Finding and claim cards, one thing each, with source and page. As every guide says. For coursework the temptation is to skip these and write a summary card per reading, because a summary is what the seminar seems to want. Write the summary if you must — as its own card, marked as a summary — but write the one-thing cards too, because the summary is what dies and the one-thing cards are what survive. A year from now you will not want to know what the piece was about. You will want the specific claim on page 214.
Your own cards, in your color, dated. The thought you had on the bus after the seminar. The connection between this week's reading and one from another course. The objection nobody raised in discussion. These are the cards a dissertation eventually comes from, and they are the ones a notebook loses fastest, because they are written on whatever was to hand. Give them cards, give them the date, and link each to the reading that provoked it.
Disagreement cards across courses. The theory seminar and the methods seminar assigned readings that contradict each other and nobody noticed, because nobody is in both. You are. A card that says Smith (theory, wk 4) and Lee (methods, wk 6) disagree about whether X can be measured is a card that, two years on, is a chapter.
Organizing by what it says
The filing key is the question the card bears on, not the week. On paper: a box, with divider cards for questions, and a card filed behind the question it answers — the same question piles as the literature review (graduate guide 1), kept between papers rather than built for one. When a new question appears, a new divider. When a card bears on two questions, a second card with a pointer, or — in software — two tags.
Two things about the questions. Let them emerge. Do not design the taxonomy in September; you do not know the field yet. Add a divider when three cards want one. By the end of the year the dividers are a map of what you actually think the field's questions are, which is worth more than any list you would have written in advance. Let them be wrong. A divider that turns out to hold two different questions is split; two that turn out to be one are merged. The cards do not change.
Beside the questions, keep one more thing: an index card per source with its cards' titles listed. This is the reverse lookup — from a source to what you took from it — and it is what answers what did Smith actually say without rereading Smith.
The disciplines that make it survive
Five, and none of them costs more than a minute a day.
- Cite on the card, at the time. Never later. Later does not come.
- Title every card in one line. A card you cannot title is two cards (student guide 3), and a card without a title cannot be scanned.
- File within the week. Cards left in a pile on the desk for a month are a pile, not a system. Ten minutes on Friday.
- Date your own thoughts. Not the reading cards — those carry the source's date — but the cards in your color. A dated sequence of your own thoughts across a year is the record of how you came to think what you think, and it is the thing a dissertation's introduction is written from.
- Keep everything. The reading that seemed irrelevant is the reading you will need. A card costs nothing to keep and cannot be recovered once thrown out.
What the system gives you
By the end of the first year: eight hundred to two thousand cards, filed by question, each citable, each one thing, with your own thinking threaded through them and dated. That is not a set of notes. It is the reading for the qualifying exam, organized; the beginning of every literature review you will write; and — in the disagreement cards and the cards in your color — the raw material of a dissertation topic, which most students arrive at by staring at a blank page in year two and which you will arrive at by reading your own box.
By the end of the degree, the box is the thing you take with you. The dissertation was one arrangement of it.
The bag on the bus
A system survives a semester only if it can be fed from wherever the reading happens: the library, the bus, the kitchen at midnight. On paper this means a small stack of blank cards in every bag and the discipline to write the citation on each. In software it means the same thing — that a card can be made from a phone, from a PDF's selected passage with its page, from a web page's selected text with its address, or from a photographed page of a book in the stacks — and that a card made offline on the bus is on the desk by morning without anything being done to it. A card that could only be made at the desk is a card that mostly does not get made.
Backing it up
A box of cards can be lost in a move. A system in software can be lost to a dead laptop or a cloud account that lapses. The rule is the same for both: a copy, on your own disk, that you can open without anyone's permission. For paper, a photograph of each drawer every term. For software, an export of everything — every card as a plain file, with its fields and its links — into a folder you control, on a schedule you do not have to remember. Software built on these principles writes that export by itself, and the test of it is whether you can open the folder in any text editor and read your cards. If you cannot, it is not a backup.
Doing this in software
Software that works from cards does the parts of this that erode on paper: the citation filled from an identifier or a barcode so that citing costs nothing; a passage from a PDF or a web page becoming a card with its page or address already on it; a card made on a phone, offline, and synced; search across every card in every project; tags for the questions, and saved views for the questions you keep asking; and an archive written to your own disk on a schedule. It keeps your own cards dated by the moment they were made. What it does not do is decide which question a card bears on, notice that two seminars' readings disagree, or have the thought on the bus. Those are the reading. The software is the box.
Everything here works with a pack of paper cards. Leyden & Beam does the same steps on screen, keeps every order you try, and never writes a word for you.
Start your 14-day trial