Turning eleven sources into an argument

The reading is done. Eleven sources: four books, five articles, two things from the web you are not sure you are allowed to cite. You have notes on all of them, some of them good. What you do not have is a paper, and the notes are not turning into one.

This is the most common place a research paper gets stuck, and it gets stuck for a reason nobody tells you: reading produces summaries, and a paper is not a summary. Eleven summaries laid end to end are an annotated bibliography. An argument is something else — a claim, the reasons for it, and the evidence for the reasons — and it does not come out of the sources by itself. You have to build it, and the building has steps.

This guide is those steps, on index cards. You can do it with a pack of cards and a table. If you use software that works from cards, the steps are the same.

Why the notes will not turn into a paper

Look at your notes for one source. They run down the page in the order the source said things. That order was the author's argument, not yours. Every source's notes are in that source's order, and eleven arguments in eleven orders do not add up to a twelfth.

To make your own argument, the pieces have to come out of the sources' order and go into yours. That means the pieces have to be separable. Notes on a page are not; cards are. So the first job is not writing — it is cutting the reading into pieces that can be moved.

Step 1: Cut each source into cards

Take one source and its notes. Make a source card: full citation, and a number — S1. Then go through the notes and make a card for each thing worth keeping:

• a quotation card for an exact passage you might quote, with the page and "S1" on it;
• a fact card for a finding, a number, a date, in your own words, with the page and "S1";
• and, when the source makes a point you want to argue with or lean on, a card that says what the source claims, marked as the source's claim, not yours.

One thing per card, one line of title across the top. If a note holds two things, it becomes two cards. (Guide 3 is about that line and why it matters.)

Do this for all eleven. It takes an hour or two, and it is the hour that makes the paper possible. When you finish you have somewhere between forty and a hundred cards, and — this is the point — none of them is in any order at all.

Step 2: Sort by what they are about, not where they came from

Spread the cards out and sort them into piles by subject. Not by source. Two cards from different books about the same thing go in the same pile. A card that does not fit any pile starts a new one.

You will end with somewhere between four and ten piles. Give each pile a title on a card of its own, face up on top: the cost of the canal, what the merchants wanted, the 1830 vote.

Something has just happened that could not happen in your notes: sources that never met are now sitting together. The article from 1998 and the book from 2015 are in the same pile because they are about the same thing, and if they disagree, they are now disagreeing in front of you instead of on pages 40 and 212 of two different documents. Disagreements between sources are the raw material of an argument. Mark each one you find with a card that says what the disagreement is.

Step 3: Turn each pile into a claim

Now take one pile. Read its cards. Ask one question: what is true about this that the cards show?

Write the answer as a full sentence with a verb, on a fresh card. Not the pile's title — the title was a topic. The sentence is a claim. The cost of the canal is a topic. The canal cost three times its estimate, and the overrun was known before the vote is a claim.

Put the claim card at the head of the pile with the evidence under it.

Some piles will not yield a claim. The cards are interesting and they show nothing in particular. That pile is background: it may give you a sentence of context somewhere, but it is not part of the argument. Set it to one side. Some piles will yield two claims that contradict each other; keep both and mark the pair, because a contradiction you can name is halfway to a thesis.

When you have done every pile, you have a row of claim cards, each with its evidence beneath it. That row is the skeleton of the paper. It came out of the sources, but it is not in any source.

Step 4: Find what is missing

Look at the row and count.

A claim with one card under it is a claim with one witness. Either the witness is very good, or you need another — go back to the reading with the claim in hand, which is a different and much faster kind of reading than the first pass, because now you know what you are looking for.

A claim with nine cards under it, all from S3, is S3's claim, not yours. It is fine to agree with a source. It is not fine to summarize one and call it a paper. Find what S1 and S7 say about the same thing, or narrow the claim to what you can show from more than one place.

Evidence cards that support no claim go in a pile of their own. Do not throw them away. Do not force them in. A paper that uses every card is a paper that was written to use the cards, and it reads that way.

Two of your eleven sources may turn out to contribute nothing. That is normal; research reads more than it uses. The sources you did not use are not wasted — they told you where the edges of the subject are — but they do not have to appear.

Step 5: Put the claims in an order that argues

A row of claims is a list. An argument is a list in an order, where each claim makes the next one easier to believe.

Try the obvious order first: background, then the main claim, then the objections. Read the claim titles aloud in that order. Then try another: the surprising claim first, the background where it is needed, the objection answered last. Cards let you do this in a minute. Do it three or four times and keep the order that reads best aloud. Guide 6 is about this step, because it is the one that turns a competent paper into a good one.

The thesis — the one sentence the whole row exists to support — is usually visible by now, and it goes on its own card at the top. If it is not visible, guide 4 is for that.

Step 6: Write from the row

Write the paper in the row's order, one claim at a time: the claim as the paragraph's first sentence, the evidence beneath it as the paragraph's body, the counterpoint beside it answered in the next paragraph. Between claims, one bridge sentence saying why this one follows from the last.

The introduction is written last, because until now you did not know what you were introducing.

What you have when you finish

A paper with a claim, reasons, and evidence from more than one place — which is what "an argument" means — and beneath it, a row of cards that shows where every sentence came from. If anyone asks how you got from eleven sources to this, the cards are the answer, in the order you made them.

Doing this in software

Software that works from cards does the same steps with less table. Each source's citation comes in from its DOI or ISBN, so the source card writes itself; a quotation card carries its source and page as a link, not a handwritten "S3"; a pile is a tag or a pile on a board; and a claim card with nothing under it is something the software can point out, because a link that is not there is a fact about the structure. Nothing in the method changes. The table just gets bigger and the cards stop falling on the floor.