How to Start a Nonfiction Book: What to Do in the First Two Minutes

You Open the Book at Page 1, Like Everyone Else

You're holding 340 pages on a subject you actually want to learn. What do you do in the first minute?

Almost all of us do the same thing: we open to the first page of text and start reading.

Except that book has five things printed in it that tell you almost everything that's coming. They're right there, free, and you skip them because they don't look like reading. The table of contents looks like an administrative list. The preface looks like a formality the author owed someone. The last chapter looks like a crime.

This isn't a named method with a certificate behind it. It's what anyone who's read enough nonfiction eventually starts doing, so the book stops catching them off guard.

Five Things, In This Order

The table of contents. You don't skim it like a list — you read it like the outline of an argument. The chapter titles laid end to end are the skeleton of the case: where the author starts, what they go through, where they land. Watch for groupings: if eighteen chapters sit under three parts, those three parts are the thesis. And watch the proportions: if the subject you bought the book for gets one chapter out of twenty, you've just learned in ten seconds something you'd otherwise have discovered on page 200.

The preface or introduction. It's the only place where the author talks about the book, not from inside it: what they're trying to prove, who it's for, and, in the good ones, what it doesn't do. The last paragraphs of an academic introduction almost always contain a chapter-by-chapter summary, written by the person who knows them best.

The last chapter or the conclusion. Here comes the objection: I'm spoiling it for myself. A nonfiction book doesn't have a twist to spoil — it has a conclusion the author spends 300 pages defending. Read first, the rest of the book becomes the argument for something you already know. Only the direction changes: instead of gathering pieces and hoping they add up, you're checking a claim you already hold in your head.

The index, if it has one. Two lookups, not ten. Take the terms that matter to you and see how many pages they point to. A concept with twenty-three references is the spine of the book. One with a single reference is mentioned in passing, no matter what the cover promised.

The year of the first edition, not the printing you're holding. For some subjects it's a footnote. For others it's the verdict: a book about memory written in 1998 and one written in 2019 aren't the same object, even if they share a title on the spine.

The Part That Does the Work Is the Questions, Not the Map

If you stop here, you have a map. The map is useful, but it isn't what makes the difference.

Do one more thing, the one that takes thirty seconds: write down three questions you want the reading to answer. In your head works; in the margin works better. Not questions like "what's this book about," but questions with an actual answer: why does the author claim working memory is the bottleneck, not attention? What did they look at to say that? What counterargument do they dodge?

The difference between reading with questions and reading without is the difference between searching and receiving. When you search, you have a stopping criterion, and you notice when you've hit it. When you receive, you move through pages with a vague sense of understanding, and that sense can't be contradicted until someone quizzes you.

The questions have a second job, later on: after each chapter, you answer them from memory, book closed. That's the opposite of rereading, which produces familiarity without producing memory. I wrote separately about why you reread a chapter and retain nothing from it.

This procedure even has a name, fixed steps, and an acronym: SQ3R, from 1946. Surveying and questioning are its first two letters, and I broke it down step by step separately. You don't need the acronym to look at a table of contents.

Why It Has to Come Before, Not After

The clearest experiment on this is fifty years old, and it involves a balloon.

Bransford and Johnson (1972) gave students a paragraph written in plain words, with flawless grammar, and completely incomprehensible: someone trying to send a message to a distant window, with a speaker hanging from balloons. Without context, participants rated its comprehensibility at about 2.3 out of 7 and recalled around 3.6 ideas out of 14.

A second group saw, before reading, the picture that explained the scene. Ratings jumped to about 6.1, and recalled ideas to 8 out of 14.

The decisive part is the third condition, the one summaries tend to leave out. One group got the exact same picture, but only after they'd finished reading. Their recall stayed at the level of the group that got nothing at all. The information was identical. Only the timing had changed.

That's why the two minutes belong at the start, and why you don't get any of it back by reading a summary at the end. The frame has to already be there when the text arrives, because the frame decides what from the text actually sticks.

To be fully honest: the balloon paragraph was built deliberately opaque, and a chapter on monetary policy isn't nearly that closed off. The effect on a real book is smaller, even if the direction has repeated across dozens of studies since.

The Arithmetic of Two Minutes

The claim is that the two minutes pay for themselves within the first ten pages. Nobody has measured that exactly, so look at the numbers yourself — they're simple.

Ten pages of nonfiction run to about 3,000 words. At 238 words per minute, the adult average for nonfiction from a meta-analysis of 190 studies (Brysbaert, 2019), that's roughly 13 minutes. To claw back two minutes from that, you'd need a gain of about 15% over that stretch. Or, more realistically, just not rereading two paragraphs because you didn't know where they were going.

Over the whole book, the stakes are even smaller: 340 pages is about seven hours of reading, and two minutes is half a percent of that. The bar is so low that "is it worth it" stops being a meaningful question.

Speed isn't even the main payoff. Often what the two minutes tell you is "this book isn't about what I thought" or "I only need chapters 4, 7, and 11." That's where hours get saved, not percentages.

There's a deeper reason too. Just, Masson, and Carpenter (1980) compared graduates of speed-reading courses with readers who simply skimmed. On familiar texts, the graduates caught the gist better. On technical texts they knew nothing about, the advantage vanished: prior knowledge was doing the work, not the technique. Rayner and colleagues' 2016 synthesis lands in the same place — the biggest lever is how much you already know about the subject, not how you move your eyes. The two minutes are the cheapest way to manufacture a bit of that "already."

Fiction Is the Exception, a Research Paper Is a Different Protocol

For fiction, you do almost none of this. The author, the year, maybe the first page. That's it.

The reason isn't sentimental: in a novel, the order in which you learn things is the content. A mystery read with the ending already known isn't a more efficient version of the book — it's a different book. Masson (1982) showed that readers who skim stories keep the thread but lose exactly the details that turn a scene into a scene. In nonfiction, that loss is acceptable. In fiction, it's the whole point.

For a research paper, the order flips again. The abstract is the conclusion, printed up top. Then the figures and their captions: in a serious paper, the result lives in the figures, and the caption is the densest sentence in the whole text. Then the discussion, where the authors say what they think it means and, in honest papers, what they can't claim. You read the methods when you want to check the claim, not to find out what the paper is about.

What the Two Minutes Don't Give You

First trap: after surveying, the words feel familiar, and that produces a feeling of knowing. Recognition is cheap and has nothing to do with being able to reproduce something. If you close the book after the contents and the preface, you have a map of a territory you've never actually walked.

Second: the map is the author's. The table of contents shows you the structure they want you to see, and an author hiding behind vague chapter titles hands you a false map.

Third, and it concerns what comes next. Surveying decides what deserves careful attention, but it doesn't read for you. Duggan and Payne (2009) showed that, under time pressure, a fast pass over the entire text beats careful reading of half of it. Under time pressure. When you have time, careful reading wins, and surveying is what aims it. I've written separately about exactly where skimming stops working.

None of this makes you read faster line by line. It makes you read the lines that matter, with a question in mind. Next time you open a new book, give yourself two minutes before page one, and see what's different by chapter three.

See what a reading plan for students looks like