The Previous Reader Wrote No She Wouldn't in the Margin

I buy my copies secondhand, and not for the reason people assume, which is money; the reason is that a used book arrives with an argument already in progress. The copy of Middlemarch on my desk cost two dollars and forty cents at a shop in a strip mall an hour outside Sacramento, and it contains, in a small upright pencil hand, the complete history of somebody's quarrel with George Eliot: broad agreement for two hundred pages, a hardening somewhere in the middle, and then, beside the decision the entire novel has spent six hundred pages arranging, three words in the margin — no she wouldn't — underlined twice, hard enough to have furrowed the page beneath.

I have taught that novel eleven times and I have never received a better response to it than that. I think it is wrong. I also think it is unmistakably a reading: it takes the book seriously enough to hold it to a standard, it locates the precise sentence at which the reader's trust broke, and it stakes a claim about what a person is like that the reader was evidently prepared to defend against the author herself, in pencil, alone, in a room, for no audience at all.

Now open the same novel in an e-reader and look at what the passages with the little dotted underline have in common — the ones the software helpfully reports as having been highlighted by some large number of other readers. They are, almost without exception, the detachable sentences. The aphorisms. The lines that survive being lifted out of the paragraph they came from and would look at home over a photograph of a window. What they have in common is not that they are the best sentences in the book, though some of them are; it is that they are the sentences a reader can agree with in isolation.

That is the claim, and I want to state it plainly before I qualify it, because I think it is the quiet thing that has happened to reading in my adult lifetime. The marginal note recorded an argument. The highlight records an agreement. We have swapped one for the other almost completely, and the swap has changed what the activity is for.

Consider what each one requires of you. To write in a margin you must have formed a view that the sentence in front of you is answerable — that a book is the kind of thing that can be answered. You must also be prepared to be wrong in your own handwriting, permanently, in a copy that may outlive you and be sold for two dollars and forty cents to a stranger who teaches. To highlight, you need only recognise something. The gesture is closer to nodding than to speaking, and the reward it offers, when the software tells you that four thousand other people underlined the same line, is the specific pleasure of finding you were not eccentric.

The objection here is real and I have had it put to me by people who read more than I do. Underlining is not new; readers have been drawing lines under sentences for as long as there have been printed books and pencils to hand, and the great annotators were always a tiny minority of cranks, scholars and people who could not stop themselves. The romance of marginalia is a survivorship illusion — we have the marked-up copies because somebody thought them worth keeping, and the ninety-nine clean ones went into the fire. Moreover the aggregate highlight data is a genuine gift; no scholar of the nineteenth century has ever known, with numbers, which sentences ordinary readers stopped on.

All of that is true. It does not touch what worries me, which is not the ratio of annotators to clean readers but what the equipment now makes easy. My students arrive fluent in the highlight. They can produce, on request, the eleven sentences from a novel that struck them, and the eleven are astonishingly consistent from one student to the next, and every one of them is the sentence in which a character is most confident. That last part is the thing I cannot stop noticing. The most quotable line in a novel very often belongs to whichever character the novel is preparing to disassemble. Confidence is quotable. Doubt is not, because doubt only means anything in the paragraph it lives in, and a paragraph does not fit over a photograph of a window.

So when I set a novel now, I ask for the sentence you did not believe. Not the one you loved. The one where you thought: no, a person would not do that, or say that, or stop there. It is a harder assignment than it sounds and the first attempts are terrible, because a student who has spent eight years being rewarded for recognition has no practice at all in refusal, and refusal is the entire muscle. By November it starts working. What comes back is not better literary criticism, exactly. It is the beginning of the belief that the book is on the other side of the table rather than on a shelf above it.

Under the pencilled no she wouldn't, in the margin of a book that cost me two dollars and forty cents, I wrote: she would, and that is the tragedy — you have found the place where it hurts.

I sold the copy back at the end of term. It is somewhere in a strip mall an hour outside Sacramento with two people arguing in it, and whoever buys it next is going to have to take a side.