Reading with a pen, and why a highlighter is not one

Lesson 5 asked you to do a great deal inside one chapter. Classify the book, state its claim in a sentence, outline its parts, find the terms and settle them, gather the argument from the sentences that carry it.

Try that on a chapter with nothing in your hand and you'll lose most of it before the chapter ends. A term you settled on page 40 is needed on page 300. An objection you formed in the second paragraph is gone by the fourth. This lesson is about the thing that makes lesson 5 possible, which is a pen.

What a mark is for

Start with the one distinction the rest of the lesson hangs on.

A mark must record what kind of thought you had. This is the structure. This I doubt. This I did not understand. This connects to something else. This is a word the author is using in his own way.

A highlight records one thing only: something here. It doesn't say what, and it doesn't say why, and a month later you won't know either, because what made the sentence worth colouring was in your head at the time and the colour did not capture it.

That is the whole argument for marking and against highlighting, and it is not a claim about effort or neatness. It's a claim about what each record contains. A judgement is a thing you can retrieve. A location is a thing you have to reconstruct, which means doing the reading again.

Harvard Library's reading habits put the same point as an instruction: abandon the highlighter, write in the margins, develop a personal symbol system, and write down the questions you could not resolve so that they stay visible.¹

There is a second reason, and it is the one people miss. The mark does most of its work at the moment you make it, not later. To write "this does not follow" you have to decide that it does not follow. To write "he means something narrower by accuracy here" you have to notice the shift and name it. The note is a by-product of a judgement you were forced to make. Highlighting asks no judgement at all, which is exactly why it feels so restful.

A system, and a page to use it on

Here is a five-symbol system. It isn't the only possible one and the symbols don't matter, but the five kinds do, because they are the five kinds of thought that turn out to be worth telling apart.

Now a real page. This is John Ruskin, from a lecture called "Of Kings' Treasuries" that he gave in Manchester in December 1864 in aid of a library fund, and published the next year in Sesame and Lilies. It's about how to read, and you can open it free at Project Gutenberg.²

And, therefore, first of all, I tell you earnestly and authoritatively (I know I am right in this), you must get into the habit of looking intensely at words, and assuring yourself of their meaning, syllable by syllable—nay, letter by letter. For though it is only by reason of the opposition of letters in the function of signs, to sounds in the function of signs, that the study of books is called "literature," and that a man versed in it is called, by the consent of nations, a man of letters instead of a man of books, or of words, you may yet connect with that accidental nomenclature this real fact:—that you might read all the books in the British Museum (if you could live long enough), and remain an utterly "illiterate," uneducated person; but that if you read ten pages of a good book, letter by letter,—that is to say, with real accuracy,—you are for evermore in some measure an educated person. The entire difference between education and non-education (as regards the merely intellectual part of it), consists in this accuracy.²

Read it once before you look at what follows. It is the opening of one paragraph, and it is harder than it looks.

Ruskin does not stop where I have stopped him. The same paragraph runs on for another two hundred and twenty-one words, and I should tell you what is in them.

Some of it is still about reading. An uneducated man, Ruskin says, may know many languages by memory and talk them all and "truly know not a word of any,—not a word even of his own", which is the British Museum sentence again in the other direction. And a seaman has only to speak a sentence "to be known for an illiterate person", which is the plainest thing in the lecture about what Ruskin means by that word.

Most of the rest is class. Words have a "peerage", with "true descent and ancient blood" on one side and "modern canaille" on the other, a well-educated gentleman can tell them apart, and Ruskin finishes by saying that a false accent or a mistaken syllable is enough, "in the parliament of any civilized nation", "to assign to a man a certain degree of inferior standing for ever". His next words are "And this is right", though the same sentence goes straight on, "but it is a pity that the accuracy insisted on is not greater, and required to a serious purpose", and the paragraph after says it's right that a false Latin quantity should raise a smile in the Commons and wrong that a false English meaning should not raise a frown.

I cut it for room, not to protect him. It is all at the link, and it is worth reading before you take him as an authority on anything except accuracy. Lesson 7 sorts one objection to it, and finds that the method this course teaches reaches the first half of that sentence and not the second, which is a limit worth knowing about the method.

The same page, marked

Six marks. Every one of them is a judgement I had to make in order to write it.

1. Vertical line beside you must get into the habit of looking intensely at words, and assuring yourself of their meaning. That is the claim. Everything else in what I have quoted either supports it, decorates it, or overstates it, and I want to be able to find it without rereading the rest.

2. X beside (I know I am right in this). Ruskin is offering his own certainty as a reason. It is not one. Marking it's not a complaint about Ruskin, who is entitled to be sure of things. It is a note that this particular sentence gave me nothing to weigh, so at this point in the paragraph the claim is still unsupported, and what I want to know is whether the next two sentences repair that.

3. ? beside For though it is only by reason of the opposition of letters in the function of signs, to sounds in the function of signs. I have read that four times and I can restate the paragraph around it and not that clause. That is lesson 3's signature for syntax, so I did lesson 3's fix for syntax: find the main verb, find its subject, put the qualifications to one side. The structure comes apart at once, and it says that it is only by reason of one opposition that the study of books is called literature. And I am still holding nothing, because what I lack is what that opposition is. A signature sent me one way and the fix sent me straight back, which means the signature was not the diagnosis. So I ran lesson 3's own discriminator instead, which is the question I had skipped: would a definition fix it? There's no word here I cannot define. Letters, signs, sounds, opposition, all ordinary. What I am missing is a fact the sentence assumes and never supplies, that literature is built on littera, the Latin for a letter, and that letters are signs standing for sounds while sounds are signs standing for meanings. Tell me that and the clause opens. By lesson 3's three signatures that is plain missing background: I know every word and cannot say what the clause claims, and the fix is not in this text. Lesson 3's own bond checkpoint is the neighbouring case, where a word you can point at has a definition that gets you part of the way and then stops; here there is no word to point at in the first place. The ? is still the right mark to have made at the time, because none of that was available to me while I was reading. By lesson 2's rule about a first pass, I mark it and keep going, and the mark is what sends me back.

4. O around literature and around illiterate. Ruskin is using both in a sense of his own, and the text puts both in quotation marks, which is the signal worth learning to spot. The whole sentence turns on it: for him an "illiterate" person isn't one who cannot read but one who reads without accuracy, or, in the part I cut, speaks without it. If I take the ordinary sense, the sentence about the British Museum reads as nonsense. That is lesson 5's fifth rule arriving in the margin.

5. X beside you are for evermore in some measure an educated person. Ten pages read accurately, and you are permanently educated in some measure. That is a much larger claim than anything he has given a reason for, and the hedge "in some measure" is doing a great deal of quiet work.

6. Arrow beside The entire difference between education and non-education (as regards the merely intellectual part of it), consists in this accuracy, pointing to my note on lesson 5. Ruskin's route to a word is its history: later in the same lecture he tells you to get good dictionaries and hunt a suspicious word down through Greek, Latin and French. Lesson 5's rule 5 takes the other route, into the sense this author pinned down, which is often not the dictionary's at all.

   So the arrow records a difference of method, and I want to be careful about how far it goes. Two paragraphs above this extract Ruskin writes that you should "go to the author to get at his meaning, not to find yours", and then "Judge it afterwards if you think yourself qualified to do so; but ascertain it first", which is Adler's fifth rule and the whole of lesson 7 arriving a century early. After the part I have cut, he says what he wants from a word's history is the meaning "which in a good writer's work it must still bear". The two routes were always meant to end in the same place.

   So what the arrow catches is not a disagreement about where meaning lives. It's a question about how much of the work a word's history can be asked to do, and Ruskin puts the limit on it himself. In the paragraph immediately before this extract he says the words are the rock and the author's meaning is the metal you crush them for. Shortly after the dictionary passage he says the meaning he wants is the one a good writer's work must still bear. So the history is a tool in his hand rather than the place he is going.

   The live question is how often you need that tool before the book itself settles the word, and Adler's answer is almost never. That is the thing I would have lost.

Notice what the marks did not need to be. They are not neat, they aren't sentences, and four of them are single characters. Marking is not note-taking and it is not a summary. It is a running record of the kinds of thought a page provoked, made fast enough that it does not stop the reading.

The one-page record

Marks live in the margin of one book. A reading record is the page you write when you close it, and it is what you'll actually consult.

Keep it to one page and give it these seven lines.

1. The book, the edition, and the date you read it.
2. What it claims, in one sentence, in your own words.
3. Its three or four key terms, with the author's sense of each.
4. The argument, in three or four lines.
5. What you doubted, with page numbers.
6. What you did not understand, with page numbers.
7. What it connects to.

Lines 5 and 6 are the ones people leave out and they are the two that earn the page. A record of what convinced you is a record of your own agreement. A record of what you doubted and what you could not follow is a map of where to go next, and it is the thing that will still be useful when you come back to the subject in two years.

Books you don't own

Adler's practical caveat on all of this is that you should not mark a book you do not own, and it is obviously right.³

The method survives intact anyway. Keep a sheet of paper beside the book and use it as the margin, with the page number beside every mark. You lose the thing the margin does for free, which is putting the judgement next to the thing it is about, and the page number buys most of it back.

Two things make this work rather than being a chore. Write the page number first, before the thought, because a thought written without one is a thought you can't place. And keep the same five symbols, so the sheet reads the way a margin would.

What people get wrong

"Highlighting is studying." It is among the most common study techniques and among the least effective ones. This course does not re-teach that evidence, because How to Learn Anything already has it in its lessons 1 and 6. What this lesson adds is the reason: a highlight cannot record a judgement, and a judgement is the only part worth keeping.

"My notes should be neat." Neatness is a cost with no return here. A mark that takes four seconds gets made; a mark that takes thirty doesn't, and the one that doesn't get made is worth nothing however tidy it would have been.

"I will remember why I marked it." You won't, and the interval is shorter than you think. How to Learn Anything calls this the fluency illusion: while the page is in front of you the sense of understanding is strong, because the page is doing the remembering, and it goes when the page does. I have no evidence for the stronger version you sometimes hear, that the feeling is strongest exactly where the understanding is weakest. What I can tell you is that it is strong enough, often enough, that you should not spend it on remembering why you marked something.

"Marking a library book is fine." It isn't yours, the next reader inherits your judgements whether or not they want them, and a margin full of somebody else's certainties is a worse place to read than a clean page. Use the sheet.

"The marks are for later." Most of their value is spent at the moment of reading, in the judgement you had to make to write one. A reader who marks and never reopens the book has still read it better than one who did not, which is worth knowing on the days you'll not get back to it.

Practice

Connections

Lesson 2's first pass through a hard book says not to stop for what you do not understand. The ? mark is what makes that rule survivable, because it lets you keep going without losing the place you would otherwise have stopped at.

Lesson 3's three obstacles are three different marks. A word you can point at is one thing, a relation you can't see is another, and a sentence that will not resolve is a third, and marking them the same way throws away the diagnosis.

Lesson 5 is what the marks are in service of. The terms you circle become the terms you come to; the vertical lines become the outline; the passages you doubted are the raw material for lesson 7, which is judgement.

Lesson 7 comes next, and its first rule is that understanding has to come before any verdict at all. Your X marks are not yet disagreements. They are places where you owe yourself a reason.

What this lesson does not do is keep notes across years. The one-page record serves one book. A system that makes a hundred of those records findable five years later is a different subject with its own course, Note-Taking and Personal Knowledge, in Term 2. This lesson stops at the edge of one book on purpose.

Go deeper

• Harvard Library, "Interrogating Texts: Six Reading Habits". Free, and its second habit is this lesson in three paragraphs, including the instruction to abandon highlighters and the reason for writing the unresolved questions down.
• How to Learn Anything, lessons 1 and 3. Lesson 1 has the evidence on highlighting and on rereading, including the study in which highlighting hurt on questions needing inference; lesson 3 has the retrieval crossover. Lesson 6 of that course rates highlighting again alongside summarisation. This course leans on all of it and repeats none of it.
• Mortimer J. Adler, "How to Mark a Book", Saturday Review of Literature, 6 July 1940. The original statement of marginalia as thinking rather than decoration. It is short, it's still in copyright on the evidence of the renewal register, and every reproduction we could find is unauthorised, including one that carries an open-licence badge belonging to the site rather than to the essay. So this course cites it and does not link it, and if you find a copy, notice that the famous sentence is usually quoted with its opening clause removed.

Sources

1. Harvard Library, "Interrogating Texts: Six Reading Habits to Develop in Your First Year". Read in full. The annotation habit, which says to abandon highlighters in favour of marginal notes and a personal symbol system, and to write down the questions you could not resolve so they stay visible.
2. John Ruskin, "Of Kings' Treasuries", in Sesame and Lilies, 1865, Project Gutenberg #1293, public domain, transcribed from the 1894 George Allen edition, where the lecture is headed "Of King's Treasuries". The lesson quotes the opening of the paragraph and stops, and says in the body that it has stopped and what it left out. The paragraph quoted above is verbatim from that text, including its dashes and its italics on know, which are the transcription's rendering of Ruskin's emphasis. Curly quotation marks and apostrophes have been set as straight ones and the double spaces between sentences closed up, which changes no word.
3. Mortimer J. Adler, "How to Mark a Book", Saturday Review of Literature, 6 July 1940. Used here only for the caveat that one should not mark a book one does not own, given in this lesson's own words. research/SOURCES.md records this essay as read at quotation level only and not confirmed against a scan, so no sentence of it is quoted here, and it is cited rather than linked for the reason given in Go deeper. That entry also records two things worth knowing about this essay: the sentence about mutilation and love that circulates everywhere is a clipped version of what Adler wrote, and almost every copy online misdates the essay to 1941.