How to read a (good) book in one hour.

I’m teaching a class at Harvard this semester, and was satisfied to discover that even there, students balk at my tendency to assign what seem like humanly impossible amounts of reading for class. Because I basically just want everyone to like me, I frequently backpedal on this otherwise gratuitous display of pedagogical power. I do however, intend for students to read the books I assign, and because I am myself an incredibly slow reader, I sympathize with my students, especially when I assign things I haven’t read and I think: now why did I do that? But I also find that it helps to make explicit what I mean by “read.” For this purpose I have frequently relied on Paul Edwards excellent handout, “How to Read,” which counsels that one can read a book in 6-8 hrs. I’m upping the ante, though, because I think you can read a book in one hour and gain a reasonably good sense of the argument, stakes and main areas of investigation, even if the details elude you. Here ‘s my snake-oil, use only as directed, and I disclaim all warranties, implied or expressed.

To read a book in one hour requires a particular kind of book. It works for most scholarly books, especially in history and social science, the denser the better. It works less well for books in philosophy or for heavily argument driven texts that require the reader to follow along (I would not recommend trying to learn calculus in this manner). More importantly, it requires the apparatus that a scholarly book gets when it is published– i.e. it does not work for dissertations, drafts, self-published works or poorly-published ones. Indeed, a well-crafted scholarly book is fantastic machine, one that can be readily approached, understood, extended and critiqued. In this era of the crisis of scholarly publishing, it seems to me that presses should be doing a lot more to indicate that they can turn an otherwise messy manuscript filled with hard-to-find but good ideas into a scholarly hot rod tricked out with everything necessary to teach generations upon generations, connect up scholarly communities, and parse out complex topics into loadable modules of delicious knowledge. Publishing a scholarly book is not about making it available– it is about making it readable, and this is what you pay for, or should be paying for anyways. If you can follow these steps, especially with a work of history or ethnography, then the book is a well-produced scholarly work.

How to read a book in 1 hour.

  1. Read the whole book at once: Start by flipping through it, read the TOC, the preface and forward. if there are any, look for subheadings and for a general sense of whether the book has internal divisions (parts, chapters, subheadings that do not appear in the TOC), and whether it has a conclusion or other kinds of sections, interludes, or breaks in the text. Browse the notes to see if they contain merely references or extended parts of the argument. If the book does not contain an index, you can stop here: the only thing left to do is sit down and read from cover to cover, as slow or as fast as you permit yourself.
  2. Turn to the index.
  3. You will make two lists. Begin by looking for the largest entries, those indented with sub-headings, and lots of page references. Write them all down: people, places, things, concepts. In a normal academic tome (300ps) there should be anywhere between 10 and 30 pages of index, so this list can range from 5 terms to more like 100. But really, start with the longest and most detailed, which should yield a good list. This is your list of the main subjects and problems of the book.
  4. Now go through the index again, and look for entries that do not have subheadings, but have more than 3-4 page entries. Some authors go crazy with the subheadings, so the first list might be a lot longer than the second, other authors (or index makers) are content to list everything once, with page refs. You have to exercise some judgment here. If your first list is very long, then for your second list pick out those entries which are not people, institutions or events, but analytic or conceptual designators– i.e. look for entries that are analytic sounding: “assemblages” “neo-liberal shenaniganism” “trading zones” “network forums” etc. If your first list is very short, it very well might already contain these terms, and the second list will be a list of people, places or things that reappear throughout the book.
  5. Note at this point that you have two lists of terms which you can use in class to remind you of the details, even if you haven’t yet read the book. The index is the Platonic ideal of the text, use it.

  6. With your lists in hand, turn to the Introduction. But don’t start at the beginning. Read the last few pages of the introduction, where most likely there will be a series of paragraphs here dealing with the content of each of the chapters. Read carefully, noting which chapters relate to which entries on your two lists. If your author has chosen to express their individuality here and forgo such a list, you can wing it by looking at the beginning and end of each of the chapters to see whether the author gives you a hint there.
  7. Note that you still haven’t “read” very much yet, but that you should already have a deepening sense of the main themes of the book, and a map, complete with precise coordinates of where to find the main arguments and the main subjects of the book.

  8. Now read the introduction carefully. Make sure you are clear what the author thinks the main arguments and sub-arguments are, and that you could reconstruct them if asked, even if you can’t offer any details or reasoning behind them.
  9. If there is a conclusion, read that carefully too. I know this sounds like cheating, but it isn’t. It is a rare scholarly book that demands of its reader that they wait until the end for the argument to make sense. {Aside: Indeed, many graduate students make this mistake in writing, assuming that it is necessary to defer and defer and defer until you get into the very heart of the most detailed detailage before revealing the a-ha! of the argument. No no no, give it up, right at the beginning and let the reader work through your example to convince themselves you are correct!} Read the conclusion for how it tries to tie up the arguments presented in the text (which you haven’t yet read) with the promises made in the introduction. Note especially if the author makes clear what the significance of the argument is beyond the text, which will help you care about the details.
  10. Now return to your two lists. The shorter of these two lists (the one with the analytic entries) should now give you a very good guide to where the theoretical meat of the book lays. Having read the intro and conclusion, you can now turn directly to each of those sections (you have the technology!) and “read from the inside out.” The longer list (filled with people, places and things) in turn gives you a good sense of where the data is, and how it is distributed across the chapters (if you go back and look at all the subheadings in the index). “Reading from the inside out” means literally starting in medias res, looking for the precise places where the author has made it a point to connect theory and data. Read the paragraphs leading up to it and following it. Note the references to empirical material marshaled or referred to, and decide which of those things you need to read more about– turn to list two, and find the places where you can follow up. After running through the entries of the shorter list, you will have read a fair amount of the most important parts of the book.
  11. Note that this approach is fractal in nature: with a good index you can make progressively longer and more focused lists that give you “random access” to the text, and allow you to dig deeper and deeper until you approximate the actual cover to cover manner in which a text seems (wrongly I hope I have convinced you) that it was meant to be read.

Needless to say, this is a strategy that works only for good books, and for books that are primarily dense with detailed empirical material, which most histories, ethnographic and other forms of social science research usually are. It is less useful for philosophical works, completely useless for books that do not have indices (like much work in French! damn them!), and it will only confirm the badness of a bad book. However, if you are faced, as many students are, with reading as many as 4-5 books in one week, this is one way to avoid ending up in a class with a vague sense of what a book is about and a detailed understanding of only the first 30 or so pages. I am of course curious to hear from people how this approach fails.


Christopher M. Kelty is a professor at the University of California, Los Angeles. He has a joint appointment in the Institute for Society and Genetics, the department of Information Studies and the Department of Anthropology. His research focuses on the cultural significance of information technology, especially in science and engineering. He is the author most recently of Two Bits: The Cultural Significance of Free Software (Duke University Press, 2008), as well as numerous articles on open source and free software, including its impact on education, nanotechnology, the life sciences, and issues of peer review and research process in the sciences and in the humanities.

24 thoughts on “How to read a (good) book in one hour.

  1. Funny. We used to call this “The Graduate School Method” back in the day, although with some modifications.

  2. Oh my god this is so opposed to everything that I stand for in education it isn’t even funny. Ah well, I suppose this means I’ll have to write up my own method explicitly.

  3. As a reader, I would tend to have the same reaction as Rex. As a current graduate student at Harvard, I’ll be applying this method to tonight’s reading!

  4. While in principle I agree with Rex on this, still, this sort of method is actually a good way for students to orient themselves to a work, and to read a work more quickly and with a greater strategic appreciation for what the work is trying to do, even when they intend to do a proper thorough read. As well, even sticklers for close, detailed reading of texts (and I tend to be one of these…) use methods like this when orienting rapidly to a new field of literature, for example – I use techniques like this regularly in order to sift through massive piles of material to ferret out the works that I’ll then read in great detail…

  5. omg, this method is so opposed to everything I stand for in education too! I guess that’s why we get along so well, Rex, and why you too will eventually join me on the dark side of the force.

  6. I think it is fair to say that not all books need to be read in the same way. CK is very clear that this approach is only suited to a specific type of literature, read for particular purposes. One could similarly write a piece entitled: “how to choose what to read” about how to read a book at the bookstore before deciding whether to buy it. Personally I feel that most anthropology books published these days don’t deserve to be books, but are only published as such because of the pressures of our discipline. I often advise students to see if authors published an “article version” of the book, as I find the shorter format often forces authors to think more clearly about what exactly it is that they want to say.

  7. Moral judgments aside, many people (myself included) find themselves with reading loads that are totally impossible if they use their preferred or habitual reading style. I think this is fantastic for those sort of keeping your head above water kinds of needs. I’ve done things like this on a catch as catch can basis but it’s useful to have it systematized and the written lists thing is a really good idea for preparing oneself to talk about the book.

  8. I think there’s some great advice here. Yes, it offends the sensibilities of crusty academics like myself who live and love books. Here’s the thing: Given enough time, students should definitely read the whole book. Given less time (and they often are — I know there were plenty of weeks in grad school whee I was expected to read 3 full books and an array of supporting materials in a week, all while pursuing research for my paper in the class, and working full-time to boot) students should do a “foreshortened” reading, getting the TOC, intro and conclusion, dipping in to a couple chapters, etc. Given even less time, something like the above is good. The point is, something’s better than nothing, and the more something you can manage in the time you have available, the better. The point is to get as much as possible with the resources a student has.

  9. Hmm, I applied ckelty’s methods to this article, and I didn’t understand a thing. Am I doing something wrong?

  10. This is disgusting. If I read this entry while doing web sweeps for potential job applicants, your application would be pulled from the pile.

    That said, I teach similar tactics, but for very different ends. I teach my students to use similar skimming and evaluation techniques to decide if a book or article will have something important to offer them and their given research project. I’d fail any student who did this and claimed they were “reading” a book.

  11. good lord if ya’ll are digusted by this, wait ’til I tell you how I teach the students to grade their own papers. I can’t even remember the last time I read a student’s paper 😉

    and then there’s my method for teaching them to reason by avoiding valid inference of any kind and my method for picking a research topic by using the LOC Wheel of Fortune, and my method for reading by running an automatic co-occurence algorithm on a text and talking only about the top five results.

    don’t you people read Derrida: il n’y a pas d’hors du text dude. Move beyond hearing-youself-read…join me, together we will rule the galaxy!

  12. As a grad student, I’ve tried variations on this method for around 3 years. I haven’t found a way to make it work for me. I’ve never found a book with an introduction + conclusion that I can read in an hour. I think one major problem is not just if the book is heavily argument-driven, but also how familiar the reader is with the material. New types of arguments–which often come with new vocabularies–simply take more time to process.

    If you’re assigning so much reading that the first pass becomes the only pass, then I disagree that Edward’s method is a good way to read ethnography. I feel that what gets lost is the interaction between the ethnographic details and the theory, which is where the strength of ethnography lies. If it’s an undergrad class, you deprive them of the opportunity to discover what makes ethnography great, and if it’s a grad class you encourage the kind of over-theorizing that is the bane of early grad studies in anthropology.

    I’m replacing my snarky last paragraph with it’s essential observation: I think a syllabus that pushes the limits of possibility in terms of reading, especially coupled with the (in my experience false) promise of Edward’s method, has a high risk of conveying a strong sense of failure and inability to students. This can be heightened by things like the difficulty of the material or one’s isolation from one’s peers.

    There is certainly a benefit from learning to get a broad but shallow grasp of a topic quickly, and I wouldn’t suggest that ambitious reading schedules or some training in how to quickly assess the arguments of a book is bad. However, I think that such a pedagogical approach needs to be applied carefully. Profs need to acknowledge the kind of demand that this places on students, both educationally and emotionally. Students need to know that this kind of course is an exercise in broad, shallow learning, and given course goals and objectives that match the approach. The limits that this approach places on learning should also be acknowledged, either by incorporating other approaches over the duration of the course, or by letting students know upfront that there is more information than they will be able to fully process, and that the course acts as an introduction and guide to their further studies.

  13. carmen, you make a good point about coming to the information for the first time– certainly for students completely unfamiliar with some form of writing or some form of argument, there is a steep uphill learning curve, and I appreciate that. However, I of course am still correct 🙂 because I think the alternative for most students is that they spend far too much time on the wrong parts of a book, for precisely this reason. They do not learn first where to read in a book, and instead submit entirely to it as if it were a novel that required starting at page 1 and soldiering through, never skipping, until that crucial final paragraph that makes it all clear. Not all books (indeed, very few I would argue) actually build an argument from beginning to end. Most are more like collections than single unified arguments, and the index is a powerful way to find out which kind of book you are holding.

    I also object to your taunt that I am teaching a broad shallow course. If you only knew how little I was teaching them 😉

    but seriously, there are some important disciplinary differences, I think, in that in history courses what looks like broad and shallow is in fact an attempt to array a number of different perspectives– so about 5 of the 13 required books in my class rehash exactly the same material, but in different ways– if my students don’t have a deep understanding of that material by the end, i certainly will have failed. By contrast, in this course, the focus was on carefully reading texts (and only 4 of them!) that proposed innovations in a theory, and the historical aspect of “public spheres” about which there is a GINORMOUS amount, was left out… I think those students also went deep, but in a different way…

  14. Sorry ckelty, I didn’t mean “broad and shallow” as a taunt or insult at all. I think it’s a fine thing, and any given course negotiates a tension between “broad and shallow” and “narrow and deep”. But I think that any course that pushes the boundaries in terms of reading load inevitably leans towards the “broad and shallow” side of things (caveat: I’m assuming a short course, 3 or 4 months). I think it’s inevitable that a student’s grasp of any individual author’s perspective is going to be shallow if they have, say, 4-10 authors to read each week for 12 weeks.

    I think, though, that you’ve side stepped my point a little, which was more about the effects of unrealistic expectations combined with unrealistic work loads than about the objective value of teaching students to quickly grasp the gist of an academic argument (I agree with you on that point).

    From my perspective, the one-hour reading timeline is ridiculously over-optimistic. I got some new books today, and I tried a perusal of the index, not actually writing the page numbers, but noting the entries according to your steps 2-4. It took me 8 minutes to get from A to C. Optimistically this leaves me 1/2 an hour to read the introduction and the conclusion (a total of 63 pages) “carefully”. I take “carefully” to mean, basically, that you have a strong grasp of what the author is saying, understand to some degree why they are saying it, and grasp the logical processes of their arguments. Honestly my “careful” reading speed is around 2-6 minutes per page, depending on my familiarity with the material, which puts me at 2-6 hours just to get through the intro and the conclusion. I have really tried to increase that, but the truth is that my comprehension plummets at 30s-1minute per page.

    The reason that I brought up the whole emotional/educational demands argument was because of your own description of your course load as an “otherwise gratuitous display of pedagogical power”, and I suspect that it is the rather cavalier attitude towards your students needs (which in turn I suspect was more to add some humour or levity to the article than any actual callous disregard on your part) that has left some of the commenters here making such strong statements.

    However, there is something there in your first paragraph that I think needs to be addressed generally. I think that what we lose touch with as we progress through the academic ranks is how poorly prepared we often are for what is expected of us after transitions (1st year undergrad, 1st year grad, etc.), and how difficult the process of learning how to learn can be. The sentiment behind what you’ve said here is good–students should get some guidance in how to aproach the mounds of information in our world–but the way that we take this sentiment and turn it into teaching needs to be done with careful attention to the level of the students, to the tools that we can realistically expect them to have, and with attention to where idealized methods for learning break down. If you suggest to your students that they should be able to read a book in one hour, you are placing a tremendous personal responsibility on them to handle what you admit is an ambitious course load at best. When they are unable to do it (a near certainty) if you haven’t laid a groundwork for more reasonably expectations, you risk teaching them that they are failures instead of that they need to learn (or how to achieve) efficiency in reading.

    I feel like I’m going on at way too much length here, but it’s an issue that I’ve given a lot of thought to on my travels through ridiculously unrealistic course loads. I really don’t mean to be insulting, and if you are sensing anger or frustration, it is directed at my own experiences, not at your course.

  15. carmen, well spoken, and I agree that I haven’t directly addressed the issue of unrealistic expectations. And yes the pedagogical power quip, while meant as a joke, does contain that grain of truth that a teacher’s own insecurity can translate into unrealistic expectations and a forgetting of just how hard it is to learn how to learn, as you put it. I like to think I am aware of this, and I have had the luxury (for which I give thanks every day) of always teaching small classes in which I am allowed to shower attention on the students, both in and out of class. The result is that I usually catch the cases where students are struggling, and try to find ways to help them, if they are willing to be helped. Indeed, I gave this method to my current class as a response to a student’s concern, not just as some introductory lecture meant to indicate that they better keep up or else feel worthless… but I can certainly sympathize that this is not the experience everyone has of unrealistic expectations… and in larger classes, I can see how the method might look like some kind of stupid algorithm that would be a substitute for real teaching. I certainly shouldn’t make assumptions about the kinds of teaching and learning that people get based on my own experience of elite universities…

  16. This supports everything I support in education. Learning how to learn is necessary. Reading every word of a text is not necessary. We are not hermeneutic Biblical or Qur’anic scholars. There is nothing holy about an entire book. Books are collections of information, and there a several ways of extracting and digesting that information. There is certainly something to be gained from reading a well-written book from start to finish. There is also something to be gained from skimming a book or using ckelty’s method. An ethnography of the fetishization of reading among “crusty academics” would be interesting. The semiotic ideology at work here is fascinating.

  17. This is actually great advice! I plan on using it in the future, even if I plan on reading the book, this is a great way of getting an idea of what the author has to say/their essential arguments and ideas.
    I don’t understand why people are so offended, its actually kinda funny!!!! but thank you ckelty!

  18. So wait, you all are in a course together LOL does the professor know who you all leaving comments actually are?

  19. Pingback: Senses and Sounds
  20. You can also save time by tearing out the index and marking the parts that would go to form your lists. This is not advisable if you don’t own the book, of course.

  21. Mortimer Adler was flogging something very like in How to Read a Book, seventy years ago. Good idea then, good idea now.

  22. This is not the graduate school method. The graduate school method is:

    0) Read the author’s wikipedia page.
    1) Look the book up on jstor. Pull some reviews from the NYRB and the like and read them.
    2) Then, read some more articles from jstor and the like
    3) Prepare three or four incisive comments. Venture two, and keep one or two in reserve if pressed.

    You’re set.

Comments are closed.