Skip

Less Wrong: The Best Textbooks on Every Subject
March 25, 2012 11:59 AM   Subscribe

For years, my self-education was stupid and wasteful. I learned by consuming blog posts, Wikipedia articles, classic texts, podcast episodes, popular books, video lectures, peer-reviewed papers, Teaching Company courses, and Cliff's Notes. How inefficient! [...] What if we could compile a list of the best textbooks on every subject? That would be extremely useful.
Less Wrong, a community dedicated to rationality, is compiling a list of The Best Textbooks on Every Subject.
posted by Foci for Analysis (49 comments total) 170 users marked this as a favorite

 
Cool... now I'm of course going to have to look at all four of the books recommended for the subject I'm teaching a one-day class on in June, and decide if the recommendation is valid. It's possible I have trust issues. Or maybe I'm just into compulsive over-preparation...
posted by SMPA at 12:03 PM on March 25, 2012


I learned by consuming blog posts, Wikipedia articles, classic texts, podcast episodes, popular books, video lectures, peer-reviewed papers, Teaching Company courses, and Cliff's Notes.

Ok. As much as I consider most of those sources to be valid sources, I just can't give anybody any kind of credibility when they tell me they were educated from blog posts, Wikipedia, podcasts, and Cliff's notes.

That's the equivalent of, well, Cliff's Notes.
posted by karathrace at 12:06 PM on March 25, 2012 [2 favorites]


karathrace, totally depends on the topic, no? I do programming and things move so fast that you're lucky if there is a blog post of Stack Overflow Q&A to guide you. Also, the OP isn't the one who compiled the list, the entire LW community has.
posted by Foci for Analysis at 12:08 PM on March 25, 2012 [5 favorites]


It would indeed be extremely useful to have a "best textbook" on every subject. However, that would not obviate reading Wikipedia, classic texts, popular books, etc because:

1) You don't learn something by reading about it once in a canonical form and then going on with life. You need to read about it here, hear someone talking about it there, try it out somewhere else, go back and read something you read earlier in light of what you now know and so forth. Multiple sources of information give you a better perspective.

2) Not everyone learns the same way. One person's classic is another person's dry tome and vice versa.
posted by DU at 12:09 PM on March 25, 2012 [16 favorites]


Isn't there a highly-favorited Askme thread like this? How about an autodidact tag? Also, eponysterical.
posted by StickyCarpet at 12:11 PM on March 25, 2012 [1 favorite]


Why is learning from peer-reviewed articles a problem? Textbooks take a stance and a theoretical perspective just like everything else. Just because it purports to teach doesn't make it objective.
posted by ChuraChura at 12:16 PM on March 25, 2012 [2 favorites]


I'm sure the IP address for this website will be blocked in Texas.
posted by hippybear at 12:16 PM on March 25, 2012 [5 favorites]


The AskMe in question.
posted by Jpfed at 12:21 PM on March 25, 2012 [5 favorites]


I'm sure the IP address for this website will be blocked in Texas.

To be fair, its only because they don't like reading.
posted by karathrace at 12:22 PM on March 25, 2012


I'm sure the IP address for this website will be blocked in Texas.

To be fair, its only because they don't like reading.


You obviously haven't followed the news about the Texas Board Of Education and the way they're meddling in textbooks, have you?
posted by hippybear at 12:26 PM on March 25, 2012


Meanwhile, let's note once again how absurd the prices are. Textbook publishing is an utterly broken model, at least from the students' perspective.
posted by twsf at 12:30 PM on March 25, 2012 [5 favorites]


Less Wrong asks the wrong question. The AskMe thread asks the right question.
posted by stbalbach at 12:38 PM on March 25, 2012 [2 favorites]


Isn't the whole point of being an autodidact learning from sources you've assessed (at least somewhat) independently? If you're looking for a list of the "best textbooks" on a particular subject, find a professor whose views you trust or agree with, and then get one of his or her syllabi. Simple. Why trust a crowd of people with day jobs to determine your learning material when there are thousands of people nationwide being paid to thinking about exactly this problem as it relates to their field of study?
posted by anewnadir at 12:41 PM on March 25, 2012 [3 favorites]


I'm picturing a compendium of textbooks/great books on a variety of topics, with graphics. So you click/find the topic that you want, and you get several rotating books which provide a great selection of material to read, from basic to a bit more advanced, as well as a commonly regarded 'bible' on top, in case you only want to read one book.

That would be helpful, because I tend to do a ton of reasearch on topic of the day, guided to a variety of sources from Amazon reviews to blog posts, to forum threads. Then I typically buy the top few books in the fray. I do this on topics from bread baking to organic gardening to quantum mechanics (for dummies). I can see the value of this sort of thing, I just wish it was better organized/presented.
posted by tatiana131 at 12:41 PM on March 25, 2012 [1 favorite]


The AskMe thread has 237 answers and 1585 favorites and is vastly superior to the (interesting) LessWrong thread. The hive mind at this website has a huge resource advantage.
posted by bukvich at 12:49 PM on March 25, 2012 [2 favorites]


I was gonna ask how classic texts, video lectures, and peer-reviewed papers could be replaceable by textbooks, until I realized that most of the subjects are fairly technical. For less technical subjects, this approach would eliminate what's good about classic education, which is that you're immersed in the worldviews of the original authors and don't take their fundamental assumptions for granted.
posted by stoneandstar at 12:50 PM on March 25, 2012 [1 favorite]


Less Wrong asks the wrong question. The AskMe thread asks the right question.

Specifically, the AskMe poster is pretty clear about what they mean by "best" -- it seems to me the LessWrong poster isn't.

I had a bad feeling about that reading the post, and looking at some econ related recommendations and commentary deepens it.
posted by weston at 12:51 PM on March 25, 2012


Its in threads like these I wish a certain .nu website still existed... :(
posted by Another Fine Product From The Nonsense Factory at 1:02 PM on March 25, 2012 [4 favorites]


Must approve of the recommendation of Bishop's Pattern Recognition and Machine Learning text. One of the finest books I have on my shelf.

On preview: looks like people are going to take away many different things from this list.
posted by JoeXIII007 at 1:12 PM on March 25, 2012 [1 favorite]


Textbooks take a stance and a theoretical perspective just like everything else. Just because it purports to teach doesn't make it objective.

I don't think objectivity has much to do with it. A good textbook starts with what you'll need to know to go deep, then they take you deep. Not much else (other than a class or tutor) does that. Most things are either introductory, or shallow, or assume you're already an expert. Trying to overcome that shortfalling by reading a lot of those things can result in a patchy and poor grasp of the subject.

Textbooks aren't as good as being taught by someone who knows the field, but they're better than a lot of things you can do on your own.
posted by -harlequin- at 1:15 PM on March 25, 2012 [6 favorites]


Meanwhile, let's note once again how absurd the prices are. Textbook publishing is an utterly broken model, at least from the students' perspective.

Solved.
posted by -harlequin- at 1:19 PM on March 25, 2012


Another Fine Product From The Nonsense Factory: "Its in threads like these I wish a certain .nu website still existed... :("

"The disappearing virtual library" (Al-jazzera)
posted by stbalbach at 1:45 PM on March 25, 2012 [2 favorites]


"A good textbook starts with what you'll need to know to go deep, then they take you deep. Not much else (other than a class or tutor) does that. Most things are either introductory, or shallow, or assume you're already an expert."

Exactly! In my younger days I used to read a lot of nonfiction just for the hell of it, and this was a constant annoyance. A thousand introductions and histories of the subject, but very few sources of detailed knowledge. Textbooks are very challenging and time consuming, but they're the main course in any knowledge feast.
posted by Kevin Street at 1:54 PM on March 25, 2012


StickyCarpet: "Isn't there a highly-favorited Askme thread like this? How about an autodidact tag? Also, eponysterical"

It's linked to on the page, under "elsewhere".
posted by Red Loop at 2:01 PM on March 25, 2012 [2 favorites]


What the "best textbook" is depends so heavily on your level of understanding the moment the textbook is picked up. I don't really know of a single textbook that bootstraps anyone from novice to expert in one continuous set of chapters.

The best solution to this is to make all books available for free. Then you can download a bunch, sample and jump around and find what you need. Then come across a new problem, sample and jump around, etc. Google books helps a lot in the searching. Other sources help a lot in the actual reading. Shame about most of those other sources lately.
posted by Chekhovian at 2:17 PM on March 25, 2012


And by the way, how exactly will this tie into Less Wrong's Harry Potter Fan fiction?
posted by Chekhovian at 2:19 PM on March 25, 2012


I don't really know of a single textbook that bootstraps anyone from novice to expert in one continuous set of chapters.

I don't know about finally achieving "expert", but O'Reilly books are really good ways for novices to break into a huge variety of computer subjects and come out with pretty good working knowledge.
posted by hippybear at 2:28 PM on March 25, 2012 [1 favorite]


I have to say the advice on the textbooks for the History of Western Philosophy was pretty good.
posted by oddman at 2:33 PM on March 25, 2012


I will now have to read The Great Conversation. I still think A History of Western Philosophy is a fantastic, fantastic read and a great summary of the major philosophical movements, and was disappointed to see that the linked article that reviewed the "inaccuracy" of Russell's book did so on the grounds that it left out Eastern philosophy. Yes, the History of Western Philosophy did not cover Eastern philosophy and therefore it is inaccurate. And this from a reviewer of philosophical books.
posted by Lutoslawski at 2:42 PM on March 25, 2012 [2 favorites]


So, that review lists the lack of Eastern philosophy as an "omission", not an "error"/"inaccuracy", and points out that Russell doesn't only not cover it, he "[sweeps it] aside as mere unstructured opinion." The errors seem to be the logical fallacies he complains about at the end of the review. The review agrees with you that it's a great read, however, one that "still stands head and shoulders above the rest."
posted by Casuistry at 3:10 PM on March 25, 2012


I'd like LessWrong more if they were weren't so evangelical. Take this as an example - what would be wrong with listing good books on a subject? Why do they have to identify *the best*? And why not suggest original sources, instead of textbooks? Critical reading is a fundamental academic skill, but this thread implies that knowledge would be best transmitted directly from one brain to another.
posted by Joe in Australia at 3:41 PM on March 25, 2012


Like several people observe above, I think that "best" isn't a very useful word. What does it mean?

I've been taking college math classes for the last couple of years. Unlike various sciences I've studied, where the gap between reading the textbook and reading current articles is pretty small, math is mostly a textbook subject. So what makes a "good" math textbook?

Virtually all college math starts with a calculus sequence, and virtually all introductory calculus textbooks cover the same material in identical order. They have similar explanations, similar proof outlines, and similar exercises. Students disagree widely and vocally in their assessments of any given textbook. Some see the book for what it is: a neutral carrier of the subject's rudiments. Others consider the book as an affront to their dignity and intelligence, or an unfair impediment, or a useless ream of ink-smeared paper.

So, what's the "best" introductory book for calculus? A standard, colorful, illustrated, 1000-page tome (of which there are dozens, in dozens of editions)? Something that dumps students directly into terse rigor? A soft book that lets students get by on intuition?

Then, of course, students go on to more advanced classes, and start running into books that command enormous respect and are justly considered classics. But a lot of these books expect a lot from the reader and demand a lot of work before they give up their rewards. Here is an Amazon review of Strauss's book on partial differential equations for undergraduates. Here is the reviewer's later retraction of his negative comments. When was the reviewer more correct in his/her assessment?

And let's not forget the question of scope: writing on the same standard subject, say, linear algebra, one author could take an analytic approach and another could take an algebraic approach. They will define standard concepts from different perspectives, reach important results in different order. In one author's framework, result A will be obvious and result B will require a bit of proving, while in author B's framework this will be reversed. Again, which one is "best"?

In math and other dry subjects, at least, it'd be much more useful to make a list of time-tested classics and let students make of them what they will. Everyone likes Walter Rudin in retrospect; no one likes working through their Rudin for the first time. Everyone likes Herstein on algebra, and so on. It's not as if no one knows what the classics of introductory math are supposed to be. But the accepted pillars of the subject are hardly the "best" reading for all readers at all times.

Expecting to find a "best book" is evidence of unrealistic expectations. It obscures the fact that learning is hard work, and it is never complete.
posted by Nomyte at 3:46 PM on March 25, 2012 [4 favorites]


And why not suggest original sources

I had this thermo prof that expected us to read the original papers on things from like 1910...good god...not a useful thing to do most of the time. And have you tried reading The Principia? He derives everything using geometry, because that's all everyone else could do at the time, even though he figured out everything using calculus. Sometimes you gotta kick away the ladder and use secondary sources, it tends toward masturbation otherwise.

what would be wrong with listing good books on a subject?

C'mon dude, be serious...would be discussing if it were an effort to list things that are decent, or kind of okay, or pretty reasonable?
posted by Chekhovian at 3:50 PM on March 25, 2012


I don't know about finally achieving "expert", but O'Reilly books are really good ways for novices to break into a huge variety of computer subjects and come out with pretty good working knowledge.

I think programming is a somewhat atypical example. And most of the O'Reilly books I've read generally seem to start assuming you have familiarity with the ideas of programming in general. I don't recall them spending vast amounts of time on the philosophy behind loops and conditional logic. But yeah, they are great books. They even have one on Mathematica. It's pretty good.
posted by Chekhovian at 4:02 PM on March 25, 2012


> For years, my self-education was stupid and wasteful. I learned by consuming blog posts,
> Wikipedia articles, classic texts, podcast episodes, popular books, video lectures,
> peer-reviewed papers, Teaching Company courses, and Cliff's Notes. How inefficient!

The autodidact-since-college-anyway mind I know best cannot help learning from whatever it's exposed to. Wherever it's coming from, it's coming in like a tidal wave that never stops and all one can hope to do is grab an occasional gasp of air. If learning from wikipedia is wrong, that means mustn't ever look at wikipedia. OMG, learning might occur!
posted by jfuller at 4:05 PM on March 25, 2012


I think the point of listing the "best" textbooks is just to save time, for people who are looking for a deeper understanding of the subject than the average layman, but who are not themselves experts. Reading one textbook will never be as good as taking a college course or earning a degree, but it can give you a pretty decent mental map of a field of thought.
posted by Kevin Street at 4:08 PM on March 25, 2012 [1 favorite]


Actually, thinking some more about this issue, what's frustrated me in recent years is the lack of "intermediate" resources in some fields. Say you have a reasonable grasp of some issue but want to learn more its often very inefficient to build your way up to expertise. Most lower level resources will mostly reproduce things you already know, while many upper level things will be too advanced or too specialized to be useful. At least the way it worked for me was that I either spent vast amounts of time skimming through lower level resources looking for the few new snippets I didn't miss, trying to fill in the inevitable gaps one ends up with in ones knowledge, or hours and hours staring at terse high level stuff that fails to point out the the implicit lower level assumptions it makes.

I suppose trying to become an expert on something is inherently time consuming, I just wish it didn't feel so inefficient.
posted by Chekhovian at 4:16 PM on March 25, 2012


Autodidacticism is an efficient method for consuming large amounts of raw data. Transforming this data into useful knowledge requires thoughtful interpretation, which is most often achieved by discussing and arguing with others about these interpretations...
posted by ovvl at 5:35 PM on March 25, 2012 [1 favorite]


And why not suggest original sources

Oh, and never try to learn Fourier transforms by reading Fourier's original work. Its the mathematical equivalent of picking up a neolithic hand axe to chop your wood. It may work, but you won't like it.
posted by Chekhovian at 8:19 PM on March 25, 2012 [1 favorite]


And have you tried reading The Principia? ... it tends toward masturbation otherwise.

Well, it's kind of fitting considering Newton's views on relationships.
posted by Jpfed at 8:24 PM on March 25, 2012 [1 favorite]


Newton's views on relationships

I suspect that in that case absence of evidence probably did mean evidence of absence/abstinence, or maybe you drink enough mercury and weiner doesn't work anyway.
posted by Chekhovian at 8:32 PM on March 25, 2012


I just can't give anybody any kind of credibility when they tell me they were educated from blog posts, Wikipedia, podcasts, and Cliff's notes.

That (and peer reviewed papers) is pretty much where my entire knowledge of immunology comes from. Of course I see antibodies as a chemical reagent in a laboratory assay so the notion they might also be used by the immune system to fight infection seems almost quaint to me.

On the other hand, my younger brother is working on his dissertation in Biology right now and the other day I had to explain B cell repertoire to him because he didn't have clue one about it. In all fairness, he's not doing anything with jawed vertebrates, so the idea of an adaptive immune system probably seems quaint to him, but I'd have thought it would have come up somewhere along the line.
posted by Kid Charlemagne at 9:08 PM on March 25, 2012


Ah. Glad to see some love for Griffiths' text on electrodynamics in there, if only for the reason that it's literally the only time I've ever put down a Physics text, and thought "Gee, that was thorough and clearly-written."

The people who write Physics textbooks are bloody awful at it.

(Feynmann's a bit of a wildcard, because his approach and formulation are pretty different from anything else out there; I wouldn't call this much of an exception though, because his books aren't particularly useful if the professor or student isn't familiar with Feynmann's particular style... I suppose that I shouldn't frown too heavily upon what appears to have been a noble attempt to reform the tired, broken, and needlessly-opaque state of Physics jargon, but I never really bought into the hype that surrounded Feynmann's published lectures...)
posted by schmod at 10:11 PM on March 25, 2012 [1 favorite]


Would this be a fair summary of adaptive immunity? (I came up with analogy when I was fairly young and I haven't gotten around to checking it since, so please go easy on me if it's nonsense)

Imagine an infection as a horror movie. There's some terrible pathogen that's going around killing innocents. A few guys with their heads about them, already known as badasses, try their hand at slaying the beast, but none of the established methods work! Fire, bullets, not even crushing what seems to be the head. The beast keeps reproducing, and villagers keep dying.

Some people get their bodies taken over by the beast's hideous larvae. Most of the villagers are too paralyzed with grief and horror to deal with this, but Mr. T is stone cold enough to do what has to be done, perforating anyone taken over with his trusty shotgun.

The Beta family- a bunch of outcast weirdos into body modification and stuff like that- have been observing all of this, and they know it's time to act. All of them try out different methods to kill the monsters (which by now are everywhere)- most of which don't necessarily make much sense, but when the shit has so thoroughly hit the fan you've got to get a little crazy. And as soon as one of them figures out that the monsters are vulnerable to forks dipped in turpentine, he tells the whole village. Soon, everyone who will listen is dipping forks in turpentine and sticking forks in monsters, and the monsters are driven back!

Now, the next time a monster comes to town, they'll try fire, bullets, crushing the head, and forks dipped in turpentine. And most of the time, that works. Except when it doesn't, and we get a sequel.
posted by Jpfed at 10:20 PM on March 25, 2012 [1 favorite]


"For years, my self-education was stupid and wasteful. I learned by consuming blog posts, Wikipedia articles, classic texts, podcast episodes, popular books, video lectures, peer-reviewed papers, Teaching Company courses, and LessWrong. How inefficient!"
posted by Termite at 4:56 AM on March 26, 2012


"The people who write Physics textbooks are bloody awful at it."

The introductory books are pretty good (And numerous!), but the higher up you go, the worse the writing gets. Maybe because those books are written by busy scientists who don't want to waste time explaining things when they can just drop an equation or graph in there.
posted by Kevin Street at 1:08 PM on March 26, 2012


The phrase "a community devoted to rationality" conjures for me some kind of chilling dystopian Village of the Damned scenario.
posted by ThatFuzzyBastard at 8:55 PM on March 26, 2012


Lepore's logic text is good, but it's a weird pick for best logic textbook because it's so unique. It has very little about derivation and proof, and an awful lot about translation. It's more linguistics than logic... not what most philosophers have in mind when they think of what a logic course should be.
posted by painquale at 9:27 PM on March 26, 2012


I think it does depend of the topic. I got an A in a political science class by using Wikipedia and never opening the book once. I do not understand it of course. I'm sure we all know people who have degrees and don't seem to understand anything. I think cliff notes and Wikipedia are great when taken in context of what they are made for.
posted by booklovers at 10:36 AM on March 27, 2012


« Older Trouble in Ponydise   |   "The Hunger Games" costume design Newer »


This thread has been archived and is closed to new comments



Post