Fiziks vs. Humanities: The greatest book never read
By Thoreau
Newton’s Principia Mathematica is probably the most important science book ever written. Despite that, I’ve only read a few excerpts from it. I have the Big Blue translation sitting on my shelf and I keep meaning to read it, but it’s hard to get around to. I love the parts that I have read, and I want to read the rest, but it’s never an urgent item for me.
In physics, our attitude is that the original text doesn’t matter much, just the ideas that have stood the test of time. If somebody else can write a book that makes it easier to understand, we teach from that. So we declare Newton’s Principia to be the most important physics book ever (because of the ideas it put out there) but we feel no need to read much of it.
This is totally different from humanities. In humanities, people make a point of reading the original thinkers. They don’t just say “Well, philosopher so-and-so influenced lots of other people and got the ideas rolling, so let’s read somebody influenced by him and maybe a Cliff’s Notes version of the original.” They actually sit down and read the original, or a translation of the original.
I guess the difference is that humanities scholars are interested in people and ideas, while we evil fiziks types claim to be interested in the external world. But despite that, we teach physics as a series of theories with different domains of applicability (e.g. Newtonian mechanics, relativity, quantum mechanics, relativistic quantum mechanics, ray optics, wave optics, quantum optics, etc.). There’s a lot of history behind this, with the older theories (usually) being taught first. We also engage in a ton of hero worship, telling all sorts of stories about the great physicists. (A few of those stories might even be true!) Despite all that reliance on history and that culture of revering our great predecessors, we never sit down and read the original works, a habit that would be second nature to a humanities scholar.
I’m not saying we should always do that, or add a bunch of original works to an already-crowded curriculum. But it is interesting just how rarely we do it.
Comment by Aaron Boyden —
May 17, 2009 @ 6:42 pm
The most important contributions are generally new ways of looking at things. New ways of looking at things in the sciences often involve tools in some way, or at least mathematics. Tools get better, and it makes sense to use the latest tools. Mathematical approaches don’t always improve as much (though they certainly improve), but the original way they were described is rarely critical.
In the humanities, new ways of looking at things rarely involve tools. Thus, it is less likely that subsequent developments will put you in a position to grasp the original insights much better with methods of your own. Insights also seem to be harder to come by in the humanities (possibly partly because you generally can’t do it by finding a new tool, or a new use for an old tool or an old mathematical trick), and harder to understand. Trying to figure out what the originator of the insight was thinking is a very valuable way of trying to grasp the insight, in a way it is much less likely to be in the sciences.
Comment by rj —
May 17, 2009 @ 7:42 pm
Dr. Thoreau, meet Robert Merton. Certainly not the only person to discuss the issue, but the only one whose book I happened to be skimming for other purposes just this past Friday.
Comment by hf —
May 17, 2009 @ 8:01 pm
Interesting post. In engineering, when we’re mocking physicists in class (yes we do that) one of the themes is how you’re too wrapped up in who originally discovered what. It appears that the closer a discipline gets to immediate practical application, the less interest there is in the history.
Another thing about the humanities though – their work is, to some extent, the writing itself, and proper valuation of that requires reference to the original text. You’ll notice that humanities articles often have block quotes, something you’d rarely see in science and never in engineering. It also shows up in attitudes toward plagiarism – we tend to view it as wrong mainly because the student has evaded doing their own work, but for the humanities profs, it’s stealing, and it drives some of them to real lip quivering rage.
Comment by Seward —
May 17, 2009 @ 9:10 pm
Thoreau,
In humanities, people make a point of reading the original thinkers.
Actually, that isn’t necessarily the case. It depends on the discipline and the various viewpoints within disciplines.
Comment by Seward —
May 17, 2009 @ 9:14 pm
Thoreau,
I guess one can trace that back to the onset of the Ancients vs. Moderns conflict.
Comment by kid bitzer —
May 17, 2009 @ 9:30 pm
as someone in the humanities, i’d say the practice ranges pretty widely, as do the motivations.
i mean: why does a specialist in french renaissance lyric keep rereading villon and the pleiade?
how about: cause they’re not making that stuff any more?
it’s not really an option to use a more up-to-date or streamlined presentation of that same content, the way we use leibniz’s notation for the differential instead of newton’s. the notation, so to speak, is the content.
maybe in other parts of the humanities it’s different: philosophy might have some constant content that would be amenable to improved presentation.
there, i wonder whether it’s true that all philosophers read the originals. how much did kant know about plato? how much did he care? how much does, i don’t know, john searle, know about aristotle? how much does he care?
historians–well, they have to read old stuff if it’s their source material. but if you want to get a really deep understanding of agincourt, do you spend much time reading victorian discussions of it? yeah, you have to read the gesta henrici. that’s like a geologist looking at precambrian rocks. but do historians have to read stuff between now and then?
anyhow–if i knew more about the humanities, i’d keep trying to make the case that there is no general “humanities” attitude towards reading the greats.
Comment by Aaron Boyden —
May 17, 2009 @ 9:53 pm
Philosophy is my field, and I can report that philosophers are still interested in Plato and Aristotle. Kant was certainly interested in them; he thought Aristotle was the last word on logic. I tend to think myself that most of Aristotle and some bits of Plato are obsolete because we have in fact made progress in logic, but Plato was a freak who understood so many things so much better than almost anyone else that he remains worth reading even though you have to slog through a few obsolete bits. At least he usually makes them fun to read as well. (Reading Aristotle, on the other hand, is IMHO usually a waste of time, but my opinion on this is certainly far from universal among philosophers).
Comment by John Emerson —
May 18, 2009 @ 1:01 am
Science tends to be cumulative and decisive in a way that the humanities are not. That is, many concepts and theories (e.g., phlogiston) are refuted once and for all, and many other concepts and theories (the inverse square law of gravity) are confirmed once and for all.
To me as a non-scientist, the advantage for scientists of reading the original texts, especially in well-edited annotated editions filling in the background, would be seeing how great scientists of the past actually worked. People in general, including many working scientists, often have mystified, mythic, idealized (and to a degree, erroneous) pictures of the history of science.
Comment by DCA —
May 18, 2009 @ 2:59 am
The Principia is pretty heavy going; a good alternative, that helps the modern reader without losing any of Newton’s procedures, is “Newton’s Principia: The Central Argument” by Densmore and Donohue.
On the more general point, it is also true that only rarely is the first statement of something in physics the clearest version:
this applies to Maxwell even more than Newton. In part the problem in both cases is that the mathematical concepts that now seem most applicable to the ideas being presented were not available when the ideas were first presented.
Comment by josephdietrich —
May 18, 2009 @ 5:01 am
I suspect a lot of it is also the fact that the ideas in many of the humanities can be lost or distorted by summation or restatement in a way that F=ma cannot.
Comment by kid bitzer —
May 18, 2009 @ 6:05 am
#10–
i disagree. a lot of rectilinear motion gets lost in translation.
Comment by Phersv —
May 18, 2009 @ 6:13 am
There are some significant books in mathematical logic that almost nobody reads but which are still essential: Frege’s Begriffschrift or Russell and Whitehead’s Principia Mathematica are good examples. What they did was very important but few logicians have to read the original version anymore, apart from a few selected parts.
Comment by Seward —
May 18, 2009 @ 7:06 am
kid bitzer,
I don’t really know much about it beyond what I read in my historiography course, but I could imagine that if one does cliometrics that one could largely avoid traditional sources.
Comment by mds —
May 18, 2009 @ 8:40 am
[Doffs hat to kid bitzer]
Comment by kid bitzer —
May 18, 2009 @ 9:00 am
vielen dank’, mds.
but it only stands to reason:
poetry is what gets lost in translation;
there’s poetry in motion;
ergo…
#13– i wonder if the cliometricians are intentionally striving to be more scientific. i.e., maybe they would also vindicate thoreau’s original proposal, that there’s a correlation between inquiring scientifically and shedding one’s reverence for the earliest statements of one’s discipline.
(i also think some humanists are easily fooled into thinking that they can dress up in intellectual lab-coats as it were, throw a bunch of numbers at a topic, and get a spiritual promotion into the ranks of scientists. to my mind, that betrays a different confusion about what makes science science.)
Comment by Tom Scudder —
May 18, 2009 @ 10:09 am
As I understand it, people explaining Kant are a lot more read than Kant himself.
Comment by mds —
May 18, 2009 @ 11:06 am
Well, I think that’s more of a social science phenomenon, which is usually distinguished from the humanities per se*. And even then, it depends. Many sociologists, for example, are serious about observation, data collection, and statistical rigor. On the other hand, modern economics seems to be dominated by people who swagger because they use calculus, which seems to be much more of the “Look! I’m a scientist” mindset. So we get, for instance, JK Galbraith derided by many of his younger fellows for not being mathematically sophisticated enough to build elaborate irrelevant-to-false models rather than wasting his time on observation and description of the economy.
*I would consider cliometrics to be more of an inclusion of historical data in economics / econometrics than the domain of historians with maths. Though “I’ve run the simulations, and Julius Caesar does seem to be inevitable” would be cool.
Comment by kid bitzer —
May 18, 2009 @ 11:16 am
yeah, i agree with all of that (i.e. 17). my point was that cliometricians do not (so far as i know) make a good case of humanities types who ignore their disciplinary forebears.
“I’ve run the simulations, and Julius Caesar does seem to be inevitableâ€
oddly enough, people tell me that this was what Leibniz was aiming for. (or maybe: thought we could do if only we knew as much math as god knows.)
Comment by Thoreau —
May 18, 2009 @ 11:29 am
That’s why I read this stuff. I want to see them work their mojo so I can get some inspiration.
Interestingly, I find that the best stuff in a lot of the great works tends to be at the beginning. Carnot’s most important arguments were early in his treatise. I’m reading Alhacen (Ibn Al Haytham) now, and maybe my attention span is short, but it seemed like his most effective argument was his use of what we now call Occam’s razor early in book 1 to argue that vision does not involve rays sent out from the eye. The snippets of Newton that I read involved the centripetal force, and I found his explanation as convincing and understandable as anything else I’ve seen.
That’s not so different from modern science. Whereas once upon a time they’d produce a book, now we produce a series of papers. Often the first few papers in the series are the most important.
Comment by rm —
May 18, 2009 @ 11:39 am
a spiritual promotion into the ranks of scientists
I think that’s more of a social science phenomenon
Some people in colleges of education are humanists pretending to be social scientists pretending to be scientists.
I always point out to the Freshman Composition inmates that MLA requires the page number even when you don’t quote, but in APA you can breezily wave your hand at the author’s name and cite the year it was published. MLA wants the first name spelled out and the Title Capitalized, but doesn’t care that much about the year. APA uses names like ID codes and puts the year up front. There are the two cultures, and I insist to them that it’s all fascinating.
Comment by Neel Krishnaswami —
May 18, 2009 @ 11:56 am
To be fair, if you have to read one thing by Frege, the Begriffschrift is less important than his Grundlagen der Arithmetik. This is because his philosophy is in the Grundlagen, and the Begriffschift is almost purely symbolic.
To apply his ideas in a modern context, you need the latter, and not the former.
Comment by Derek Copold —
May 18, 2009 @ 12:13 pm
That’s why I read this stuff. I want to see them work their mojo so I can get some inspiration.
More important, it would seem to me, is to see what they got wrong. Not for the sake of sneering, but for the sake of caution. These were great men, no doubt, but they were human and liable to human error.
Comment by kid bitzer —
May 18, 2009 @ 12:26 pm
i’ve read parts of galileo’s ‘two dialogues’, and certainly enjoyed watching a really smart dude think out loud.
did it teach me how to do science? no, but then that’s not what i do anyhow.
would it teach college students, esp. those interested in science? possibly; might be worth a try.
at the same time, i think the record of the st. john’s college system in this regard is not stellar. their kids actually read euclid and the principia. and i don’t think they turn out a lot of scientists. but it’s not a great test case anyhow, since the applicants are so highly self-selected. the young scientists burning to get into mit don’t apply to st. johns. what would happen if you took mit kids and had them read al haythm and carnot and galileo? dunno.
Comment by jim —
May 18, 2009 @ 1:15 pm
I don’t know about Newton, but it is worth reading Einstein’s original papers. I don’t think I would want to learn the subject from them, but once you have the ideas, it’s interesting to see how he thought about them.
Comment by mds —
May 18, 2009 @ 1:47 pm
First, this is awesome. Second, my hazy (hazen?) recall suggests you’re probably right, if you’re primarily interested in the optics groundbreaking. Though Alhazen’s Problem is interesting to mathematicians, and there’s stuff for fans of the history of medicine in the Book of Optics, too. Sigh. I look at the work of medieval Islam on things like hydraulics, chemistry, the structure of the eye, etc, and I wonder what might have been. Perhaps we should ask the cliometricians.
That’s because AFAIK they only read the Old Masters and reproduce a few of their experiments. It seems more an exercise in snootiness than a way to genuinely learn the scientific method, mathematical tools, etc. Euclid is probably not particularly effective as a way to study geometry or proofs to begin with.
Comment by dhex —
May 18, 2009 @ 2:20 pm
in terms of discussed more than read within the humanities – or anywhere, really – i would think kuhn is near the top.
followed maybe by the bible?
Comment by Seward —
May 18, 2009 @ 2:56 pm
mds,
Well, I think that’s more of a social science phenomenon…
Where humanities and social sciences divide is pretty blurry.
Anyway, I always figured Galbraith should be made fun of for advocating price controls.
Comment by someotherdude —
May 18, 2009 @ 3:15 pm
@ mds:
“Perhaps we should ask the cliometricians.”
Well I know folks in the humanities and social scientists who would say, colonialism is a bitch! and it stunts growth and transfers much more social and material wealth to those doing the colonization…Islam was most inventive when it was the main colonial actor.
I don’t know, just a thought.
Comment by Seward —
May 18, 2009 @ 3:32 pm
mds,
Reading Euclid might be a good way to study Hobbes however.
Comment by kid bitzer —
May 18, 2009 @ 4:10 pm
i’m pretty sure that reading euclid is much better as a way to learn euclidean geometry than reading newton is as a way to learn newtonian mechanics.
(but i’ve never tried to teach classical mechanics from the principia, so don’t know first hand).
Comment by Gary Farber —
May 18, 2009 @ 6:23 pm
“…to argue that vision does not involve rays sent out from the eye.”
This is so disappointing.
Comment by VM —
May 18, 2009 @ 7:30 pm
@ doktor T 11.29am
Comment by mds —
May 18, 2009 @ 8:44 pm
Well, the rules obviously don’t apply in the case of your mutant powers, Mr. Farber. Please don’t zap us with your eye rays.
Comment by Ian —
May 18, 2009 @ 9:46 pm
Newton makes for ugly reading, though the underlying argument is gorgeous. (Philosophy students read Newton, since we seem to have taken the history of science on ourselves) For beautiful writing that made a major contribution to science, the guy you want to be reading is Darwin.
Reading Aristotle, on the other hand, is IMHO usually a waste of time, but my opinion on this is certainly far from universal among philosophers.
Among philosophers, you’re in the extreme minority. If you’re interested in the history of philosophy, well, there’s a reason why the medievals called him The Philosopher. If you’re interested in what philosophers are working on right now, it’s worth mentioning that quite a lot of current work in ethics is avowedly Aristotelian.
Aristotle tends to be pretty dry stylistically, but he’s not nearly as bad as Kant. The Nicomachean ethics is something even normal people might want to consider reading.
Comment by Donald Johnson —
May 18, 2009 @ 10:06 pm
“For beautiful writing that made a major contribution to science, the guy you want to be reading is Darwin.”
And the cool thing about Darwin is that he’s still worth reading even from a strictly scientific (and not just historical) viewpoint, or so I’ve heard. (I’ve read Darwin myself, but wouldn’t want to speak on behalf of evolutionary biology but iirc actual biologists have said this.) He’s outdated and wrong on some subjects (notably heredity), but, for instance, people arguing about punk eek still find that he in some sense he was there first.
Comment by Kevin J. Maroney —
May 18, 2009 @ 11:35 pm
A lot of literary theory is read in interpretation rather than the original. Not everyone agrees with Judith Butler that the language of literary theory has to be dense and inpenetrable, and there are a great many books explaining the more recent French theorist like Levi-Strauss, Derrida, Lacan, Baudrillard, or the earlier great incomprehensible or difficult German literary critics Kant, Nietzsche, Freud, and Jung.
Comment by Randolph —
May 19, 2009 @ 1:51 am
Sometimes reading original papers of experimentalists helps in understanding the ideas and limitations of the experimental work. (I am thinking of work–which I am not going to cite here–which has serious sampling problems.)
In design and the arts, one studies “precedent” to learn process; this is why artists still trace the old masters, for instance. This is also true in software engineering and the various design disciplines that depend on it: if one wants to find out how the WIMPs GUI was invented, and why it has been so widely adopted, there is no better place to go than the original researchers, and if they are not available, go to their papers and code. Perhaps this is less important in more mathematically pure disciplines.
Comment by Randolph —
May 19, 2009 @ 1:06 pm
And today we have Krugman on reading economists in the original. One thought comes to mind: in a field where so little is reliable, and there is so much bad research, it’s important to go back to the good sources.
Comment by Seward —
May 19, 2009 @ 1:19 pm
Randolph,
EconTalk just had a special multi-part discussion of Smith’s The Theory of Moral Sentiments.
Comment by Seward —
May 19, 2009 @ 1:19 pm
That you can read along with…
Comment by D.A. Ridgely —
May 19, 2009 @ 3:51 pm
I’m shocked, shocked I tell you, that someone around here has been reading Frege! Whether Mr. Krishnaswami is correct or not about old Gottlob, perhaps we can at least agree that no one should ever have to wade through all of Principia Mathematica (or, while we’re on the subject, all of Michael Dummett’s vastly longer writing on Frege.) I don’t care if the concept horse isn’t a concept or Hansen’s Disease isn’t Hansen’s disease. And don’t even get me started on whether the present King of France wears a wig!
Comment by Randolph —
May 19, 2009 @ 9:27 pm
Seward, interesting. Delong and Krugman both care about economic history–I believe Delong has mentioned reading Smith at one time, and he has also mentioned Hayek. The people who got us into this huge mess, on the other hand, by and large took the attitude which Thoreau describes, that only the ideas mattered. Economics, we are relearning, is a historical science, and so history and historical sources must be studied. Delong and Krugman have been better able to respond to the on-going train wreck in the discipline than the pure Chicago-school economists.
Comment by Seward —
May 21, 2009 @ 2:03 pm
Randolph,
Hayek, Mises, Smith, Scumpeter, Bastiat, etc. were all mutlti-disciplinarians. So was Milton Friedman for that matter; his history of the federal reserve reads just like regular old history IMHO.