Philosophical Diamonds


“I don’t know any academic field whose writing regularly indulges in sentence structure as complex as in analytic philosophy.”

That’s Geoffrey K. Pullum, professor of linguistics at the University of Edinburgh, discussing the writing of analytic philosophers in a recent column at The Chronicle of Higher Education. He will be co-teaching a class on “philosophy of the language sciences” with philosopher Brian Rabern (Edinburgh) in the fall, and aims to “keep in mind that when we expect our students to read original work in the field of analytical philosophy, we are asking a lot.”

His main example of the complexity of analytic philosophy writing is a single sentence from Ruth Millikan’s Beyond Concepts: Unicepts, Language, and Natural Information:

In arguing for his analysis of non-natural meaning, Grice made the mistake of arguing from the sensible premise that a hearer who believed that a speaker did not intend by his words to produce in the hearer a certain belief or intention would not acquire that belief or intention to the invalid conclusion that a hearer who merely failed to believe that a speaker intended by his words to produce a certain belief or intention in the hearer also would not acquire that belief or intention.

Pullum’s point is not to insult this kind of writing, but to highlight how difficult it may be to read for students and others unfamiliar with analytic philosophy. He writes:

That is an 86-word sentence, so by the usual standards of readability it’s off the charts, even for high-school students. Yet it is perfectly formed; don’t imagine that I’m criticizing it. It’s just extraordinarily complex and demanding.

He then provides an exegesis of the sentence, noting that it is both “subtle” and “mind-crunchingly difficult.”

I sometimes think of such sentences as diamond-like: hard but clear, or, complex in their structure or vocabulary but with a meaning that’s transparent upon close inspection. I would imagine that many sentences in the writings of analytic philosophers are like this, especially for students.

I picked two books off the shelf and opened them at random. It took a few seconds in each to find such “philosophical diamonds”:

How can a sentence which comes as close as “Vixens are female foxes” does to being a definition of “vixen” be about vixens rather than about the word “vixen”? (Timothy Williamson, The Philosophy of Philosophy, p.49)

If a moral theory can be quite straightforwardly true, it is clear that, if it is self-effacing, this does not show that it cannot be true. (Derek Parfit, Reasons and Persons, p. 43)

Bonus points to Parfit for using both “straightforwardly” and “clear” in that sentence.

If you have your own examples of philosophical diamonds, let’s hear them. Noting their prevalence would be a good reminder of just how challenging it can be to put together and teach an effective course in analytic philosophy.

Alexis Arnold, “Complete Book of Crochet” (Crystallized Book Series)

guest
25 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
sahpa
sahpa
3 years ago

If I may ask a pedagogical question rather than give a diamond example…

I’m a (late) PhD student, and I’m beginning to design some sample syllabi for potential courses. It strikes me that, in an upper-division course (on philosophy of action, say), it might actually be good to assign a difficult text twice (something by Davidson, say). So, for example, it could be very useful to the students to read X, then read Y (which may be a response to X, or it might just explore some different aspects of the same phenomenon), and then re-read X in light of having read Y. The goal being, of course, to deepen their understanding of X.

Of course, practically speaking, I feel like the odds that anyone but the most ambitious and already-interested students will read X again are vanishingly small. That’s a shame because re-reading X in this way might actually cause them to become interested, now in a better position to appreciate its insights. Any tips on how to incentivize it?Report

Joe
Joe
Reply to  sahpa
3 years ago

Perhaps making your concern that most students won’t re-read X explicit would be helpful? And then alongside that you could do a reading check for each instance of reading X where the extent to which they test one’s comprehension increases. Report

Caligula's Goat
Caligula's Goat
Reply to  sahpa
3 years ago

I would imagine that a good lecture and discussion of X after Y would get you most of the work you want without the confusing double-reading. Intuitions are terrible gauges for reality, however, so I think that the best thing you could do is to actually teach a class in this way and to specifically ask your students for feedback on this at the end of the quarter (I wouldn’t announce this experiment until the end of the quarter when you’re soliciting feedback).

I’ve never done things your way by that’s merely anecdata. Report

Heath White
Heath White
3 years ago

To me, this highlights (clarifies!) the distinction between ‘clear’ meaning “not vague or ambiguous” and ‘clear’ meaning “not requiring a lot of processing power.”

If your goal is nailing down the truth, probably you are only interested in clarity-1. If your goal is effective communication, especially effective pedagogical communication, you need to be interested in clarity-2 as well.Report

Jonathan Weisberg
Jonathan Weisberg
3 years ago

That Millikan sentence *does* merit criticism. It’s utterly trivial to render the same passage readable. Just add a few periods, remove a few words, and italicize the central contrast. (Writing 101, in other words.)

====

In arguing for his analysis of non-natural meaning, Grice made a mistake. He argued from a sensible premise to an invalid conclusion. The sensible premise: a hearer who believes the speaker does not intend by his words to produce a certain belief or intention will not acquire that belief or intention. The invalid conclusion: a hearer who merely fails to believe the speaker so intends will also not acquire that belief or intention.

====

Maybe analytic philosophers do produce more diamonds because we’re conveying more complex ideas. But we could often achieve the very same clarity with prose that flows like water instead.Report

Tim
Tim
Reply to  Jonathan Weisberg
3 years ago

I agree with your assessment: the Millikan sentence is overly complex for no appreciable gain. I agree that we shouldn’t, as a discipline, allow this. I also think there’s a very clear way to accomplish this: when you’re refereeing, don’t accept papers rife with this kind of sentence. Reject (or recommend resubmission) on the grounds of excessively complex sentence structure.

It’s also not just about style points or pedagogy. It’s about making our field accessible. Yes, sentences like that are hard for undergraduates. They may also be impossible for anyone who isn’t a near-native English speaker. Allowing this kind of sentence to be normal amounts to walling our field off from all such people. Report

Animal Symbolicum
Animal Symbolicum
Reply to  Jonathan Weisberg
3 years ago

I think you’re exactly right.Report

John
John
3 years ago

My former advisor taught me the distinction between clarity and precision in writing (a distinction in the zone of what Heath White draws above). It’s one I try to convey to my students.

Milikan’s sentence is precise, that is (roughly), it says exactly what she means it to say and, when very carefully parsed, communicates that, leaving virtually no room for ambiguity. But her sentence is not clear: it is exceedingly and unnecessarily difficult to understand. Many graduate students tend toward precision at the price of clarity. The opposing tendency is less common among philosophers. Great philosophical writing (not necessarily great philosophy simipliciter) manages to wed clarity and precision, sacrificing neither. Doing that certainly takes more work than simply prioritizing one over the other, but my sense is that the long term benefits – for the philosopher and the discipline – are worth it.

We might also think here of a distinction between excellence in thinking and excellence in teaching or communicating (though, one could also parse the distinction under the heading of ‘teaching’). There is no deficiency of reasoning or thought in the sentence but I don’t think it succeeds fully as a piece teaching. Report

Dale E Miller
3 years ago

“I don’t know any academic field whose writing regularly indulges in sentence structure as complex as in analytic philosophy.”

Perhaps someday Pullum will discover Continental philosophy. I don’t mean that as a shot at Continental philosophy, and I admit that I’m not well versed in the subject myself, but I would have thought it uncontroversial that Continental thinkers’ syntax is at least as complex as that of analytic types. Report

Sam Duncan
Sam Duncan
3 years ago

If grammatical complexity and length of sentences make for philosophical diamonds, then Kant’s first Critique is studded with diamonds on every single page. That’s especially true of the original German. Sadly for our own language it just doesn’t seem set up to grow diamonds nearly as well as German. Perhaps then Heidegger is right that German is the only properly philosophical language among modern ones.
Joking aside I think that pretty much every single one of these is just an example of bad writing. I can easily think of ways to reword every single one of these passages that are clearer and at least as precise if not more so. I would never say that good writing has to be easy to read. I’ve a deep admiration for Kant and a grudging respect for Heidegger, so I can’t say that. Not to mention that outside philosophy I quite like Gene Wolfe. But I would say that good writing is necessarily intentional writing. By that I mean that ideally every aspect of it should serve a larger goal. That means that if it’s difficult or unclear there better be a good reason for that. In Being and Time, for instance, Heidegger uses bizarre and roundabout jargon precisely because he’s trying to draw our attention to aspects of experience whose familiarity makes us completely overlook them. I can’t think of similarly good reasons that any of these sentences need to be as hard to read as they are.Report

Jonathan Weisberg
Jonathan Weisberg
3 years ago

People sometimes point out that analytic philosophy is math-like in some ways. But we don’t have many of the conveniences mathematicians rely on to make their texts readable. Key philosophical concepts tend not to have universally accepted definitions, for example, and standard notations are few. We also discourage math-style writing more generally. Many readers count it against an author if they have lots of numbered definitions or named examples.

I wonder how many of our diamonds are a product of the need to do math-like work in ordinary prose. (Imagine a calculus textbook where every instance of “lim” is replaced by the epsilon-delta definition of convergence; every instance of d/dx is similarly replaced; and all formulas are written out in prose.)

I also wonder how math prose would compare to analytic philosophy’s, if some of its complexities weren’t suppressed with conveniences like numbered equations and definitions.Report

Animal Symbolicum
Animal Symbolicum
Reply to  Jonathan Weisberg
3 years ago

I think the reason we’re in this linguistic mess is that too many philosophers envision themselves as doing “math-like work in ordinary prose.” This vision is a delusion. You come close to identifying why it counts as a delusion: universally accepted or not, definitions in philosophical inquiry do not (perhaps cannot) play the same role as, or have the same theoretical or disciplinary significance as, definitions in mathematical inquiry. (Another, related source of the delusion is a misunderstanding of the semiotics of formal symbolism, a mistaken view of mathematical languages as “abbreviated natural languages” or “cleaned-up natural languages” or “just like natural languages only clearer and more precise.”)Report

Jack Woods
Jack Woods
Reply to  Jonathan Weisberg
3 years ago

Look at Paul Halmos (say, his Algebraic Logic book or Naive Set Theory) or, say, much of Alonzo Church’s or Dana Scott’s work. These mathematicians almost never use symbolizations unnecessarily, yet their prose is eminently readable and clear.

I think one of the major differences between (good) mathematics prose and (good) philosophical prose is that the reader is expected and trusted to keep up in reading much mathematics, so if it’s clear and terse and you don’t get it, it’s your fault, not the authors. Whereas in philosophy, many people don’t trust the reader to not misinterpret them (which, given some persistent disciplinary norms and the refereeing system, is kind of fair.)Report

Jonathan Weisberg
Jonathan Weisberg
Reply to  Jack Woods
3 years ago

Really good points about differences in trusting the reader, thanks.

As a reader, I think I’m more likely to distrust *the author* in philosophy than in math. And I guess, on reflection, I feel like that’s justified. Original philosophical work often seems to me to contain mistakes that render the work largely useless. (Yes, mine too.)

So maybe it’s like this: the non-mathematical aspects of what analytic philosophers do are easy to screw up. So readers learn to distrust authors. And authors in turn learn to distrust readers. To defend against being misunderstood then, they write terribly.Report

Mike Titelbaum
Mike Titelbaum
Reply to  Jonathan Weisberg
3 years ago

FWIW Jonathan, you don’t need to just imagine the various scenarios you mention above. Here’s a passage from state of the art algebra around 825 AD:
“Squares and roots are equal to numbers, as this example shows: A square and 10 roots equal 39. You investigate the technique by asking: ‘What is the square to which you join 10 of its roots so that the whole collection equals 39?’ The way of investigating this technique is to divide in half the number of roots. But there were 10 roots. So take 5 and multiply them by themselves to get 25 which you add to the 39 to obtain 64. Now the root of this number is 8. Take from it the half of the roots, 5, and three remain. Three therefore is the number of one root of this square which is necessarily nine. And so you do all problems of this type, except that if there are two or three squares or even less than one square, you convert them all to one square.”
In this (translated) passage, al-Khwarizmi uses a method we now call “completing the square” to solve the equation x^2+10x=39.Report

Jonathan Weisberg
Jonathan Weisberg
Reply to  Mike Titelbaum
3 years ago

Yikes! Reminds me of these old exams from St. Andrews:

https://standrewsrarebooks.wordpress.com/2017/12/11/the-exam-season-is-upon-us/#jp-carousel-12745

I’d definitely fail the philosophy ones.Report

Jack Woods
Jack Woods
3 years ago

That seems totally fair (and really depressing.)Report

Doc F Emeritus
Doc F Emeritus
3 years ago

I agree analytic philosophers produce complex sentences – my list includes Benson Mates, Sellers and others from back in my grad school days.

But put it in context: have you really read Derrida or Kant???? Derrida, especially is virtually impenetrable. As a former colleague said of him, “life is too short to work through this.”

The sad thing is that such impenetrable continental and analytic writers lend credence to the reputation philosophy has for being b.s. that only other philosophers can understand. I mention this a propos to the other posting here on recommended books of philosophy for non-philosophers. Without mentioning any named, there are some books and articles recommended for non-philosophers that I would say are challenging for grad students in philosophy. This might show that we philosophers are a bit out of touch with those outside our field.Report

Hermógenes Oliveira
Hermógenes Oliveira
3 years ago

Another fine example, from Dummett’s “Frege: Philosophy of Language” (p. 16):

“Moreover, those clauses in the inductive stipulation of the truth-conditions of open sentences which relate to the quantifiers are stated in terms of the truth-conditions of the open sentence to which the quantifier is prefixed which result from holding fixed the assignments to the free variables other than the one which becomes bound by the new quantifier, and allowing the assignment to the latter free variable to run through all the individuals in the domain.”

And that is from someone who fancied himself a master of English “Grammar and Style” (1997, Bristol Classical Press).Report

RJB
RJB
3 years ago

If anyone wants to improve their writing of complex ideas, my favorite book is “Joseph Williams’ “Style”.. I like it because it focuses on simple steps to improve your writing, rather than just telling you what it should look like at the end. Way way better than Strunk and White, which I know Pullum hates, and I say this as a Cornellian. The advice I’ve found most useful is to remember that every sentence tells a story, and thus has a character and an action. Make the character the subject and the action the verb. Otherwise you end up with subjects that are massive noun clauses that take forever to parse, and then wait in suspense for the verb, which ends up being “is” or something else uninformative. This wouldn’t quite fix Millikan, who embeds unnecessary complexity in subordinate noun clauses, but it helps surprisingly often.

On a more opinionated note, my field of accounting is keenly interested in reducing the costs of processing information. Anyone who imposes unnecessary processing costs on their readers fails one of the basic tenets of good accounting, and I would argue good philosophy as well. Report

Joel David Hamkins
3 years ago

To my of thinking, the concept of “philosophical diamonds” is a step on the road towards what I see as an unhealthy infatuation with difficult or technical ideas that one sometimes sees in philosophy. Most of us have seen instances where a weak philosophical argument was dressed up in pseudo-mathematical robes, seeking more respect than deserved and sometimes getting it, unfortunately, from those suffering under the infatuation. We should be on guard against this.

I propose that we celebrate instances of clear philosophical writing, which succeeds in explaining even a very abstract or difficult argument in plain language. The philosophical literature is full of this! But it may be sometimes difficult to recognize, because when the writer has truly succeeded in this task, then we don’t even notice, for the clarity of the writing makes the idea seem not so difficult or technical after all.Report

Steven French
Steven French
3 years ago

Might I suggest that just as with ‘good prose’, what counts as ‘clear’ or ‘precise’ philosophical writing might be a function of historical context (18C vs 21C), geography (Oxford vs well, pretty much anywhere thats not Oxford!) and so on. And sometimes, the ideas we’re trying to convey *are* difficult and complex and slippery … and may have to be limned in different ways.
I’m sympathetic to the idea of encouraging writers to express themselves as clearly as possible but I bristle at claims that what we’re trying to do really is simple or straightforward – often, perhaps most times, its not!Report

Colin McGinn
Colin McGinn
Reply to  Steven French
3 years ago

No, it’s not.Report

Eric
Eric
3 years ago

I write like the example sentence all the time but its not that hard to fix if you really want to:

In arguing for his analysis of non-natural meaning, Grice made a mistake. His mistake was to argue from one sensible premmise to an invalid conclusion. The sensible premise is:

1. Suppose a hearer believes that a speaker does not intend by his words to produce in the hearer a certain belief or intention. If so, then that hearer would not acquire that belief or intention.

The invalid conclusion is:

C. Suppose a hearer merely fails to believe* that a speaker intends by his words to produce a certain belief or intention in the hearer. That hearer would also not acquire that belief or intention.

*(as opposed to believing that not)

In fact, it was only after I did this that I was really able to parse what the sentence said. I don’t think philosophy needs 86 word sentences. It just causes people to write them, and we don’t always take the time to fix it.Report

Conor Mayo-Wilson
3 years ago

The fallacy: ~B(p) does not entail B(~p).

Sometimes symbolism doesn’t help, as Jack Woods says. But here one can render the fallacy vivid in a few symbols, and it’s much harder to read a corresponding English sentence like, “Not believing p does not entail that one believes p’s negation.” Millikan’s sentence is similar to algebraic equations and geometric problems that are described verbally; see Mike Titelbaum and Jonathan Weisberg’s posts above.

Report