Ted Bunn’s Blog

I’ll admit to being a bit of a grammar geek (in addition to several other kinds of geek), so I was interested in Geoffrey Pullum’s takedown of Strunk and White in the Chronicle of Higher Education.  I have positive impressions of Strunk and White from school, but I haven’t actually looked at it much in recent years.

Some of the review’s criticisms are kind of silly, I think:

Notice what I am objecting to is not the style advice in Elements, which might best be described the way The Hitchhiker’s Guide to the Galaxy describes Earth: mostly harmless. Some of the recommendations are vapid, like “Be clear” (how could one disagree?). Some are tautologous, like “Do not explain too much.” (Explaining too much means explaining more than you should, so of course you shouldn’t.) Many are useless, like “Omit needless words.” (The students who know which words are needless don’t need the instruction.) Even so, it doesn’t hurt to lay such well-meant maxims before novice writers.

I want to defend “Omit needless words.”  What it means is that you should constantly ask yourself as you’re writing (or better yet as you’re editing) whether you’re including unnecessary words or not.  It’s not just a matter of knowing which words are unnecessary; the more important point is that this is something you should pay attention to. This is actually one of the most useful of all S&W’s maxims, and one of the hardest to follow.

The really scathing part of the review is the stuff about grammar and usage (as opposed to style).  Pullum makes a convincing case that S&W are utterly incoherent on a lot of these points:

What concerns me is that the bias against the passive is being retailed by a pair of authors so grammatically clueless that they don’t know what is a passive construction and what isn’t. Of the four pairs of examples offered to show readers what to avoid and how to correct it, a staggering three out of the four are mistaken diagnoses. “At dawn the crowing of a rooster could be heard” is correctly identified as a passive clause, but the other three are all errors:

• “There were a great number of dead leaves lying on the ground” has no sign of the passive in it anywhere.
• “It was not long before she was very sorry that she had said what she had” also contains nothing that is even reminiscent of the passive construction.
• “The reason that he left college was that his health became impaired” is presumably fingered as passive because of “impaired,” but that’s a mistake. It’s an adjective here. “Become” doesn’t allow a following passive clause. (Notice, for example, that “A new edition became issued by the publishers” is not grammatical.

Pullum’s completely right on all this. He also has great examples of sentences from S&W that manage to simultaneously violate four or five of the authors’ own rules and yet sound just fine.  (Yes, I know that’s a split infinitive.  No, I don’t care.)

The stricture against the passive voice is a constant problem for scientists, because we’re also often taught to avoid writing in the first person when describing our procedures.  So should I say “The decay rate was measured” or “I measured the decay rate”?  Either way, someone will be mad at me. Once you realize this, it’s actually kind of liberating: since someone will be mad either way, just do as you please.

By the way, my favorite usage guide is Bryan Garner’s Dictionary of Modern American Usage, which I first learned about from a lengthy review by David Foster Wallace in Harper’s in 2001 (not linkable, as far I can tell).

The University of Richmond is currently considering various options for changing the required courses for first-year students.  For over a decade now, all first-year students have been required to take a two-semester core course, which is taught in many sections by many faculty members, but which has a common syllabus in which everyone studies, talks about, and writes about the same Important Texts.  Proposals have been made to change core, possibly replacing one or both semesters of it with first-year seminars on a wide variety of different topics, so that both instructors and students could choose to study topics of particular interest to them.

I’ve had a number of discussions on these options with my colleagues, including several members of the committee charged with developing them.  In at least one case, I think I utterly failed to make a colleague understand my point of view, and since he’s a very smart guy, I conclude that I explained myself poorly.  Here’s my attempt to do better.

I agree with the general proposition that it’s a good idea to require first-year students to take rigorous, writing-intensive courses in which they engage with big, important ideas.   These are all goals of core as presently constituted.  But there’s another aspect of core, namely that all students should simultaneously engage with the same big, important ideas — that is, that all of the 30+ core sections should read the same books.  I’m completely unconvinced of the merit of this, and I think that there are big disadvantages associated with it.  Something like the freshman-seminar model, in which students in different seminars study different things, seems to me better in virtually every way.

I’ll get into detailed arguments below.  First, though, I want to make one thing very clear: I don’t think that core as it exists is a bad thing, just that it’s not the best thing we could be doing with the resources at hand.  If we were talking about replacing core with nothing, I think I’d be opposed to that replacement.  But I think we can and should replace core with something better, and I think that a set of first-year seminars would be such a thing.

To me, the main disadvantage of core is that few faculty members have expertise in all or most of the various areas studied.

I don’t think that most proponents of core dispute this fact, but many deny its importance.  The argument goes that the point of core is to “engage with” the texts, and to use them to get practice thinking about big, difficult ideas.  The instructor is not supposed to be an expert but rather a facilitator in this process of intellectual maturing, so expertise doesn’t matter.

I don’t buy this.  Presumably the texts in core were chosen because the ideas in them are actually important and worth understanding.  Those ideas are in many cases also difficult to understand.  Someone who’s spent a lot of time studying Nietzsche, or Plato, or Darwin, or many of the other core texts, will be better able to facilitate students’ understanding of these difficult ideas than someone who hasn’t.  In each case, there are fruitful modes of thought for engaging with the ideas, and other modes of thought that are not fruitful.  A first-year student engaging with these texts needs a guide who can steer them in the fruitful directions.  That is, they need an instructor who’s got expertise.

If I put in a huge amount of effort over a long time, maybe I could reach the point where I could do a barely adequate job guiding a student through Nietzsche.  But with the same amount of effort (or less) I could do a great job guiding the student through Galileo.   Which is a better use of faculty resources?  Which gives the student a better experience?

(Incidentally, people have told me that I’m selling myself short in the above statement.  I honestly don’t think I am.  I have many flaws, but a low opinion of my own intellect is not one of them.  I think I’m a pretty smart guy.  I just don’t think that being smart is enough to make up for a lifetime of not studying something.)

A related issue is that core as presently taught is largely housed in the humanities, with relatively little participation from other parts of the university.  This year, roughly half of the core instructors are from departments having to do with languages and literatures; if you combine those with history, philosophy, and the arts, you get about 3/4 of the instructors coming from the humanities broadly construed.  There’s one instructor from mathematics, one from leadership, and none from the natural sciences or business.  Maybe that’s OK, but I think it’d be better if the first-year core courses were spread out more broadly.  I don’t know for sure that that’d happen with a first-year-seminar model, but the odds have got to be better.

Now let me discuss a few of the arguments I’ve heard in favor of core:

1. The intellectual climate of the student body as a whole is enhanced, because all students campuswide can discuss the same body of work.

In principle, I guess that’s possible.  I’d like to see some evidence that it actually makes a significant difference. Do first-year students actually discuss the core texts with others who are not in their core class?  Does that add to the intellectual climate more than the alternative, which is students having a bunch of different intellectually rigorous experiences that they can discuss with their friends?  Personally, I doubt it.  I think that exciting, rigorous, demanding course work has the potential to improve the intellectual climate, but I’m not convinced that there’s significant value added in that course work being uniform across campus.

If you have actual data to suggest otherwise, please show it to me.  (If you have anecdotes, on the other hand, please don’t.  I’ll make a deal with you: I won’t mention my anecdata if you don’t mention yours.)

2.  The particular ideas in these particular texts are so important that all students must read them in order to be considered educated.

In fairness, I’ve never encountered this argument firsthand; I’ve just heard it by hearsay.  So maybe nobody really believes this.

Anyway, the problem with this argument is that there are literally hundreds of texts as important as the ones on the core syllabus.  If  a student can’t consider herself well-educated without reading, say, House of Mirth, then surely she can’t consider herself well-educated without reading Adam Smith, Galileo, Dante, Dickens, Mary Wollstonecraft, Lao-Tzu, Hume, Einstein, and so forth.  (By the way, I don’t mean to pick on House of Mirth — it’s one of my favorite novels.)  The scholarly world is full of big, important ideas.  We can have a system in which students and faculty can choose from a broad range of these ideas, while still guaranteeing that everyone is grappling with big, important ideas.

3. The fact that all core sections study the same texts acts as a sort of “quality control”: in a seminar system, it’d be harder to assure that all students were getting an equally rigorous experience.

We can if we choose impose uniformity of expectations on seminars.  We can mandate a certain amount of writing, and we can have a faculty committee vet the syllabi to decide if the topics and readings are hard and significant enough.  If we do that, I don’t see that the quality-control issues are any worse than with core as it exists now.  Once again, if there’s any non-anecdotal evidence about the degree of uniformity of core expectations, I’d be interested to hear about it.  (It’s taking all my self-control to abide by my earlier promise to keep my own anecdotal evidence to myself, by the way.)

The issue of quality control comes up in other places as well, of course.  Take the general-education requirements, for example.  In order for a course to be designated as meeting one of the field of study requirements, its course description must be approved by some faculty body, and then after that we trust our faculty colleagues to behave professionally and do what they’ve promised to do.  I don’t see why the quality-control issues are significantly different for a first-year seminar program.

To summarize, I’m strongly in favor of a demanding, writing-intensive, first-year experience for students, in which they engage with difficult, big ideas.  But I think we should choose a model for that experience in which students have a wide variety of different big ideas that they can choose from, rather than all students having to study the same thing.

Here are two things I thought were pretty funny:

1. A former student sent me this spoof article showing that the value of pi has changed over time.  Although it’s mostly making fun of the recent (and scientifically legitimate, by the way) studies of whether fundamental physical constants vary over time, I think I detect a hint of mockery of the old Creationist claim that the speed of light has changed over time.  (The claim used to be that the speed tended to infinity at, conveniently enough, about 6000 years ago.)

2. From Andrew Jaffe: This erratum was published in the Astrophysical Journal:

As a result of an error at the Publisher, the term "frequentist" was erroneously changed to "most frequent" throughout the article. IOP Publishing sincerely regrets this error.

You probably have to be someone like me who uses the word “frequentist” in casual conversation to find this funny.

I’ve written about this paper a bunch of times, but since it’s now officially accepted and won’t be changing any more, I thought I’d post a link to the final version on the arxiv. (There’s a link in the upper right to download the PDF.)

The 2009 Goldwater Scholarships were announced today.  I’m pleased to announce that two University of Richmond students were recognized: Matt Der was awarded a scholarship, and Anna Parker received an honorable mention.  In both cases, the honor is extremely well-deserved.  If you know these two students, please make sure to congratulate them.

although it is correlated with it.

(I saw this on the Freakonomics blog.)

I generally think of the Economist as having excellent science reporting, but I have to admit that I only read it sporadically these days, so maybe my information is out of date.  This article on some recent experiments in the foundations of quantum mechanics is way, way below the standard I’d expect from them.

The article claims that

Now two groups of physicists, working independently, have demonstrated that nature is indeed real when unobserved.

Your skepticism level should spike whenever you read an article claiming that experimental results have  confirmed some deep-sounding-yet-vague philosophical idea.

When the article gets down to details, it says a bunch of things that are just wrong:

In the 1990s a physicist called Lucien Hardy proposed a thought experiment that makes nonsense of the famous interaction between matter and antimatter€”that when a particle meets its antiparticle, the pair always annihilate one another in a burst of energy. Dr Hardy's scheme left open the possibility that in some cases when their interaction is not observed a particle and an antiparticle could interact with one another and survive.

This makes no sense.  It’s quite common for a particle and its antiparticle to interact without annihilating each other.  There’s nothing paradoxical about that.

Later on, we have

The stunning result, though, was that in some places the number of photons was actually less than zero. Fewer than zero particles being present usually means that you have antiparticles instead. But there is no such thing as an antiphoton (photons are their own antiparticles, and are pure energy in any case), so that cannot apply here.

Where to begin?  First, I don’t believe the statement that the researchers measured “less than zero” (or even “fewer than zero”) photons.  Did they have a photomultiplier tube that clicked -17 times? The actual articles by the two experimental groups contain no claims of anything that could be described in this way.

Second, the idea that fewer than zero particles means antiparticles isn’t really true.  There was a theory due to Dirac a long time ago in which positrons (the antiparticles of electrons) were regarded as “holes” that were sort of like negative numbers of electrons, but that theory doesn’t really fit into modern particle physics in any nice way as far as I know.

Finally, can we please ban the phrase “photons are pure energy”?  There’s no meaningful sense in which that’s true.

Now I know that science journalists have a tough job, particularly when reporting on very abstract things such as quantum physics.  I have no objection at all to their simplifying things to make them clearer.  But the above statements aren’t simplifications; they’re just wrong.

These experimental results seem to be a confirmation of a “paradox” in which joint probabilities of two quantities don’t obey the rules that you’d expect classically. Like other similar paradoxes, it’s a nice example of the ways in which quantum mechanics is “spooky,” but I don’t see how the results say anything new about whether reality exists when you’re not looking (whatever that even means).

There’s a common problem in journalism about the foundations of quantum mechanics.  As my friend John Baez said a long time ago,

Newspapers tend to report the results of these fundamentals-of-QM experiments as if they were shocking and inexplicable threats to known physics. It’s gotten to the point where whenever I read the introduction to one of these articles, I close my eyes and predict what the reported results will be, and I’m always right. The results are always consistent with simply taking quantum mechanics at face value.

The Economist article goes pretty far (if not all the way) down this route.  It takes a clever and interesting experimental result and blows it up into something much spookier than it is. The result does follow precisely the pattern Baez describes: it’s exactly what standard quantum mechanics predicted all along.

Please note that I’m not criticizing the experimenters here: as far as I can tell, they’ve done a difficult and very interesting experiment and described their results accurately.  I don’t see anything in their papers that would justify the silly interpretation placed on it by the Economist.

By the way, I got the Baez quote above from a web page by Matt McIrvin which quotes Baez.  I can’t find this quote on Baez’s own web site, not that I looked all that hard.  Matt has great things to say about the interpretation of quantum mechanics, by the way.  If you’re at all interested in this subject, please read this page in addion to — or better yet instead of — the many silly things that have been written on the subject.  John Baez has said a bunch of really smart things on the subject in various newsgroup posts over the years as well.  He collected some of them here, but as far as I know he hasn’t written up anything more polished.

 My friend Walter makes the following very sensible point about my recent article on evolution and the second law of thermodynamics:

The paper about thermodynamics and evolution is neat. Though of course it will not convince any creationists as their argument was never truly science based anyway.

This is certainly true.  I don’t harbor any hope that even a single creationist will read the article and see the error of his ways.   So then who is this article written for, and what’s the point of writing it? There are actually a few answers.

1. The people who are most likely to read the article are people with an interest in the teaching of physics.  It’ll appear in the American Journal of Physics, which is kind of a funny journal: it publishes some articles explicitly about physics education research (comparing different instructional techniques, etc.), but its main emphasis is articles about physics, written in a way that are of interest to physics educators. The main reason I wrote the article is because I think that it raises some interesting physics points that educators may find useful.  In particular, I like the fact that you can use back-of-the-envelope estimates using standard undergraduate statistical mechanics techniques to get an answer to this interesting question.  I wish I’d worked all this out before the last time I taught statistical mechanics: I would have liked to use it in my class.  I expect I’ll get another chance.  Anyway, I hope that other statistical mechanics instructors find it useful.

2. My article is a followup to an earlier article, which I really liked but which had an unfortunate error in it. I wanted to set the record straight.

3.  Despite Walter’s correct assertion that I’m not going to win any creationist hearts and minds with this article, I do hope it fits in in a tiny way to the ongoing arguments on this subject.  Walter’s quite right that fundamentally the creationists’ argument is non-scientific, but on the other hand creationists do frequently try to make scientific (or scientific-seeming) arguments in support of their beliefs.  In particular, the claim that there is a conflict between evolution and the second law is a scientific claim, and the right way to refute it is scientifically.

When creationists make scientific claims, they like to use the traditional signs of scientific authority to bolster these claims.  For instance, they prominently refer to the academic degrees and positions held by their advocates.  Moreover, they use the peer-reviewed literature, and prominently note that they’re doing so.  I’m not criticizing them for doing this: peer review (for all its flaws) is the main way that scientific quality is evaluated.  But this does suggest that, when a claim is made in the creationism/evolution debates that has the form of a scientific argument but is scientifically incorrect, it’s worthwhile to have an authoritative, peer-reviewed refutation of it. When a creationist tries to use the authority of the peer-reviewed literature to make an incorrect point, a scientist can fight fire with fire.

Another tactic used by creationists (and other people who reject well-established science) is to claim that scientists are unwilling to even look at arguments that go against the standard orthodoxy.  That’s another reason it’s worthwhile to examine the creationist argument and take it seriously enough to show why it’s wrong.

The intended audience is not the already-convinced creationist, of course, but the innocent bystander who may not know much about the subject and who hasn’t made up his mind.  I don’t expect such a person to read the article (although if they’ve had enough undergraduate-level physics, it’d be great if they did); I want scientists and science teachers who are talking to such people to be able to say, “Yes, scientists have looked carefully at this argument, taken it seriously, and shown why it’s incorrect.”