Steve Reads

Approximately N+1 people have recommended The Omnivore’s Dilemma to me. From all the descriptions I’ve heard, Pollan continues a thread that he started in an article for the New York Times Magazine some years ago, that any number of Google searches couldn’t dig up. The premise of the article was that most American meat is a thinly disguised version of petroleum, when you count the pesticides applied to the corn that the cows eat, and the gasoline that moved the meat to your supermarket, and the antibiotics fed to the cows to treat the sores in their stomachs — which got there because cows’ stomachs have evolved to eat grass, not corn.

So I’ve finally started reading Omnivore’s Dilemma. It promises to be just as good as Pollan’s writing usually is. Pollan is an exceptional science writer. If Omnivore’s Dilemma delivers, as I’m sure it will, I may hit up The Botany of Desire next. Though I’m inclined to read some more-detailed economics, instead: Economic Behavior and Institutions is somewhere in the queue. And Akerlof’s Explorations in Pragmatic Economics has caught my eye.

Like Tyler Cowen, I found a good chunk of NOMYSAWM oversimple. A lot of it could be rebutted with “Well, you have your story and I have mine,” and both those stories would be equally valid until we had some models and some numbers to look at. If you’re looking for the models and the numbers, and you’re of a quantitative bent, check Bowles’s Microeconomics. (Slee cited Hal Varian’s Intermediate Microeconomics along the way, so I put in a request for that at the library.)

In particular, Slee argues at one point that the market for lemons guarantees that jobs will tend to be dominated by those who are already part of the old boys’ network: since corporations can’t be sure of the quality of incoming applicants, they’ll tend to rely on easier-to-acquire measures of the applicant’s quality — such as that the CEO knows him. So the same people will always end up with the jobs, hence on top. In this way, Slee argues, rational choice leads directly to persistent inequality.

That’s one story. Another story might run like so: those who grew up poor (for instance) are much less knowledgeable about their own value than are those who grew up in privilege. Hence they’re likely to do work for cheaper than those with more knowledge. So if the cost of searching among low-income applicants is low enough, it can be worth a company’s while to engage in that search and thereby undercut their opponents. Over time, firms may well form which have a reputation for spotting diamonds in the rough. The free market rides in and saves the day once more.

Those are just stories, with more or less intuitive plausibility. As Henry Farrell put it recently, “$4.50 and a logically consistent story that accords with intuition will get you a venti latte at Starbucks.” We need data.

Slee’s job isn’t to provide that data. His job is to break through the fog of our existing cultural intuitions — such as the classic idea of “choice.” And in that he succeeds admirably.

My man Jeff Beene recommended in the strongest terms that I check out The Secret Parts of Fortune, so I picked it up from the library. I don’t think one is necessarily supposed to read it straight through, so I’ve been ingesting an essay here and there while reading other books.

Rosenbaum plays a particular character. He’s irascible, well-read, pretty convinced of his own intelligence, and funny. Imagine David Foster Wallace as a crusading journalist and you’re almost there.

The Secret Parts of Fortune is good bathroom reading. In my experience with it so far, it never gets very deeply into any one subject. Though at times it applies a good, quick hatchet where a good, quick hatchet needs to be applied — like when Rosenbaum takes down Elisabeth Kübler-Ross. It needed to be said, and Rosenbaum said it.

I’m looking forward to reading other essays that Jeff has raved about, like the one where Rosenbaum hunts down the reclusive J.D. Salinger and finds that the entire town where he lives conspires to keep him anonymous.

I suspect that Rosenbaum’s essays are more about Rosenbaum than about his subjects. But his subjects are interesting enough that they can’t help but shine through. And Rosenbaum’s not a bad character to follow, in any case.

P.S.: Last night I read the piece where he decides to figure out what Skull and Bones’s actual secrets are. It’s a great piece, and quite funny.

Having dropped North’s book, I immediately picked up Tom Slee’s No One Makes You Shop At Wal-Mart: The Surprising Deceptions of Individual Choice. The title is actually ironic: Slee’s claim is that the choices of other consumers often force you to do things you don’t want to do. A “choice” shouldn’t be viewed as an atomic economic act; real choices are entangled with what everyone else chooses to do. You choose to buy Us Weekly at the grocery store, and so do a million of your compatriots, and pretty soon that’s all that’s available in the checkout aisle. No one “chose” for that to happen, but that’s how it worked out. Choices are constrained by other choices.

Slee wraps this all up beautifully: you should think of “best response” rather than “preference”: what you choose to do is not a direct expression of what you prefer, as naïve choice theory would have it; rather, what you choose is the best response to everyone else’s choices — and theirs are best responses to yours.

The classic example of a best response that leads to a disappointing outcome is the prisoner’s dilemma: each prisoner, when deciding how to act, realizes that no matter what the other prisoner does, it would be in his best interest to rat his partner out. If he rats the other one out, and the other one rats him out, then he’s better off than if his partner ratted and he stayed quiet. Likewise, if his partner stays quiet, he’s better off ratting his partner out. So no matter what his partner does, he should pick the outcome that makes life worse for the both of them.

Slee’s book is — so far, a bit less than halfway through it — the best use of economics in a mass-audience context that I’ve yet seen. And it’s entirely rigorous. The argument is perfectly simple and correct. It should be valuable to anyone who believes that the free market will apply a balm to all woes.

P.S.: Tyler Cowen and Cosma Shalizi have reviews.

As a fan of the economic focus on institutions of which Samuel Bowles is a part, I was excited to read one of Douglass North’s books. North is the founder of this school, it would seem, and won a Nobel for it.

Unfortunately, the above-named book hasn’t enough mathematics to satisfy those with a quantitative bent, nor enough elegance to captivate anyone. I had to put it down after 80 pages (of 140 or so). I haven’t felt like precious moments of my life were being lost so viciously since I last hurled Cryptonomicon across the room.

Is economics yet competent to make predictions of how much college costs will rise in the next, say, 10 years? The best I’ve seen — and I admit I’ve not looked far — is the supposition that the rise in costs must be rational, or else people wouldn’t pay that much. Here “rational” means “justified by the increase in lifetime earnings.” Wouldn’t it follow, then, that if two universities yield identical expected lifetime earnings, it’s irrational to attend the more expensive one? Assuming, of course, that you have accurate information about the lifetime earnings from both.

There are studies looking at lifetime earnings from the class of 1972, comparing those students who were accepted to elite schools but decided not to attend with students who attended the elite schools, but that’s the class of 1972. It’s not clear how much that has to say about students in the class of 2012. So there may be an information deficiency.

There are shorter-term estimates of earnings, in devices like the U.S. News and World Report college guide; USN&WR publishes average earnings for those who are a year out of college, if I recall correctly. If the folk wisdom is correct, though, the prestige of the college you attended only affects your earnings in your first job. Everything thereafter is based on your work experience.

Then I have two questions:

1. Will there always be too little information for students to make rational decisions about which college to attend?
2. Does economics have anything to contribute here, in the way of actual predictions?

It must be awfully hard to interview Philip Roth, as I’ve mentioned before. First of all, Roth pours great scorn on the idea of interpreting the author into his books or his characters; that scorn leaks out of practically every page in Exit Ghost. So if you go into the interview with that in mind, you’re going to have to keep to the text of the book. Which is fine: that’s a discipline that more reviewers should have, John Updike among them. Updike’s review of Ann Patchett’s new novel in this week’s New Yorker manages to suck all the life out of Patchett’s earlier novel, Bel Canto, by psychologizing about Patchett when he ought to keep his head in the story. What makes Bel Canto magical is that during the time spent reading it, the rest of the world disappears; there is no world outside that novel. Patchett planned it that way. The characters are tucked off in a house that’s mostly disconnected from the world, in a nameless South American country, and they themselves forget that there is another world outside their doors. The great beauty of Bel Canto is that you and the characters are equally enveloped. (The fact that Updike can’t grasp this may explain why I’ve never been able to read more than a few pages of his writing.)

I mention all of this because Hermione Lee interviews Roth this week. He is bellicose as always about the role of the novelist. He’s adamant that the author’s role is to dig for the particulars of characters, not to deliver sermons; he’s been making that point at least since I Married A Communist.

Roth’s latest is filled with bile for the juvenile way in which critics interpret novels as one-to-one renderings of the authors’ real lives; one of his characters pens a long letter to the New York Times in which she claims that the world would be a better place if all the literature classes of the world disbanded, and all the book reviewers went home. Leave readers alone with their books, she says. She’s particularly incensed that the Times ventured off into the hinterlands of Michigan to interview those who may have inspired Ernest Hemingway’s early short stories. (She may, in fact, be referring to this article.)

So if I were interviewing Roth, I’d try my best not to draw any connections between the world of the books and the world of the world, and try to dive as intensely into his characters as he himself has: to really seek to understand the books, rather than understand Roth. I realize this may be overly limiting, but it would be a refreshing change of pace: I invite anyone to find me an interview with Roth that manages to stick entirely to the details of the books. He obviously encourages us to conflate the man and the character, so I have to blame him. At the same time, this should direct interviewers to ask him about particular moments in Zuckerman’s life as an author — rather than particular moments in Roth’s life that would have led him to mold Zuckerman in a particular way.

I’m on the bus down Route 1 right now into Boston. Per usual for this time of day, it’s a parking lot. This just serves no one well. I can’t imagine that people driving to Boston prefer to sit in traffic like this. If there were a massive train station where route 1 meets Interstate 95, or even a massive bus station, with clean transportation that picked up often — say every 5 minutes between 7 a.m. and 9 a.m. — I’m certain that it would get huge ridership.

The trouble there is that people would still have to drive their cars to that intersection, park and get on the bus. So they’d still have to own a car. Long term, the only way to get people out of their cars is to build lots of housing near mass-transit hubs. If there were a big transportation hub right at that intersection, the housing would probably follow. An intelligent transportation policy would incorporate mixed-use development right around that hub from the start.

What about on the other end? Not all of these drivers are heading into Boston. Some are going into Cambridge, some to Somerville, some to Arlington, etc. Using mass transit would involve quickly getting them to their destinations in and around the city. Seems to me that the state would have to commit to a mass-transit policy that loses a lot of money in the first few years: set up buses to leave frequently to take people from South Station to the 100 most popular destinations, even if those buses are normally empty. There would have to be a sacrifice in the initial period just to get people to ride buses in the first place.

Again, it would have been smart for the state to have set aside land for train tracks in the rights of way alongside I-95 and Route 1, but they didn’t do that and now it’s not an option for them to do it: buildings have crept right up to the side of the road. Condemning them for train tracks would be prohibitively expensive. (The Golden Banana would be out of the state’s price range.)

Seems to me that the smartest intermittent step would be a massive investment in buses. Buses are a smart transitional form of transportation between cars and trains. Right now I’m on a bus with 40 other people; assuming the average car has even 2 people in it (which I’m sure is an overestimate), this bus has taken twenty cars off the road.

I am, of course, no more informed than anyone else who travels to Boston, so all of this is speculation. I’d love to hear from someone who has actual experience with transit planning. Per usual, I doubt there are any technical problems here — only political ones.

 . . . may well do for some contemporary novelists what Mark Twain’s essay “Fenimore Cooper’s Literary Offenses” did for Fenimore Cooper — namely, permanently shut the door on his reputation. The “Manifesto” is certainly making me rethink my love of Paul Auster. Though actually, on reflection, I don’t think anything will weaken the horrifying perfection that is In The Country of Last Things.

(Cached: Twain essay, “Reader’s Manifesto”.)

There is a new version of C++ on the horizon. There are those of you here who will find this interesting. I bequeath this gift of a link unto you.

Sweet, sweet prognostication: Joel Spolsky (“Joel on Software”) has a new piece on where the web is going: specifically, how the history of the computer industry up to now will repeat itself in the development of Ajax apps. It’s very smart, and seems to me perfectly sensible.

For some reason I used to think Spolsky was a wanker. Now I’m confused why I thought that. He labeled Groove (my former employer) “architecture astronauts”, which turned out to be perfectly correct (at the time; they learned their lesson). His standard for whom to hire — “smart, and gets things done” seems to me quite correct. (I should hunt around to see what he has to say about technology management. My colleague Martin Martin has some excellent thoughts on the matter.) Spolsky’s piece on “The Absolute Minimum Every Software Developer Absolutely, Positively Must Know About Unicode and Character Sets (No Excuses!)” is educational, lucid and fun.

So I don’t know what my problem was. Spolsky seems to rock pretty hard. Maybe I should check out his book.

P.S.: The Ajaxian blog looks on Spolsky’s piece with bemusement.

Reading over Mark-Jason Dominus’s excellent discussion of the Russell/Whitehead proof that 1+1=2, it occurs to me, not for the first time, that a lot of historical scientific and mathematical works would be easier to read if their idioms were translated into contemporary ones. At the very least, updating Russell’s and Whitehead’s dot notation (using dots where we’d use parentheses to group subexpressions) would make things easier. Dominus gives a few other one-to-one translations:

The symbol has not changed; it means that the formula to which it applies is asserted to be true. ⊃ is logical implication, and is logical equivalence. Λ is the empty set, which we write nowadays as ∅. ∩ ∪, and ∈ have their modern meanings: ∩ and ∪ are the set intersection and the union operators, and x∈y means that x is an element of set y;

Those are the least arguable translations from a work like the Principia. More arguable would be restating Russell and Whitehead’s results using the modern notion of the ordered pair: would doing so erase what makes the Principia interesting in the first place?

I can see arguments either way. For one thing, surely it depends on which ideas we’re translating into modern ones. For another, it depends on the use to which we’re putting the translation. If the point is to understand the difficulties that Russell and Whitehead were trying to overcome, then the whole point is to use their notation and their ideas. If we’re using Russell and Whitehead’s results in proofs today, though, then translating might be a good idea. Does anyone use Russell and Whitehead today? If you wanted to cite a proof that 1+1=2, you’d probably cite a more-recent axiomatic-set-theory book, right?

(If anyone has access to a database of citations, it would be interesting to check whether any mathematics journals in the last 30 years or so have cited Russell and Whitehead for non-historical reasons.)

Oh, gnumeric. I love you. I love how minimalist you are, and how clean your UI is, and how remarkably polished you are for being such a third-wheel spreadsheet. And yet.

So let me ask you, gnumeric. Where should the Preferences option go? Would you put it under the Tools menu?

No. Hm. Maybe under the Edit menu?

No? Really? So  . . .  well, I guess I’ll just scan through the menus. Ah, right: the Format menu.

Even here, gnumeric: everyone else puts Preferences at the bottom of the menu. Why not you? And why in the Format menu, of all places?

Do you want to scorn my love, gnumeric? I’ve been so good to you, and you just keep pushing me away.

Apropos of a TPM post on the Clinton-Obama race for the Democratic nomination:

1. Isn’t it incredibly silly to even think about polling this early?
2. Given that only a handful of states actually matter in the general election, because of our broken electoral system, shouldn’t polls only focus on probable swing states from early on? I concede the possibility that, like Virginia, Texas may one day become more liberal. But in the meantime, it seems as though our polls should reflect the sad reality of our electoral process and just assume that Texas is solidly red. Maybe they already do this, come to think of it: given that the polling organizations are for-profit enterprises, I’m sure they look to cut costs where possible.
3. Who cares about Iowa or New Hampshire? Have they ever predicted the outcome of any election?

P.S.: For that matter, I wonder if polls at any point in an election are connected with the study Adam Rosi-Kessel mentioned, on what effect a music’s current popularity has on its future popularity. If polls tell us that a given candidate is (currently) “unelectable,” does that encourage others to vote against him? Seems to me that the answer is “certainly yes,” though it may only apply early in the election season: someone who gets his party’s nomination may be presumptively electable.

The Wikipedia entry for pleonasm is 4,113 words long. Just because this is what I do, I wrote a script to count the number of words in a Wikipedia entry, with some of the unimportant stuff (like the autogenerated table of contents) deleted.

I suspect there will be two people in the world who care about that. I am both of them.

Per James’s recommendation, I picked up Google’s PageRank and Beyond: the Science of Search Engine Rankings. Two brief initial notes:

1. It’s more accurate to say that it’s a book about the mathematics of search-engine rankings than about the “science.” I could think up a good way to define the difference, but I’m too tired right now. They’re tied together, of course, but this book has way more in the way of theorems than empirical results. Though it does cover a lot of computer-science issues in how to do PageRank at large scale, so it’s science in that sense. Really, I suspect that “the mathematics of search-engine rankings” wouldn’t sell nearly as well as what they chose.

2. James describes the book as “moderately rigorous.” The only way it could be more rigorous, to my mind, is if it were structured as definition-theorem-proof. As it is, it has full-frontal definitions and theorems tucked into the back, and also sprinkled throughout the text. The book assumes you’re up on your linear algebra, large chunks of which I’ve sadly forgotten.

I’m only skimming it, for the moment, but it’s fun times. The basic outline of how PageRank works is nicely summarized in another context on Cosma Shalizi’s blog, so I’ll leave the honors to him. What’s interesting is that a technology developed initially for measuring the reputations of academic journals should migrate so nicely to the web. Seems like Google’s problem is largely one of scale: while there may be tens of thousands, or even hundreds of thousands of academic journals, there are billions of web pages. So the transition-probability matrix contains billions of rows and just as many columns.

In Google’s case, its enormous matrix contains probabilities that the user will click over from one page to another. Specifically, we start by assuming a random user who visits a web page, randomly chooses amongst the n links on that page, clicks on the one he chose, and repeats. Since most web pages have only a few links (PageRank and Beyond says an average of ten links per page), this is an extremely sparse matrix: in a row with billions of columns, all but about 10 will be zeroes. Sparse matrices will be easier to manage in some cases, whereas dense matrices will be desirable in others. Likewise, I have to imagine that the web’s cliquish structure — American-history web pages don’t tend to link to astrophysics ones, and neither links to fetish porn — helps with optimization; maybe it means that the larger matrix of the whole web can be broken down into much smaller submatrices.

As you can see, my knowledge of the details here is pretty sketchy. If I gave PageRank and Beyond the time it deserved, all of this would be clear; the authors are very good about tightly integrating the mathematics and the applications. In fact, I could see this book being turned into an excellent introductory linear-algebra class for computer scientists. They could write up their own little version of Google and write code to compute the eigenvectors in practical amounts of time. If Chris Genovese taught linear algebra, I have no doubt that his class would be that cool.

I recommend PageRank and Beyond for anyone with a decent mathematics background and some interest in how Your Favorite Online Search Behemoth works.

Aaron Swartz has an excellent post up about the pesticide-industry-funded (and Republican-backed) claim that Rachel Carson’s war on DDT cost the lives of millions of Africans. Swartz’s answer: the rise in African malaria cases has nothing to do with a DDT ban and everything to do with mosquitoes developing resistance to the chemical.

Swartz’s post included below the fold.

(Hat tip to Crooked Timber for leading me to Aaron’s post.)

(more…)

I’m happy to see that the Burrito Guy near 15th and K NW in D.C. — whom I’ve recommended before — is still drawing long lines at lunchtime. Go, Burrito Guy. Keep living the Burrito Dream for us sinners.

(Photo courtesy of my friend Andrew Pratt, whom I miss.)

I’ve owned a gorgeous box set of The Iliad and The Odyssey since 1996 or so. I only read The Iliad in maybe 2002 or 2003, and only just finished The Odyssey on the 7th of this month. I can’t tell you how much of a joy it is to put a book on the to-read shelf that has been taunting me for more than a decade.

(I should apologize up front to Chris Rugen. I promised that I’d read The Odyssey along with him, and I totally didn’t pace myself to do so.)

It’s a good, fun story; in fact it’s just about the original good, fun story. I doubt I could really add anything to the plot beyond what everyone already knows. There’s the famous story of the Sirens, and Odysseus lashing himself to the mast to stop himself from falling prey to their wiles. There’s the Cyclops, whose one eye Ulysses gouges out. There are the monsters at the Scylla and Charybdis. And so forth. He’s been away at war against Troy for ten years. After Troy falls at the end of the Iliad, Ulysses and others sack the city and take to the seas to return home. It takes them ten years.

Meanwhile, after so long away from home with no word back to his wife, the locals back on Ithaca (known as “the suitors”) believe Ulysses is dead, take up residence in his house, drink his wine, slaughter his animals, bed his maids, and try to woo his wife. She — Penelope — is crafty, and tells them that she’ll pick a suitor once she’s done sewing a burial shroud for Odysseus’s father (who’s not dead, but who’s been devastated by the loss of his son and expects to die soon). Every day she weaves it, and every night she undoes what she made during the day. Eventually one of the maids lets the suitors in on the secret, so they force her to finish her sewing. She’s just on the verge of choosing a suitor when Odysseus returns home, disguised as a beggar by the goddess Athena. He wanders around his property, begging for a crust of bread and judging the character of all those who are living in his house. His son, Telemachus, is brave and handsome and everything else you’d expect out of the son of a Greek hero. Penelope has remained true to her husband for twenty years. Various swineherds weep for their departed master, still, after twenty years. All the rest of them are bastards.

After setting up appropriate dramatic tension really effectively, and leaving Odysseus within just a few feet of his family — though disguised — Homer finally lets loose in a cataclysm of blood. All the bastards die.

I love the story because it’s so elemental. Throughout The Iliad and The Odyssey, “staid servantwomen” serve hundreds of meals wherein goats and sheep are cut into four pieces and roasted over spits. After eating the meat and drinking the hearty wine, various heroes either fall into peaceful slumber on soft beds, or take baths and get lacquered with oil by nubile servant-girls. Life could be worse for the Greeks.

I enjoyed The Odyssey more than The Iliad. More happens in the former. The latter is death upon death. Some of the deaths are very creative (a lance tears through a man’s face and yanks him off a chariot like a fish on a hook), but it’s basically a primal action movie. The Odyssey has a couple more interesting characters, and at least the scenes change.

I didn’t realize before reading The Odyssey that nothing much actually happens to Ulysses during the era when the story takes place. Instead, most of The Odyssey is spent in flashback: Ulysses stops off at some island or other on the way home, and while there reminisces about all that’s happened to him. Finally he gets so homesick that he simply must return to Ithaca. And he does. Why Homer chose this method, I’m not sure. Borges uses the same method. For that matter, so does The Arabian Nights.

Fagles’s translation is workmanlike. It’s highly readable, but it’s not very inspiring. People do what they’re supposed to do, and that’s the end of it. I didn’t feel the poetry. I’ve read Fagles’s description of the challenges he faced, and I appreciate them. He could have chosen to stick to the literal meaning of the original words, thereby making the book irrelevant to modern readers. He could hack the words to pieces to make them fit with the meter. He could completely adapt The Odyssey to a modern idiom, thereby deleting a lot of the book’s original force. Me, I would have liked more music.

The Fitzgerald translation is supposed to have more music. Pope’s translation is supposedly the most beautiful Homeric translation ever written. One of these years I’ll have to take an Ancient Greek class, just so I can read Homer in the original. I remember seeing someone on Jeopardy! when I was growing up who’d learned Ancient Greek for just that reason. Back then I thought that was ridiculous. Now it seems perfectly sensible.

Philip Roth’s latest book arrived at the library for me yesterday. It’s the final (so I’m told) book in the Nathan Zuckerman series. Zuckerman, like the protagonist in many of Roth’s works, is possibly a stand-in for Roth himself. At other times (Operation Shylock among them), the protagonist’s name has in fact been “Philip Roth.” Roth — the real-life person — delights in confusing us about where the author ends and the character begins. By this point, it’s hard not to believe that Roth himself is a bitter, incontinent, impotent old man who hides in the Berkshires away from civilization, coming down from the mountain just long enough to buy groceries. Perhaps the real Roth rejects even a friend’s gift of kittens, as Zuckerman does at the beginning of Exit Ghost. Maybe too much companionship distracts the real Roth from the full-time business of being a writer, as it does Zuckerman.

Or maybe not. When the protagonist (“Philip Roth”) in Operation Shylock goes on a quest to find and destroy a man who’s claiming in major newspapers to be the famous author Philip Roth, and using the real author’s good name to spread his own (i.e., the impersonator’s own) noxious ideas, the real Roth may well be telling us to stop looking for the author in the book. Just read the book, appreciate it as literature, and let the book be enough. Maybe. I’ve not made up my mind.

If this were all that one got from Roth’s books, it would be little better than navel-gazing. Instead, his works are embarrassingly rich troves of ideas and intoxicating stories. They’re embarrassing to a non-writer because they’re like a well-executed magic trick: the writing is so simple — almost slapdash — that I’m never sure what it is that gets me hooked. By the fifteenth page of Exit Ghost, I couldn’t put it down. Roth had already set up and torn down two threads that could have been the foundation for the book, and had already introduced what might be a love interest but then again might not. As usual for Roth novels, Exit Ghost feels to me like a sequence of confidently executed quick brush strokes that quickly form a painting of entrancing beauty.

If we decide to view Nathan Zuckerman as a mirror on Philip Roth, then the real author is a husk of a man who’s seen his death approaching for ten or fifteen years. Let’s hope, then, that the character has nothing to do with the man. As a selfish reader, I hope I spend many more years in the company of this awe-inspiring author.