Jan 1, 2008
Justinian's Flea: Plague, Empire, and the Birth of Europe, William Rosen (2007)

This is the sort of book I would write. I think maybe I recognized that when I read a review of it in the Economist and was intrigued enough to put it on my list.

Viewing oneself in the mirror can be very enlightening ... if one is willing to see what's there. I've never written a book, and I probably wouldn't have the perspective to see it objectively if I had, so reading a book that's like what I might have written is the next best thing. It's particularly informative if the book is flawed, which this one is.

By idiomatic habit I want to say "deeply flawed", but that wouldn't be right. The depth of Justinian's Flea is just fine. It is broadly and shallowly flawed. One thing the review mentioned is that the author hops quickly from topic to topic. He is the sort of polymath who has so many interesting thoughts on so many topics that he can't sit still long enough to focus on just one. On the whole I like that sort of thing, and I think it's a tendency I have as well, but reading this book helps me see that one can go too far with it. At times it seems that the narrative is just one digression after another. After a while it gets disorienting.

It's at the middle scale where the lack of focus bothers me the least. At the large scale, the lack of overall structure was vaguely disturbing. Rosen does have a thesis, but it's so flimsy I have to wonder if his heart is really in it, and it's never strong enough to make me feel like the whole book serves this central argument. As much as I enjoyed the many digressions (and truly I did), I kept wanting to say, "OK, so what's your point? Where are you going with this?"

The thesis is that the plague that hit the Mediterranean world in the time of the Emperor Justinian altered the course of history drastically. The scope of the alleged alteration isn't entirely clear. Certainly, Rosen attributes the fall of Rome to it. At times he seems to trace other things to that, like the rise of modern Europe or the eclipse of imperial China (because China didn't get the plague.)

Like any historical argument that small event x was the cause of large trend y, it relies of the "for want of a nail" fallacy. Even the tiniest event can be linked to circles of effect that ever-expand as you move further away, but insofar as every event is momentous, none is special.

Rosen not only acknowledges this, but devotes the first five paragraphs of his book to explaining that very fallacy. (Characteristically, he comes at it obliquely, making a metaphor out of the gravitational perturbation of the planets.) On the third page he finally settles in and says,

OK, fair enough. But I find it curious that he feels the need to launch his book with a two-page summary of why a thesis like his can never be true ... before proceeding to build the book around that thesis nevertheless.

He does it elsewhere, too — for instance, introducing the section where he draws parallels between Justinian's Rome and Khosrau's Persia. The formula is the same: "One shouldn't make a case like such-and-such [several sentences explaining why not], but on the other hand [brief rationalization], and so I will," and off he goes with the sort of argument which he just told you one shouldn't make. Is this a tic? A personal compulsion?

I wonder because I can easily imagine myself wanting to do the same.

At the small scale, Rosen is positively addicted to parenthetical comments. He can barely finish a sentence without interrupting it with some tangentially related thought. Sometimes they are enclosed by parentheses, sometimes by commas, sometimes by dashes. Occasionally a particularly diverting one will warrant an asterisk for a footnote at the bottom of the page. The variety of available punctuation helps to disguise the reality that there are just too damn many of them: Any time you find yourself having to juggle punctuation marks in order to make your complicated sentence comprehensible, that ought to be a hint that something needs to be taken out and put in its own sentence. Or maybe just taken out.

Any regular reader of Benzene knows that I suffer from the same affliction. It's good for me to witness excessive parenthetics from the reader's point of view for a change.

But as I said, the book's flaws are superficial. On the whole it's a jolly romp through history with a quirky guide, which is a good thing.

More than most, this book would have benefited from more aggressive editing. Ironically, Rosen himself is an experienced book editor, taking the role of author for the first time. Knowing this, I wondered if he felt that being an editor somehow made him immune for needing an editor himself, but no. Right there in the acknowledgments he copiously thanks his editors, remarking that he knows better than anyone how important they are. Either they didn't do a very good job, or I'd hate to see what Rosen's writing looked like before they beat it into submission.

I don't know what all is a copyeditor's responsibility — I like to imagine that everything is — but someone failed to catch several mistakes that go beyond style or grammar. I noticed several factual gaffes and misspellings of proper names. Is an editor expected to know these? I don't know. The average reader probably won't even blink at seeing Diyarbakir misspelled "Dyarbakir", but to my eye it leaps out.

Elsewhere, we read of Khusro honing "his appetite on the relatively modest spoils of Susa", which seems strange given that Susa is in the heart of Persia, the winter capital of several earlier dynasties. In fact, the city that Khusro has spoiled is Sura, on the upper Euphrates.

Other errors require less specialized knowledge to catch. Any editor should have winced at "the deference granted to whomever writes the laws" or "a literal avalanche of demographic and population shocks", not to mention simple misspellings like "emporer" and "aquaduct". (The latter looks right to me, even though I know it's wrong, because that's how it's misspelled on AsoBrain's pirated online Cities and Knights of Catan game.)

One last complaint: The book is framed by a short prologue and an even shorter epilogue. Both are set in all italics. Unless you're going to pick a typeface whose italic is more tolerant of large blocks of text than this one is, that's a bad idea. Italic is fine for a one-paragraph abstract to introduce a topic, but the prologue runs for three pages, and it's butt-ugly. That's one mistake that I wouldn't have made.

I know I'm being unfair to the book by focusing only on all the errors, but I can't help myself. They interest me, and they're what inspired me to scribble notes with page numbers for later recollection.

I wonder if the book would be more enjoyable or less for a reader who isn't as self-conscious of the quirky writing style. Or one who knows the subject matter less well. I'm not an expert on Justinian by any means, but I'm reasonably familiar with basic Byzantine history. Although I don't know all the stories Rosen tells, nor even most of them, I do know the basic historical context. When names of supporting characters like Narses or Amalasuntha pop up, I know who he's talking about. To me, reading a book about Justinian is probably a lot like reading a book about Thomas Jefferson would be to you.

On further reflection, I wonder if the real problem is not lack of structure per se, but just that the book is not true to itself. Maybe all Rosen really wants is to hop about from topic to topic telling interesting stories loosely related to a common theme. Maybe it was someone else who suggested he build it around the plague thesis, and the resulting half-hearted attempt at a core narrative proved worse than an honest hodgepodge might have.

Sometimes you just have a list of sundry thoughts to tack end to end and there's no point in trying to connect them. Sort of like me in this post.

Justinian's general Belisarius has reconquered North Africa. The Vandal king Gelimer has fled and taken refuge in the mountains with a primitive Berber tribe.

Did he now? Maybe Rosen knows something I don't, but the standard account of Gelimer under siege says that he requested a loaf of bread, a sponge, and a lyre. The sponge was to wipe his swollen eyes. The lyre was to accompany him in singing the ode he wrote about his misfortunes. J.B. Bury's account is here, and many others tell the same story, but in no account that I know of is the ode actually about the sponge. It sure sounds to me like Rosen (or maybe his source) has misread the account and garbled it. How many authors will now copy from Rosen and propagate this error?

This footnote caught my attention:

That probably means nothing to you, but if you know anything about the study of Hun history (what, you don't?), you will recognize "some historians" as Otto J Maenchen-Helfen, and this is a little swipe at him. And just in case you weren't sure, the footnote's asterisk is followed by extended quotes from Ammianus Marcellinus, describing the Huns in the foulest terms, exactly the sort of nonsense Maenchen-Helfen set out to debunk.

I have no position on whether the Hsiung-nu are or aren't the Huns, but I happen to be a huge fan and partisan of Maenchen-Helfen, so when I read this, I wonder. What is Rosen getting at? Is he challenging Maenchen-Helfen's scholarship? Or is he just mad because it was more fun to discuss Huns when they were more like orcs than human beings? Maybe he also misses the dog-faced men Marco Polo reported on the Andaman Islands.

At the passing of William F Buckley I was shocked and appalled to read an outpouring of sentiment among bloggers and, especially, blog-commenters that they hated his use of unfamiliar words. It seems most blog readers feel that any writing that requires one to consult a dictionary is an offense, and the offender is the writer. (Perhaps this is just a corollary of the axiom that most blog readers just don't want to read anything that they don't already know to be true.)

Hmph. If you ask me, any book that doesn't send you to the dictionary at least once probably isn't worth reading.

1. Newest to me in this book was theodicy. Here's the sentence:

Now this is a perfect example of an unfamiliar word that you don't really have to look up. If you don't know what theodicy is — and I didn't — you can still get the general gist of it. You see the theo- there, so you figure it's something like religious doctrine. The sentence is comprehensible enough, so rather than stop and look up a word, it's easy enough to just read on.

If you do look up the word, however, you find that it is "defense of God's goodness and omnipotence in view of the existence of evil." So your vague idea of religious doctrine was on the right track, but it's actually a very specific and interesting question of doctrine. Now go back and read the sentence. See how much more interesting it is now? See what you missed by being too lazy to open the dictionary?

The quoted sentence is also a choice example of Rosen's parenthesizing. The long bit between the dashes could have easily been a separate sentence, but he can't resist attaching it to the exact noun phrase of "dualistic religion". The asterisk points to a footnote about how the exiled Zoroastrians became India's Parsis.

2. Also fun was seeing sequelae, which I know only as a Scrabble word (though I've never had the pleasure of seeing it in a game). Sequelae is the plural of sequela. Related to a sequel, a sequela is a secondary result of some event. Thus it is exactly the right word when Rosen says that

3. Enjoin is one of those Möbius-strip words whose meaning has evolved by such an odd pattern that it has twisted back to become the flip side of itself. Like sanction, it makes you stop and say, "wait, does that mean you're in for it or against it?" My dictionary, starting with the second definition, says, "2. forbid, prohibit [...] syn: see command." Isn't command the exact opposite of prohibit?

Usually you can count on context to make the meaning clear, but it isn't much help in this sentence:

Eventually, you figure out what he means, but it sure isn't obvious.

4. I have worked as an editor at times, and I count copyeditors among my friends and relatives, but one sentiment I don't share with the copy-editing community is their passion for removing hyphens. It's not that copyeditors dislike hyphens. When it's a question of inserting them between separate words, they love hyphens. Copyeditors live to tell you that your whole sentence is one big compound adjective that must therefore be treated with hyphens. But when it's a question of a hyphen whose removal might leave a solid word, they can't wait to take it out. The rule is that copyeditors love to smash words together. Exhibit A: "copyeditor".

Me, I'm old-fashioned. When a prefix is newly attached, I like to see a hyphen there. I prefer quasi-libertarian pseudo-intellectuals to quasilibertarian pseudointellectuals. I wouldn't mind the solid setting so much if the joining point were at least preserved as the favored spot for hyphenation when the word is broken over a line, but typesetting software never seems to figure that out, so we get breaks like cow-orker and nona-ligned.

That's my problem with protonation, a favorite word of Rosen's which he uses repeatedly in this book. I look at that and say, "protonation? what's that?" Is it the result when you protonate something? Has it something to do with tonation? (Whatever the heck protonate and tonation mean.) No, it's really just a proto-nation. But even after he's used the word several times, it's still hard for me to see it that way.

Where the main text says that Spain had been a province of Rome for nearly seven hundred years, an asterisk points to a footnote saying,

Actually, no. Rosen has Baetica and Lusitania reversed. (Here's a nice map illustrating both this point and the next.)

Shortly after that we read how the Suevi (a barbarian tribe) are "largely pinned on the western side of the Tagus River". Well, I get what you mean, but given that the Tagus runs predominantly from east to west, with a southwestward trend at the end, that's sort of like saying the Virginians are trapped on the west side of the Potomac.

Speaking of Virginia, here's a case where he could have used a bit more explanation. He's describing Rome in the time of Constantine, the emperor who left Rome to build a new capital in the east:

Whoa, are you keeping up? I see what he's getting at, but that's quite a lurch to suddenly jump to 19th century America without warning. And if you aren't reasonably well educated about early American history, you might still have no idea what he's talking about. Is his goal to illuminate for the reader the situation in 4th century Rome? If so, perhaps he'd do better by spelling it out a little more, rather than relying on a comparison that is likely to be lost on half his readership. (But my inner voice asks: who else likes to make obscure allusions?)

Rosen loves to throw in offbeat allusions to art and literature. This can be fun, but sometimes they go awry. Here he's discussing the Ostrogoth king Theodoric — well, no, actually, he's discussing a later Ostrogoth king, Totila, but he compares him to Theodoric, prompting this aside,

Nothing there is purely false, but it will give you the wrong idea. The punctuation doesn't make it clear, but "Nibelungenlied" (as it is usually spelled) is the title of the anonymous epic, not Wagner's opera. Wagner actually wrote a cycle of four operas and gave the title Der Ring des Nibelungen to the whole thing. The common thread of "Nibelung" has led many to believe that the entire series is based on the Nibelungenlied, but in fact only the last of the four operas, Die Götterdämmerung, is. Not only that, but it's based on the first half of that epic, which tells the story of Sigfried, Kriemhild, and Brünhild. Dietrich of Bern doesn't appear in the epic until the second half, which Wagner doesn't use at all. So yeah, Dietrich really is the hero of the Nibelungenlied, and the Nibelungenlied really is a primary source of one great Wagner opera. But really, Dietrich has nothing to do with Wagner.

Another character in the second half of the epic is King Etzel, whom we know better as Attila the Hun.

That Dietrich equals Theodoric is easier to grasp when you realize that the latter rhymes with "Roderick", not "meteoric". The Bern that he hails from is not the one in Switzerland; it's a Germanized rendering of Verona, the city in northern Italy.

Theodoric's daughter is Amalasuntha, queen of Italy, whom I mentioned back there a ways. (Rosen spells it "Amalasontha", following Gibbon.) I don't know the Nibelungenlied well enough to say whether she makes an appearance. Just as the Gothic name Theodoric eventually became the German name Dietrich, the Gothic name Amalasuntha is the precursor to the English name Millicent. In French, the name is Mélisande, titular heroine of another opera. (If I were Rosen, I would have tried to insert this information between dashes immediately after first mentioning Amalasuntha, but I'm content to postpone my tangential observations by several paragraphs.)

But this is not Rosen's worst misadventure in music. That comes in a discussion of architecture. I don't know a thing about architecture, so I can't judge his main argument. It seems sensible enough to me.

No problems so far, but the that asterisk points to a footnote that the editor should have just expunged altogether. It says:

A little knowledge is indeed a dangerous thing. Do I make such gross errors when I let slip parenthetic comments on topics of which I'm ignorant? I hope not. This is wrong in so many ways I barely know where to begin.

The piano is not tempered to make the twelve semitones uneven, it is tempered to make them even. (That's why they call it equal temperament.) I want to say that Rosen has it exactly backward, but even his backwardness is inexact. Although equal temperament makes the intervals even, it does not make them more pleasing to the ear; it makes them less pleasing. (Yes, that's right: tuning a piano really means putting it out of tune.) The intervals that are most pleasing are those that most closely represent simple arithmetic ratios. Equal temperament pushes intervals away from these ratios, not toward them.

This topic is testing my ability to be concise, and I'm failing the test, because now I need to explain why simple ratios and equal semitones are incompatible. A musical interval is defined not by the difference between the frequencies of the two notes, but by the ratio of one to the other. Therefore, when two intervals are added end to end, the combined interval is not the sum of the smaller intervals but the product of them. Conversely, when an interval is divided into equal parts, those parts are measured not as equal fractions of the interval but equal roots. For example, if you wanted to divide the ratio 8:1 into three equal parts, each part would be 2:1, because 2:1 = 4:2 = 8:4. That is, you're not dividing 8 into thirds, you're taking its third root.

An octave represents the frequency ratio of 2:1. Since any root of 2 is an irrational number, the division of the octave into any number of equal parts will, by definition, yield only intervals which are irrational, and thus imperfect in terms of the aurally pleasing simple arithmetic ratios. It's like squaring a circle; it simply can't be done. That is why for so many centuries the scale was not equal.

Unfortunately, the inequality of the scale, while fine for isolated melody and simple music, becomes an obstacle to complexity in music. The inspiration of equal temperament is that if you divide an octave into 12 equal parts, even though none of the intervals are mathematically perfect, many of them are close enough that they don't offend the ear too much. In exchange for this tolerable compromise of aesthetics, the equal tempered scale makes complex harmony and modulation possible.

This only works because, by happy mathematical coincidence, several integral powers of the 12th root of 2 are remarkably close to simple ratios. 2^(7/12), for instance, is not exactly 3/2, but it's very close. By the exact same math, 2^(5/12) is very close to 4/3. Additional coincidences make 2^(4/12) reasonably close to 5/4, and 2^(3/12) reasonably close to 6/5. These are our equal-tempered perfect fifth, perfect fourth, major third, and minor third. Other roots of 2 don't generate such close approximations, and that's why we divide the scale into 12 parts rather than some other number.

(The kicker to this story is that even though the rational intervals are supposedly more pleasing aesthetically, empirical studies demonstrate that anyone who has become habituated to the imperfect intervals — which is to say pretty much anyone who doesn't live in a remote jungle in New Guinea — learns to prefer them.)

Of all the bizarre asides, this one takes the prize. I have to quote the whole paragraph to give the proper context. We're back with Khusro now. He's Justinian's counterpart in Persia, sometimes known as Chosroes or Khosrau or any of several other spellings. (Transliteration of classical Persian names is pretty wild; Xerxes, for example, is Khashayarsha. I think it has something to do with the fact that they all passed through Greek on the way to us.) The discussion is of Khusro's capital, Kisfun (aka Ctesiphon).

Don't get me wrong. I'm not complaining. I actually enjoyed reading all these little details. But still ... You're reading a book which is supposedly about how the Justinian plague changed the course of history in Europe. A chapter is set aside to tell the story of Justinian's geopolitical opponent, the Persian emperor Khusro. In that chapter, we take a side trip to discuss the Persian capital of Ctesiphon, in which a long paragraph is devoted to describing the palace. Two sentences of that paragraph are set aside to discuss the famous carpet in that palace, and — oh by the way, some obscure classical composer in the 1970s wrote a sonata named after that carpet.

That's the most extreme example, but really it's that kind of book.

And by the way, "obscure classical composer in the 1970s" is an oxymoron. You can be a famous classical composer, and you can even be a renowned composer in the 1970s. But if you're a classical composer in the 1970s, sorry, but you're obscure. Go ahead and try to prove me wrong.

Oh, and "absolute zenith", way up there in the first quoted passage of this post. That's an oxymoron, too. A zenith is absolute by definition. There are no relative zeniths. See, there I go with postponed parenthetics again.

In the more recent quoted passage, yes, there's an asterisk there. Another damn footnote. It doesn't matter what it says.

I think Rosen is incorrect where he says all that remains of the palace is the arch. In these picture of the ruins, I see several walls intact in addition to the arch. That arch is awesome, by the way. It's so freakin' huge. I'm a big fan of gigantism in architecture.

11:39:47 PM  [permalink]  comment []