History Repeats Itself in the Notices of AMS

Before Communications of ACM became cool again, I’ve been a regular reader of the Notices of American Mathematical Society. I still check it out occasionally to keep tabs on the mathematical community.

This month’s issue featured a lengthy article with a lofty title “Mathematical Methods in the Study of Historical Chronology”. It covers the work of Russian mathematician Anatoly Fomenko and his frequent co-author Gleb Nosovsky, where they re-examine traditional chronology of the history of the ancient world based on astronomical records and statistics.

Fomenko’s starting point is that much of traditional dating methods are circular (events and archaeological artifacts are dated relative to each other with very few absolute references). However, intervals between noteworthy astronomical phenomena, such as solar and lunar eclipses, are sufficiently unique within the period covered by recorded history, so that they can be used to anchor chronicles mentioning these events. Zodiacs and star atlases can also be dated independently based on the perceived movement of stars, some of which have shifted considerably over the two millennia. According to Fomenko and his co-authors, doing so systematically reveals stark inconsistencies in traditional chronology, which has to be revised if not completely rewritten.

This narrative is appealing and captivating: A lonesome hero, maverick and iconoclast, rides into town and wielding nothing but a Magnum and Matlab cuts through the fog of conventional wisdom, “Moneyball” style, chasing out the charlatans and humanitarians. I am mixing my metaphors here but as we shall see it is quite fitting Fomenko’s theories. We love to root for the underdog, especially if the little guy is one of our own: David against Goliath, Frodo against Sauron, Nate Silver against the GOP.

The Notices article describes several Fomenko’s methods but oddly enough does not offer any examples of their applications. It is indeed strange, since some of these methods are at least thirty years in the making. Surely, they must have led to something interesting? I will fill the gap by consulting the magnum opus of Fomenko et al. “History: Fiction or Science?” spread across four volumes (in English; there are several more in the original Russian).

Thucydides in his monumental history of the Peloponnesian War describes three eclipses, two solar and one lunar, 7 and 11 years apart. Traditional chronology places these events in the years 431, 424, and 413 BC, which disagree in some details with Thucydides’ account (such as whether the first eclipse was total or partial). To find a solution fully consistent with his description one has to fast forward to 1039 AD.

Rather than to admit that either the original text or its surviving copies are not fully accurate, as traditional historians do, Fomenko concludes that the Athenian general actually did live in 11 century AD, less than 1,000 years ago. After assembling a few more examples of similar inconsistencies, he declares the entire history of the ancient world a fabrication done in the Renaissance.

Was it a fraud or a giant mistake perpetrated by generations of historians, building upon each other’s mistakes? Fomenko and Nosovsky identify certain patterns to the history that never was. Namely, they assert, that most of what passes for historical accounts are copies of actual events, shifted and disguised by changing or translating proper names and places. The Notices article vaguely mentions it in the context of the Bible authorship. In reality, exercises in identifying parts of history that are duplicates of each other constitute the bulk of Fomenko’s work.

This is where Fomenko and his comrades abandon the scientific method for good and stand with both feet firmly planted in quackery. Everything goes to establish equivalence between events, locales, or historical figures: a fleeting phonetic or graphical similarity (often present only if one spells the name backwards and in Russian), parallelisms in some biographical or geographical details, or just a hunch. A couple of quotes from volume IV, which covers what is traditionally thought of as history of Britain, illustrate this approach nicely (page 614):

Galfridus proceeds to tell us that the city of New Troy, or London, has been founded on River Thames. We believe the name to have been a reference to the Bosporus initially, which is where we find Constantinople. This strait is very long and relatively narrow; it does look like a river on maps […]

Let us also take a closer look [at] the word Thames. Bearing in mind the Oriental manner of reading words from the right to the left and the word “sound”, a synonym of the word “strait”. Reversed and unvocalized, it looks as “DNS” – possibly, a version of TMS (Thames).

I mentioned that according to Fomenko, much of ancient history (actually, all the way to the XVI century) was not entirely made of the whole cloth but was obtained by shifts, translations, and superposition. Where did the originals take place, and when? It turns out, astonishingly, that most of the world’s history happened somewhere in the steppes of the Eastern Europe and the Middle East. Again, from the same text (page 626):

The fact that we managed to identify the Scots of the XII-XV century as the Scythians must also imply that the term Irish had been synonymous to the term “Russian” in the epoch in question (RSS or RSH = Russia sans vocalization); the name Ireland may also have referred to Russia once.

I stop taxing your patience here (similar free associations fill thousands of pages). Hopefully it is enough to convince you that Fomenko makes “The Bible Code” look as well argued as Bourbaki.

Fomenko has quite a following in Russia, with his books widely available and debated there (I counted 65 books co-authored by Fomenko currently in print, not including his textbooks on topology and differential equations). It is not truly surprising given that Russia is known as a country with unpredictable past (Orwell’s “He who controls the present controls the past” rings as true today as ever), and its national pride has been badly hurt by the post-Cold War world order. In such an environment, Fomenko’s version of history as a world-wide forgery that ascribes to other nations the events of Byzantine and Russian past is simultaneously believable and appealing.

Promoting these views from the pages of the Notices of AMS is a different matter. Most troubling is that this is not the first time in recent memory that we are having a similar conversation. The last time the Notices got a lot of coverage in CS blogs, it was in reaction to Neal Koblitz’s article alleging that modern cryptography is a fraud wrapped in bad mathematics.

Mathematicians are known (as are men and women of other professions) to occasionally adopt extreme or radical positions. Most outrageous examples include Theodore Kaczynski (the Unabomber) or Serge Lang who was an outspoken AIDS denialist. By themselves, mathematics credentials are not a guarantee of inerrant judgement in matters remote from one’s area of expertise. Nothing I’ve seen suggested that mathematicians are particularly prone to be taken to conspiracy theories, but neither are they immune to them.

I do find it disturbing that the Notices offered their pages to views antithetical to mathematics (Neal Koblitz lamented the use of proofs by cryptographers) and to promotion of phony scholarship. The trajectory is downhill: the article of Neal Koblitz (and a follow-up, joint with Alfred Menezes) merited a serious response, while Fomenko and Nosovsky’s writings are studied as expressions of the nation’s psyche. In departure from standards of scientific publications, controversial viewpoints are not accompanied by rebuttals, other than by much delayed and necessarily curt letters to the editor. (The only source critical of Fomenko, which is acknowledged by the Notices article, appears in the references but never cited in the text.)

The most charitable explanation I could fathom was that the article was an elaborate April Fools’ joke (even though the issue became available weeks before April 1st). If this is so, consider me punk’d. Otherwise, it is the editorial board and the readership of the Notices that got duped.