Since it's the sort of thing that deserves to be posted publicly, and usually ends up going inside the cover of a work that collects dust and never gets consulted again, I thought I should reproduce my acknowledgments here. It really would have been impossible without you.

Any large undertaking incurs many debts of gratitude, and this thesis has been no exception. My profound thanks must go to my husband, Mike Knell, who initially encouraged me to return for the D.Phil. program, and who has ensured constant moral support, occasional typing assistance, and uninterrupted supplies of tea. The thesis would have been equally impossible without the support of my supervisor, Prof. Theo M. van Lint, who has seen me through with patience, humour, and frequent reminders that literature is as important as history.

I would also like to thank the following people and institutions:
     The Mekhitarist Fathers of Venice and Vienna, the Maštocʿ Institute of Ancient Manuscripts (Matenadaran) in Yerevan, the British Library, the Bibliothèque Nationale, and the Hill Museum and Manuscript Library, for graciously providing physical access to, and digital copies of, the manuscripts of the Chronicle.
     The Colin Matthew fund of St. Hugh's College, the faculty fund of the Oriental Institute, and Dr. Jean Knell, who each provided financial assistance necessary to visit these libraries; as well as the Nubar Pasha fund, who provided the financial means to obtain the necessary manuscript reproductions.
     Linacre College, for their award of a Mary Blaschko Graduate Scholarship for 2006–2008 which enabled me to begin the D.Phil. degree, their assistance with conference travel via the Old Members' Trust, and their indispensable assistance in various administrative matters.
     Dr. James Cotton of Queen Mary, University of London, who provided invaluable advice and assistance in the application of phylogenetics to my manuscript data.
     Lou Burnard and Dr. James Cummings of the Oxford Text Archive, who advised on the possibilities of TEI as a format for collation output and critical edition.
     My legion of proofreaders, including Dr. Alan Knell, Dr. Jasmin Raufer, Dr. Tim Stadelmann, Michael Kröll, and Andromeda Yelton.
     The members of the Vienna.pm Verein, for unfailing hospitality on my multiple visits to the city and the library there, and for the opportunity to speak about my work on Encritic to the Twin Cities Perl Workshop; and the members of the London Perl Mongers, who were ever on hand to assist with technical matters.

A friend pointed me the other day to a Guardian article which was full of the usual ignorant mish-mash of "solutions to help Main Street, not Wall Street" and included this:

At the same time, several steps can be taken to reduce foreclosures. First, housing can be made more affordable for poor and middle-income Americans by converting the mortgage deduction into a cashable tax credit.

The Guardian article is the only place I've seen this idea explicitly, seriously suggested, but I have seen a similar sentiment ("The government should be helping out Main Street, not Wall Street!" "The real victims here are those who took out subprime mortgages? What's in it for them? Why aren't we helping them out directly?") in very many places. But actually, the idea of "giving money directly to Main Street, and letting Wall Street hang" is precisely counterproductive to this goal.

So, let's think about this for a minute. Pretend that, suddenly, the US passes legislation to allow mortgage interest to be a tax credit. What this means is that you can tell the US government that you spent $X on mortgage interest in 2008, and they will subtract $X from your tax bill. If this makes your tax bill negative, the government will send you a check for $X - (your tax bill). It is basically saying, "the US government will pay your mortgage interest."

This sounds great, doesn't it? If you're a US citizen with a mortgage, anyway. But let's think about it for a minute. The interest on a loan is the money that the bank who owns the loan is making, as recompense for your having taken their money to buy something for yourself. Generally, when you have first bought a house, the mortgage interest comprises the majority of your payment. (Play with a mortgage amortization calculator if you don't know what I mean.) So the Guardian writer is proposing, basically, that the government should pay the bit of people's mortgages that makes up the vast majority of the monthly payment right now. It also means that the government should pay, over the life of the loan, the part that is the payment to the lender for having made the loan. And so:

1. This will make it a lot easier for people to pay the mortgage, at least in the first few years.
2. This means that most of the mortgages that have been sliced & diced into CDOs will be good.
3. This means that the banks who bought CDOs will receive the profit, that is the mortgage interest, courtesy of the US government.
4. This means that the "toxic debt" won't be toxic anymore.
5. This means that the CDOs will have precisely the inflated value that the banks assumed all along that they would have.
6. This means that the US government will directly be paying interest money to the banks who hold the debt, that is, the investment banks who bought the CDOs.
7. This means that the "bad, stupid" CDOs that the investment banks bought won't be so stupid after all—they'll be guaranteed by the full faith and credit of the United States government! There will be a party on Wall Street again after all!

It's a great way to get the money straight to Wall Street, with some incidental benefit to Main Street. But wait, I thought we were supposed to be bailing out Main Street and not Wall Street? Oops.

The best feature about the Paulson plan is that it pays a fair price (including the pessimism about Main Street being able to pay their mortgages) for these CDOs, and since it's a fair price, they will be able to sell these CDOs for at least the price paid, in future. The taxpayer will not, in the end, lose money if this plan is executed properly. (Yes I know, that's a big "if", but at least it's not a dead certainty that we'll lose money.) The plan has the side benefit, which Congress seems to have noticed, that the US government would hold a lot of slices of a lot of mortgages, and it would suddenly become possible (as it isn't at the moment) to put these mortgages back together and re-negotiate the terms with homeowners on the edge, and try to keep them in their houses. But the main benefit is that Paulson has come up with a way to put a bunch of taxpayer money on the line to keep the economy functioning, and get the money back in the end, so that the taxpayer won't be out of pocket in the end. This is the part that I love. How many government employees do we have who can still think of ways to retain the taxpayer's money in such an impressive way?

Fortunately for me, the NPR show This American Life did an hour-long broadcast where they explain it in depth, and they did a pretty good job. Even more fortunately, there is now a transcript available, so you don't have to spend an hour listening. I think it was well done, and it does a very good job of showing the natural market motivations that arose to cause a lot of people to make a lot of bad decisions, without having to postulate any criminal conspiracies.

So, what to do about it? I find it hard to know what to advocate. The Paulson Plan looks at the moment like it is going to go through. I think it could work, if it is managed intelligently, and I think that Paulson has the skills to manage it intelligently, but a) he could make a mistake, and b) he could find his hands tied by the fact that he is working for the US government now, and not for the board of Goldman Sachs, and he must therefore show that he can be persuasive to a set of people that thinks along completely different lines. One major advantage that could come of the Paulson Plan is that, suddenly, the zillions of mortgages that have no clear owner will have a clear owner, and it will be possible to re-negotiate the terms of those mortgages. One downside to that, of course, is that you'd be hard-pressed to find an American who wants to be in debt to the government (and to their primary financial interface to the US taxpayer: the IRS.) And I haven't even begun to touch the ideological arguments over whether the government should be in the business of mortgage administration, or of buying commercial securities on the open (albeit distressed) market.

I find that I am not all that concerned with "punishing the guilty". There was clear fraud, at the very least on the part of the mortgage brokers who put down fictitious salary numbers on behalf of their clients. Those would not be hard to prosecute, if the clients who were sold the mortgages have held onto their documents. That would be far more satisfying to me than arbitrary caps on executive pay, which have no economic basis and, as far as I can tell, have only been proposed in order to make the "have-nots" feel like the "haves" are being punished at least a little.

What I do care about is the systemic problem. (Of course I do; I'm an engineer.) I'd love to get my hands on the design of a proper regulatory system for the US. Over the last twenty or thirty years, the financial markets have become an amazing ground for creative engineering. This has a lot of parallel with the computer technology world, and probably for a lot of the same reasons. The only way to devise appropriate regulation for such financial innovation is to innovate. The Fed needs a well-paid team of financial strategists to go over new products as they come out, and as they go to market; the job of these strategists is to look for things like asset bubbles—sudden inflations in house prices, or sudden explosions in the number of mortgages approved; that is, precisely the sort of thing that people were spotting four or five years ago—and dig around for the causes. You can't regulate effectively until you have seen the consequences, both expected and unexpected, of a new activity.

I still think Hank Paulson is a smart guy, and I still hope that his plan works. I wish I had some more effective way than "a blog with a small readership" to debate these ideas with policymakers though.

It seems to me that the New Yorker is the irony-impaired party here; perhaps their cartoonist's theme song should be Alanis Morissette's "Ironic", filled as it was with descriptions of things that are not ironic at all. The cover is drawn like any old political cartoon; these are, of course, almost always satire, just as the New Yorker's cover claims to be. The message of most political cartoons is clear - it's putting a satirical spin on some situation its subject was in. The message of, say, this one is to satirize the (real) hypocrisy of the US about global pollution. This one is making fun of Jesse Jackson based upon things he's really said and done.

The problem with the New Yorker cover is that there is no clue, either visual or via a caption, to tell the reader "this is a satire of something that isn't real", or "this is a satire about a bunch of false images of the subject." You're expected to take a leap of comprehension that isn't normal for political cartoons, with no cue to do so. If that cartoon were printed on the cover of some sort of Fox News magazine, without caption, the message would be clear: Obama is a Muslim terrorist-loving America-hating weirdo. As it is, coming from the New Yorker, the reader is left with (not particularly funny) ambiguity, and confusion about what the message is supposed to be. Basically, if the whole image were drawn inside a TV screen box with a logo of "Fair and Balanced" or some such, it would have been a lot funnier, much more understandable, and probably not have caused a whisper of outrage.

So the critics of the critics, in a sense, are right. Americans are terrible at irony. It's just that the New Yorker is right up there with them.

I have been branching out of my own field for the past couple of weeks, and attending a comparative philology seminar -- although I have a thing for languages, I have never looked very much at linguistics. It happens that the seminar is focusing this term upon Armenian, a language I know to some extent, which means that it's easier for me to follow the talks than it might otherwise be.

The speaker for today was forced to cancel his appearance on short notice, but the ever-resourceful convenors of the seminar came up with a brilliant substitution. Who knew that Indo-European phonology could be turned into a board game?

Each player (or team) draws a card from a shuffled deck. The card has an Indo-European root word and a modern-language word (in this case, all of the modern words were Armenian) into which the root word has evolved. On the back of the card there is a sequence of stages that the evolution may have taken. This sequence is handy if a player (e.g. me) knows pretty much nothing about linguistics or phonology. On the other hand, if a player is good at phonology, he is free to arrive at a different (possibly shorter) route of evolution, as long as it is valid. For example, the card that my team drew was:

*k^wóteros > ór (the nominative / accusative relative pronoun)

The steps on the back of the card were:
1. *k^wóteros > óteros
2. *óteros > óyeros
3. *óyeros > óeros
4. *óeros > óros
5. *óros > órokʿ
6. *órokʿ > órkʿ
7. *órkʿ > ór

Now certain of these changes must happen before others; for example, the 'y' in step #3 can't drop out until the 't' has become a 'y' in step #2. Others can happen independently; for example, it doesn't matter whether the first step I play for is the loss of initial kʿ in step #1, the changing of the last consonant from s to kʿ in step #5, or the change from t to y in step #2.

So everyone has his card, and his token; these tokens all get placed on the "Start" square of the board. Each player rolls between 1 and 10 dice; each die has a 50% chance of rolling a 1, a 33% chance of rolling a 2, and a 17% chance of rolling a 0. Add up the numbers, move the prescribed number of spaces in any direction you like. Each square represents a known phoneme shift in Indo-European, e.g. "*t > y between vowels" or " *t > tʿ ". In our example above, I very much don't want to land on the latter square -- if the t in my initial word changes to an aspirated t, it cannot then change into a y, and my word evolution will be torpedoed. The winner is the first player who has travelled around the board landing on all the squares necessary to evolve his word in a workable order, without making a wrong move. Squares that represent changes that are not applicable to your word (e.g. "*d > t") are safe, but don't help you make progress.

An advanced version of the game gives players three or four words to work on at the same time. The same principle applies, but here the player must be careful to avoid squares that will derail the evolution of any of his words. On the other hand, of course, a single square might progress the evolution of all three!

One of the attendees proposed a variant in which one player's move affects the other players; this would open the possibility of player attacks on each other. I think this would be brilliant, although there would have to be a mechanism whereby a player can recover from a bad shift. Any game designers out there want to have a crack at it?

So there I was, stomping around Southwark in a very bad mood indeed. I had planned to camp out in Trafalgar Square to watch the opening ceremony for the Tour de France, but with an hour to go the square was getting unreasonably packed, and I had concluded that the last thing I wanted was to be jammed in the middle of a huge crowd for three hours. I had my camera, and I was pretty close to the stage, but I just knew that some idiot (you can tell an idiot by the way he/she is taller than you and blocking your view) would make it impossible for me to get any decent pictures. Plus, I was by myself, and who wants to be alone in a crowd?

I was now in Southwark, instead of Trafalgar Square, because I had decided to give up on this whole opening ceremony business, and intended to meet a couple of people there. They were both being suddenly and stubbornly unreachable by phone. After checking all the pubs I could find in the vicinity, with no success, I couldn't think of anything to do but go watch the opening ceremony after all. This time, I got off the tube at Westminster, and began to walk up Whitehall. The street was pretty clear, so I sat down on the curb some distance short of Trafalgar Square to watch the giant TV screen that had been set up.

It was, in short order, time to introduce the teams. Agritubel made their appearance on the big screen, followed by Saunier Duval and Milram. Around the time Milram was being introduced, I saw a small pack of riders in Agritubel colors cruise down behind some barriers right in front of me.

Holy cow. Lucky I had my camera.

Lucky my friends didn't answer their phones, in fact.

I spent the next couple of hours having gloriously clear shots of all the pro riders as their teams took turns cruising up and down Whitehall behind the barriers, while half a million people were stuffed in the middle of Trafalgar Square with no idea this was happening. I've attached my first amazed shot of David Millar to this post; the rest can be found here on Flickr (with a bonus shot of Didi Senft thrown in.)

Alas, my camera battery died with six teams to go, so there is no QuickStep and no CSC. There is consequently no record of the highlight of my evening, which was when I got spectacularly high-fived by Dave Zabriskie.

I just hope I'm so lucky tomorrow.