Stephen Laniel’s Unspecified Bunker

I’ll be traveling from tomorrow through May 12, so expect silence from me until then. Right now the itinerary as I know it involves flying from Boston to New York tomorrow, New York to Istanbul overnight (May 1 to 2), and Istanbul to İzmir May 2, then taking a train from İzmir to Selçuk and, presumably, collapsing after 17 hours of travel. I have no idea what’s happening between the 2nd and the 12th. I don’t especially care. I intend to eat a lot of amazing food, carry a Turkish-coffee buzz with me wherever I go, and drink plenty of rakı. The rest is noise.

Have a wonderful 12 days, everyone. With any luck, the Democratic nominating process will be over by then.

If you read enough, you just have to be wary of Here Comes Everybody and its ilk. By now you’ll have encountered scores of essays on why X Will Change The World, or What X Means, or What X Will Mean In The Future Once People Realize. Especially for books about the Internet. If you’re the sort of person thinking of reading Shirky’s book, you’ve probably also read Larry Lessig (Code), Yochai Benkler (The Wealth of Networks, not to mention essays like “Coase’s Penguin”), Shapiro and Varian (Information Rules), maybe Weinberger (Everything is Miscellaneous), and on and on. You’ve used the Wikipedia. You may well use Linux. You’ve learned about “the wisdom of the crowds” (Surowiecki). You’ve got “the long tail” in there somewhere, namely the idea that places like Amazon and Netflix make more money from obscure titles than they do from popular ones: the popular ones sell more individually, but the obscure ones sell more in the aggregate.

So maybe by now you’re just saturated with Internet prognostication. I know that I am. I’ve had enough of Coase’s Theorem, which says that organizations emerge when it becomes too costly to transact all your business in the marketplace — i.e., when “transaction costs” get too high. People like Benkler and now Shirky have asserted that the Internet is a third way: rather than relying on managerial commands like a firm would, or price signaling like the market, we have projects like Linux and the Wikipedia which use market-like coordination for very large projects, using mostly trivial contributions of labor from each user to construct a beautiful edifice in the aggregate. I’ve read this enough times by now; I’m ready to move on.

Finally, I’ve read a zillion times by now the adage that “freedom of the press used to apply only to those people who could afford a printing press.” The Internet changes that. It makes a lot of things cheaper. It turns one-directional broadcasting (television, say) into many-directional conversation (blogs, LiveJournal, etc.)

These are all fine observations. They’ve been made over and over. What Shirky adds to this cacaphony is an important special case of all of the above: the Internet lets us form groups effortlessly. Now we can work together on projects that we wouldn’t have known about otherwise. We can find other people for fun in the real (non-Internet) world. We can find people with remarkably obscure interests matching our own. Previously these would have taken far too much time and effort. And the payoff is far too low for any company to be interested in connecting, say, lovers of ancient Chinese art. What the Internet has given us is a set of tools that allow us to create and find these groups.

This comes with its downsides. For instance, at the same time that it becomes easier for you to find blogs devoted to 18th-century ship-in-a-bottle designs, it becomes easier for me to find backwoods militias. The example Shirky gives here is a web bulletin board devoted to encouraging anorexia among its teen members. (This was the only part of the book that actually horrified me.) In the real world, these sorts of groups succumb to social pressure and go into hiding. The web makes it possible for them to find one another; they are no longer alone.

Shirky only gives brief treatment to these groups, and seems generally in favor of them for the same reason that people favor free speech: it protects the speech we hate as well as the speech we support. I would have liked deeper coverage here. In a lot of senses, the Internet is making us reconsider the foundations of democracy: now that we’re face to face with the consequences of truly free speech, what do we do about it, if anything? Do we still stand by the free-speech absolutism that we clung to when it was more or less hypothetical? Shirky doesn’t really touch on this.

He’s quite often a techno-idealist, which is a stance he assumes professionally. As a technologist, he’s convinced that the spread of cheap communications technologies will allow protesters to connect and topple ruling elites; he uses protests within Belarus as an example. He doesn’t really follow this up with counterexamples: Great Firewall Of China, anyone? More to the point: politics will exist even after text messages amongst flashmobs are a faint memory. I’d have liked this book better had Shirky cowritten it with a political scientist.

At one point Shirky’s love of technology seems to lead him to a non sequitur. He’s just told us about how protest worked in Germany before the fall of the Berlin Wall. The protesters, he says, learned to start their rebellion in some tiny way that slips under the government’s radar, and grow slowly. By the time they got big enough to cause the government any harm, it was too late and down fell the wall. Continues Shirky:

The lesson for protesters … was that they should protest in ways that the state was unlikely to interfere with … The lesson for repressive states was the opposite: don’t even let small protests get started … These two lessons set up a cat-and-mouse game between protesters and the protested institutions that continues to this day. As in everything that involves coordinated action, social tools have changed the balance of power in this game.

A “cat-and-mouse game” doesn’t sound to me like a change in “the balance of power” at all. That sounds to me, in fact, like a classic arms race: one side is temporarily victorious until the other side evolves a response.

Had Shirky dug into this a little more, the whole tone of his book would have changed. Had he scaled out his historical perspective, he might not be as optimistic either. I’ve been reading about the revolutionary potential of technology at least since I started using PGP; it was supposed to have been used by freedom fighters in the jungles of Burma. This strain continued through O’Reilly’s publication of its collection of essays on P2P. Within there were essays on, say, FreeNet, which was designed to create a censorship-proof peer-to-peer network. Only the occasional voice was brave enough to ask whether FreeNet would even be permitted within a repressive regime. If Shirky were interested in convincing me that technology might topple existing power structures, he’d go ask how those freedom-fighters are doing.

Shirky’s is a valuable point of view, but it’s a point of view that I’ve heard too many times. Nowadays, it’s more courageous — and ultimately, I think, more helpful to the world — to write a book disagreeing with Shirky (Google and the Myth of Universal Knowledge, say, or The Cult of the Amateur) than it is to write Here Comes Everybody.

Democrats are better for the economy, says Larry Bartels, and they’re better for the poor. He backs this up with an arsenal of data on rates and causes of inequality over the last 50 years under Republican and Democratic presidents. Inequality systematically increases under Republicans and decreases under Democrats. Bartels doesn’t linger much over the mechanism which might make this true; he hypothesizes that Republicans emphasize inflation-lowering policies that help mostly businessmen, while Democrats fight unemployment that largely afflicts the poor.

So if Republicans are bad for the majority of us, why do they win elections? A good part of the answer, says Bartels, is that Americans have short memories: they respond much more intensely to economic gain in the year right before an election than they do to economic loss in the preceding three. And American political opinion suffers from an unfortunate inconsistency: people claim to be in favor of reducing inequality at the same time that they support policies which further it. No matter how you frame it, for instance, Americans have overwhelmingly supported ending the estate tax since the 1930’s, even though it demonstrably only affects the wealthiest 1% or 2% of the population. And this inconsistency doesn’t go away with education: virtually every way you cut the data, clear majorities support doing away with the tax on inherited estates.

Why, then, did the inheritance tax persist until President Bush and a Republican Congress took control? Because ideology trumps the popular will much of the time. Democrats managed to keep the estate tax around until the rare combination of circumstances that allowed it to be overthrown. Bartels’s analysis shows that ideology persists as an explanation, even after other factors like voter wealth are removed. Ideology matters. If this elicited a “well, duh” from you, then you and the author of this review have something in common. More along these lines below.

Bartels’s statistical tables contain both estimates and standard errors, and much of the time the standard errors dwarf the estimates. That is, there’s more noise than signal. Often that’s because his statistics rest on small numbers (e.g., the set of all presidents from 1945 to the present.) And yet the graphs used to illustrate the tables are notably lacking in error bars: they’re just confident straight lines. I’d have much more faith in the book’s statistical conclusions if Bartels were more careful to point out fuzziness.

That said, a good many of the relations in the book do seem pretty cut-and-dried, like the relationship between an incumbent party’s popular-vote margin and income growth in the year before an election. It is striking how reliably the latter predicts the former, and how unreliably cumulative income growth (that is, income growth over the full four years of the incumbent’s term) does so. If nothing else, Bartels’s book has pointed out some patterns that have raised eyebrows and will certainly drive a lot of research.

Unequal Democracy is, unfortunately, a highly academic book: it seems very concerned to establish ideas rigorously that the rest of the world has long since taken for granted, out of the sheer analytical joy of doing so. Thus we wait 250 pages to see Bartels announce: “I find that senators in this period were vastly more responsive to affluent constituents than to constituents of modest means.” This is why we pay political scientists the big bucks. And yet to read Bartels, political science as a discipline only understands democracies as a collection of autonomous equals. So Unequal Democracy constitutes an advance. So much the worse for political science. I have my doubts that anyone outside of political science will get much from the book.

In particular — once again, assuming Bartels has summarized the literature propertly — political science seems to have missed out on the collective-action problem in economics. As Mancur Olson noted in The Logic of Collective Action in 1965 (and I don’t think he was the first), there’s a problem when policies stand to benefit one group while they spread their harms across the whole population: the group will lobby intensely for the policy, while the rest of the population stands mute. Compact interest groups are really important, if only for this reason. Yet Bartels doesn’t even start to discuss their effect on policy. He also never stops to touch on the disfranchisement of the poor. This was a large part of The Conscience of a Liberal: Krugman asserts that our nation’s growing inequality stems in large part from weakened labor unions, which used to help bring the poor to the polls.

In short, Bartels is looking at the American political scene from a high statistical level, never descending to the foundations. And his book will not help us change the situation.

In a world where The Conscience of a Liberal and Paul Farmer exist, I can’t recommend Unequal Democracy.

Behavioral economics has gotten a bum rap for a while now. Perhaps the bummest of them all was Richard Posner’s dig in the middle of Frontiers of Legal Theory :

Behavioral economics is defined by its subject rather than by its method and its subject is merely the set of phenomena that rational-choice models (or at least the simplest of them) do not explain. It would not be surprising if many of these phenomena turned out to be unrelated to each other, just as the set of things that are not edible by man include stones, toadstools, thunderclaps, and the Pythagorean theorem.

The essence of this objection is that behavioral economics, for all its charms, isn’t a coherent set of ideas at all and serves mainly destructive ends. If that was ever true (hint: it wasn’t), it can be safely ignored now that Richard Thaler and Cass Sunstein have published Nudge. Nudge’s purpose is to use our understanding of Man As He Is to build better policies. Rather than assume a perfectly rational human who can parse long, complicated documents with his mighty, limitless brain, Man As He Is sometimes skims and can be deceived by cleverly worded contracts. Man As He Is is often aware of his own limitations: he’ll flush his cigarettes down the toilet to prevent his future self from doing what his present self knows to be harmful; he’ll promise to start exercising tomorrow; and he’ll curse himself for procrastinating. Perfectly Rational Man — whom Thaler and Sunstein call an “Econ,” to be contrasted with a “Human” — would never have these problems. Econs sit down with (notional) pencil and paper and calmly work out the costs and benefits of all available actions, then take the action that maximizes their present and future happiness subject to a discount rate (future happiness is worth less than the same quantity of present happiness). They don’t have an internal procrastinator at war with a rational planner, nor do they ever regret on Sunday morning what they did on Saturday night.

Nudge is for Humans, not Econs. Nudge realizes, for instance, that making 401(k)s opt-out rather than opt-in, and setting a reasonable default investment plan, will lead lots more people to save money for retirement. And now that they’ve been enrolled, very few people will opt out. This is what Thaler and Sunstein call “libertarian paternalism”: giving people a gentle push in the direction of their own best interests (the “paternalism” part), but never taking away choices (the “libertarian” part). People can quit at any time; it’s only the default that has changed.

Your 401(k)’s default investment plan is part of what Thaler and Sunstein call “choice architecture.” As a 401(k) administrator, I can guide your choices in any number of ways. I can choose opt-in or opt-out; if I choose opt-out, I have to choose a default plan, whereas if I choose opt-in, I have to decide how much prodding to give you. The point is that choice is inevitable. There’s no way to avoid structuring the options available to people, so the right thing to do is to pick the best default. Given this realization, most of Nudge will be entirely uncontroversial.

Thaler and Sunstein digest a mountain of psychological research and reassemble it into a convincing story about how to build policies that correct for human failings. Humans can be expected to make the right decision when faced with a routine, concrete problem — buying food at the grocery store, say — but all bets are off when we’re asked to evaluate a complicated, large-scale problem like the impact of our air-conditioner usage on global climate change. Thaler and Sunstein want to give the market itself a nudge here. They wouldn’t insist that we buy only low-power appliances. Instead, they want our appliances to give us simple, immediate feedback on our energy usage: thermometers that reveal moment-to-moment energy costs, say, and EPA fuel-economy infographics that use easy-to-understand metrics like “dollars per year.”

Econs may be able to consume any information thrown at them and correctly render a judgment from what they read; Humans have finite attention spans and would rather spend time with their families than pore over fuel-economy tables. If we want Humans to make the best choices, we have to structure their choice environment to make this possible. Nudge is Thaler and Sunstein’s brilliant contribution toward this goal.

P.S.: Here would also be the place to recommend Thaler’s absolute gem of a book, The Winner’s Curse. If you’re not sure whether you have time for a full book, and would like to get a taste first, I can’t commend your attention strongly enough to “The Law Of One Price In Financial Markets.” It’s one in a series of essays from the Journal of Economic Perspectives entitled “Anomalies”.

Microfinance is most famous as microlending, whose most famous representative is Bangladesh’s Grameen Bank. Grameen, and its founder Mohammad Yunus, won a Nobel Peace Prize in 2006 for their aid to the poor. The idea, with which most people are probably familiar, is that the bank loans some of the world’s most destitute people small amounts of money — $100 or less, typically — for some vital bit of capital. Borrowers might use the money to buy a sewing machine, for instance, which they can then use to produce far more clothing than they had produced by hand. Grameen’s default rate has been remarkably low — “the poor always pay back”, to use the phrase from Grameen II.

The economic logic here is actually revealing as a study of what’s unspoken in economic logic, hence how misleading economic postulates are. “All else being equal” (such a magical phrase), the first bit of capital that I get will yield more benefits to me than the second bit. Assuming I’m rational, I will spend the first money I get on more-productive capital, then spend subsequent bits on less productive capital. That is, the marginal returns to capital are decreasing (or at least nonincreasing). Hence, if I’m a rational bank and all else is equal, I should be more willing to lend to the poor than to the wealthy: I’ll get a greater return from lending that little bit of capital.

Needless to say, that’s not how it works: Citibank is in no rush to lend to Bangladeshi farmers. Why not? Obviously it’s because all else is not equal. Among many other things, Citibank relies on the vast infrastructure provided by advanced capitalist economies: before they loan to me, they check with credit-reporting agencies that have a special competence validating people’s reputations. Those credit-reporting agencies can follow me around because I was born with a number, namely a Social Security Number, which I can’t escape from without some work. Hence the infrastructure beneath me makes it hard for me to default on a loan without other banks noticing. This infrastructure is missing from Bangladesh. Consequently, the cost of gathering all the necessary information about a loan applicant is much higher — transaction costs per dollar of loan are astronomical if the loans are administered in the way that Citibank specializes in.

Grameen handles this in a novel way, for which they’re justly famous. It’s called “group lending”: in Classic Grameen, they loan to groups of five people. If any one of the applicants defaults, the others are forbidden from ever receiving loans again. The informational burden is transferred from the bank onto the applicants.

Can’t those five people conspire to default on loans together? Yes, they surely can, and here we run into another difficulty of the classic economic picture. If they cut and run on a loan, they could run to another microlender and get another loan — and so on for as long as they want, so long as the microlenders don’t share information. The more microlenders that service a given area, the more challenging this problem becomes. So competition actually works against microlenders here, by making collusion possible. To solve this problem, microlenders need a set of institutions that make validating reputations less costly. Credit-reporting agencies would help, as would the whole arsenal of Western identity policies. Which isn’t to say that those are the only systems that will solve microlenders’ problems, by any means; just as group lending is a novel approach to the developing world’s specific problems, so we might expect them to land on different solutions to the reputation problem.

The Economics of Microfinance is filled with interesting discoveries like this. It starts with a less-developed form of microlending, namely the Rotating Savings and Credit Association, evolves through group lending, and discusses where Grameen and its ilk (BRAC et al.) are today. Most interesting for me was microsaving, as opposed to microlending. The poor often need savings accounts more than they need loans. Indeed, they are willing to receive negative interest rates on their money, just to ensure that the money stays in a safe place. Armendáriz and Morduch give a remarkable example: in certain rural villages, savings collectors will offer to take money out of the villagers’ hands, hold it for a time, take a fee, and return the now-smaller pile of money. Presumably this negative interest rate is less negative than the alternative, namely theft or neighbors begging for a loan. Microsaving is most often used to keep money away from husbands, according to Armendáriz and Morduch. Indeed, microfinance generally is most associated with rural women; they constitute an overwhelming percentage of Grameen’s (and other microbanks’) client base.

By the end of the book, however, it’s not clear that anyone can quantify the value of microfinance programs. Would those who participate in microfinance have done just as well without it? To gauge the actual impact of microfinance, one needs to answer that sort of counterfactual — which is, for obvious reasons, difficult if not impossible. There’s also a problem of what we’re modeling: if we’re trying to quantify, say, small-business growth before and after the introduction of a microfinance program, that’s one thing, and is relatively easy to answer. If we’re trying to measure empowerment of women, that’s quite another, and it’s not at all clear that we even know how to start measuring that. Should we measure it, for instance, by the rate of reported domestic violence? Empowerment may increase reporting rates. It may also cause a shift in the balance of power at home, which may increase violence.

The difficulties are manifest, as Armendáriz and Morduch are well aware. The great virtue of this book is that it doesn’t shy away from pointing out areas of ignorance and future challenges. Anyone interested in how microfinance actually works — and how one would actually measure its success — cannot avoid reading this book.

First off, I’m going to move these exercise posts somewhere else — probably exercise.laniels.org, if I can figure out how to make WordPress not display certain posts on the front page. They don’t seem to fit nicely here.

In the meantime, on to a quick report:

Yesterday: ran my now-standard route in about half an hour (I wasn’t keeping track). Got back home, did some stretches, and ended up approximating that yoga pose where you bend at the waist and put your arms behind your legs. I ended up touching my forehead to my knees, which I’ve never done before. Must be I was really limber after running.

Today: my day off. I’ve felt exhausted all day. I’m going to try to finish The Economics of Microfinance, then go to bed early, so I can …

Tomorrow: wake up early, go running, then get to the club in time for astanga yoga at 8:30 a.m. Then do weights if I still feel energized. Class starts at 8:30 … means leaving my apartment at 8:25 after changing into a clean t-shirt … which means getting to the apartment at 8:20 … which means starting the cooldown at 8:10 or so … start running at 7:30 … wake up by 7. Yeah, I can dig that.

P.S.: Two out of three isn’t bad. I ran and did a bit of weights (three sets of lat pulldowns: 110, 110, 95 lbs; two sets of tricep pulldowns at 40 lbs and one set at 45), then waited around for people to show up for the 8:30 yoga class. I got bored of waiting, so I left.

Running was a bit tougher this time than previously, because I neglected to walk around for ten minutes before running (I did more like two minutes), and didn’t stretch. Pre-stretching would have helped a lot. Also, I think the socks I was wearing were a bit too thick, so my feet were moving all around. Hence the blister on my left instep was bugging me the whole way. So: next time I run, I’ll stretch more before and wear thinner socks. Done and done.

I went home after a late lunch to work out, and felt incomparably better at work afterwards.

• On the elliptical: five rather intense minutes just to get warmed up.
• Tricep pulldown: 40 lbs., three sets of ten.
• Leg press: two sets of ten reps at 240 lbs., then one set at 260.
• Leg extension: three sets at 95 lbs., if memory serves. (I need to carry a little notepad in with me.)
• Lat pulldown: 110 lbs. Almost three full sets of ten reps, though the third set was really hard and was more like 8/10 of a set.
• Vertical chest press: 2.8 sets of 10 at 65 lbs.
• Weather: all indoors, sadly, in the interests of time. It was spectacular outside, so I would have preferred to run.
• Clothing: the usual. No socks, though; I forgot to bring them.

Doing this one from memory; it happened a few hours ago:

• 15 minutes on the elliptical. If I don’t have something to distract me, like an iPod, I get so bored on this machine. With some distraction, I can do 20 or 30 minutes. But I made these 15 rather intense minutes, anyway. I think I’ll go running tomorrow or Wednesday; running is way more interesting.
• Tricep pulldown: 40 lbs., 3 sets, 10 reps per set.
• Leg press: 240 lbs. 3 sets, 10 reps each.
• Leg extension: something like 90 lbs. 3 sets, 10 reps each.
• Lat pulldown: 100 lbs., 3 sets, 10 reps each. (Maybe 105 lbs., actually)
• Vertical chest press: 1 set of 10 at 65 lbs., 1 set of 10 at 80 lbs., and 1 incomplete set of 10 (maybe 5) at 65 lbs.

By this point my arms were really, really tired. I tried to do pushups and managed four before totally crapping out. I can do 20 normally without much effort.

• Clothing: wind pants, white t-shirt.
• Weather: the finest indoor conditions money can buy.

I am philosophically aligned with Medawar, which means philosophically aligned with Karl Popper. Pluto’s Republic is a collection of essays in Popper’s mold, which is to say: scientific, realistic about how scientists do their work, and opposed to pseudoscience of all kinds. Medawar has even picked up Popper’s hatred of psychoanalysis, and gives anti-Marxism in the Popperian style a healthy go.

Medawar’s heart is in the right place, and he’s less astringent — less academic, perhaps — than Popper. He has great faith in the power of science to cure the world’s ills, and conversely has a deep hatred of pseudosciences that purport to solve problems without submitting their claims to rigorous examination. The scientific life, properly conceived, is the one we all should aspire to: constant self-critique and constant piecemeal explanation. Science is an endless sequence of “conjectures and refutations,” to use Popper’s phrase.

That’s one of Medawar’s repeated critiques: that a long string of philosophers has misrepresented how science works, aided and abetted by polished scientific papers. This polished picture convinces us that we can understand the world free of theory: just gather a lot of data, look at it with an uncritical mind, and voilà : out of the data’s forehead springs the goddess Theory, fully formed. Whereas if you watch how science actually happens, says Medawar — if you actually listen to chatter in the lab — you’ll hear a temporarily plausible conjecture first; this conjecture drives the experiments. Scientists then gather data and either refute or temporarily confirm their story. Onward we go, haltingly, provisionally, in a piecemeal fashion.

So induction, as it may naïvely be imagined, is a non-starter as a description of how science works. But the untruth of induction doesn’t imply the uselessness of scientific method. It just means that we’ve misunderstood what that method is about. Science, with all its flaws, is the only legitimate candidate we know for demonstrably progressing toward a fuller knowledge of the world: guess, observe, test, repeat.

On a few occasions Medawar attacks those who see rampaging technology as a great evil in the world; here I think he falls off the rails a bit. Technology may cause great evils, says Medawar, but it is also the only solution for fixing its own mistakes. Those who’ve watched fertilizer runoff destroy the Gulf of Mexico may be skeptical that what we need to fix it is more technology rather than a whole new attitude toward agriculture; likewise, is there really a technological solution to the problem of nuclear weaponry? Medawar doesn’t go into enough detail about exactly what he means in this context. In general, he’s on the weakest terrain when he ventures outside his zone of professional competence; I found his general social commentary and economic philosophizing (e.g., why economic predictions are inherently less accurate than meteorological ones) fatuous.

Overall, I think there’s a much better book waiting inside of this one. A more courageous editor would have deleted perhaps 1/3 of the essays, or would have rearranged them into new, less overlapping ones. As it is, Pluto’s Republic fuses two of Medawar’s earlier books without apparently deleting much from either. And it suffers from the disease (which, I’ll grant you, some don’t consider a disease) of many British historians: I envision the writer ambling around the lecture stage, cigarette in hand, thinking out loud to himself while his bemused students try to copy down what he says and put it in some coherent kind of order. This book is what happens when someone types up those lecture notes.

So I’d recommend against reading Pluto’s Republic, when many other fine books in the same vein are available. I suspect Medawar would agree with me that you should run out and read Popper’s Open Society and Its Enemies at your earliest convenience if you’ve not already. After that, Pluto’s Republic feels like rambly footnotes.

I routinely have to copy files from one “environment” to another, where an “environment” is a set of computers that can see each other but can’t see all the other hosts in the world. In particular, computers in one environment very often can’t see computers in another. Each environment typically has a gateway machine: you connect to the gateway, which can see all the other hosts in the environment. My desktop machine can see all the gateways, but the gateways typically can’t see one another.

Copying files amongst those environments is tricky. Suppose I want to copy a file from a machine inside environment A to one inside environment B. I need to copy it out from the inner machine in environment A to A’s gateway, then copy from A’s gateway to my desktop, then from my desktop to B’s gateway, then finally from B’s gateway to the inner B machine. The number of bytes going over the wire here is four times the size of the file we’re copying. Then you have to remember to delete all the intermediate copies. It’s a hassle.

Doing all of this with tar and ssh is actually really easy. You’d just do this from your desktop machine:

ssh gatewayA ssh innerA tar cz /foo/bar/baz \
| ssh gatewayB ssh innerB tar xz /path/to/unpack/

This combines three little tricks:

1. ssh can accept piped input. When it sees that it’s getting something on standard input on the local machine, it dutifully dumps it to standard output on the remote machine.
2. tar is the universal awesomeness. All it does is pack up a set of files into a single file. If you don’t tell it which file to pack into (which you would do with the -f argument), it packs those files to standard output.
3. The -z option to tar will gzip tar’s output.

So all told, the command above does the following:

1. Connect to A’s gateway.
2. Connect to the machine inside environment A
3. tar up the files in /foo/bar/baz, compress them, and dump the compressed stream to standard output.
4. Redirect that standard output into an ssh chain where you’re first connecting to B’s gateway, then connecting to B itself.
5. On B, unpack the zipped tar file. Note that when we tarred it up in step 3, tar stripped off the leading ‘/’ from all the files. Hence they will be unpacked on B to /path/to/unpack/foo/bar/baz. If you want them to be unpacked into /foo/bar/baz, use Pcz rather than cz in the first tar command.

There are no files laying on disk after this is done, and the number of bytes going over the wire is significantly smaller (because of the -z option). Extending this to the case when innerA and innerB are way inside a chain of isolated environments is straightforward.

Wrapping this all up into a script wouldn’t be hard. It’s a question of how to set up the interface. Here’s a possible way to interact with it:

remcopy innerA:/foo/bar/baz innerB:/path/to/unpack \
--source-gw gatewayA --dest-gw gatewayB

but at that point you’ve not saved yourself any keystrokes. You could save keystrokes by using smaller argument names: -s, say, rather than --source-gw. If we leave off /path/to/unpack, we should default to ${HOME} on innerB. Other than that, I can’t think of many ways to shorten this.

Finally, rsync may well already have this. If I’m reinventing the wheel, do let me know.

Saw it yesterday. It’s a fine film, particularly in the genre (broadly, “family movies”). I don’t get why it got such a sucky Tomatometer reading.

“Consensus: Flounders under an implausible storyline and simplistic stock characters.”

Um? This is a story about a girl who can travel to an enchanted world just by reading books. And it’s about the agoraphobic author of those books suddenly developing bravery and flying off to a South Pacific island to save the girl. It’s a fantasy movie. Seems frankly retarded to object to it on the grounds that it’s implausible.

As the years go by, I think more and more that Roger Ebert’s dictum is true: it’s not about what a movie is about; it’s about how the movie is about what it’s about. Were the filmmakers aiming for a fantastic, implausible story? Yes. Did they deliver on that? Yes. Did they do it stylishly? Yes. Will it appeal to the children who are its intended audience? Yes. Will the adults accompanying them be bored? Not at all. I give it a solid B, if not an A-.

A friend mentioned to me his self-described irrational fear of flying the other day, which inspired me to look up rates of death from automobiles and airplanes. Turns out that 1,921 people died in “Water, air and space, and other and unspecified transport accidents and their sequelae” in 2004, and that around 45,000 died in motor-vehicle accidents.

But that’s not quite the measure I’m looking for, if I’m trying to quantify whether fear of flying is irrational. The thing to look at is probably deaths per passenger-mile. For that, we turn to the DoT’s statistics, which show that the per-passenger-mile rate is actually about 50% higher for airplanes than for cars.

This contradicts most of what I’ve heard about the relative risks. Can anyone help me figure out where the disconnect is? Is it just that few people have been looking at the right risk measure?

I’m going to ape my anonymous friend littlepurple and write about my workouts. Maybe not as diligently as she does, but at least sometimes.

Today’s workout:

• 30 minutes to run a loop from my house, down Magazine Street to the River Chuck, over to the Mass. Ave. Bridge, and back up Mass. Ave. to my house.

• 105-lb. lat pulldown. 3 sets of 10 reps.

• 40-lb. triceps pulldown (not really sure what the exercise is called). Tried to do three sets of ten reps. Got to two sets of ten, plus a third set of 7.

Things to work on, notes, etc.:

• I need new socks. As I sit here, I see that both my insteps are incipient blisters. Today will be a sock-buying day. It will also be an “ask the kind people at Marathon Sports what to put on my feet between now and Monday so that I can go running without killing myself” day.
• I need to work on stretching before I start running. As it is, I start slowly because I’m not loose enough. I was hoping to get to the Boston Sports Club in time for a yoga class today, but that didn’t end up happening.
• I was hydrated properly, which was good. I waited a while between drinking coffee and running, but I should probably just skip the coffee pre-run. In practice, I think this means that I should go running first thing in the morning.

P.S.: Littlepurple says that it’s valuable to write down what I was wearing, the weather, etc. So here goes:

• Clothing: t-shirt from a standard ten-pack of white t-shirts (I am not exactly a professional at the workout gear); wind pants with something like a felt lining (designed for winter, I think); ITA zip-up hoodie (which I unzipped about halfway through the run).
• Weather: perfectly blue sky, probably in the fifties. It rained earlier today, so there were lots of puddles to skirt around.

I saw “Ring of Fire” last night. It’s in the same vein as, I gather, “Movin’ Out” or that one ABBA musical (MAMMA MIA, I guess). The trick with any of these things is that the audience has the original musician’s tune running through its collective head, so the performers have to be really careful not to sound too different from what the audience expects.

In “Ring of Fire”’s case, the fundamental trouble is that only one of the actors can even approach Cash’s bass. A couple of the ladies have pipes approaching June Carter Cash’s.

If you can keep the original out of your head, you should definitely go see “Ring of Fire.” Today I’m back at work listening to the originals, and I’m realizing just how much was missing from last night’s show.

I don’t know why, but I felt compelled.

Thanks to Chris Rugen for rocking the PhotoShop on my behalf.

The following is really no more than reheated Karl Popper, I think, but it seems worthwhile to note anyway.

Merely knowing what’s wrong is immensely valuable, and would be valuable even if we never learned what was right. I can know that the claim “Steve Laniel is eight feet tall” is false even if I don’t know exactly how tall I am. I would be more than happy with a set of sciences — from the mathematical layer, up through the physical layer, on up to the economic layer — that merely told me what not to believe.

Popper’s observation was that there are a lot of asymmetries like the asymmetry between knowing what’s true and what’s false. For instance, “maximizing happiness” (à la the utilitarians) is not the same thing as minimizing suffering, and the latter is much more vulnerable to scientific investigation than the former.

Imagine the good that would be accomplished if we could eliminate bad reasoning, even if we never actually introduced good reasoning. I don’t insist that people be good statisticians, for instance — just that they not commit basic probability errors of the sort that one finds corrected in How To Lie With Statistics. I don’t insist that they be good logicians — only that they not confuse necessity with sufficiency (e.g., “All terrorists are Arabs,” even if true, does not imply that “All Arabs are terrorists”; that revelation alone would delete perhaps 10% of all stupid web commentary).

Of course, the objection to all of this is that there’s no way we could know what’s true, so the only thing that’s even possible is to get rid of false ideas — hence that what I’m asking for here is no more than what we already do. The history of science itself is just the history of correcting mistakes. The theory of relativity isn’t true; it just describes more phenomena accurately than Newtonian mechanics did; and where Newtonian mechanics models a phenomenon well, relativity doesn’t model it poorly.

I’m just envisioning knowledge stated in negative terms, e.g., “We don’t know what a man’s mind actually is, but we do know that he doesn’t maximize his subjective expected utility.” Or “We may not know all the details of how one species evolved into another, but we do know that creationism doesn’t even count as an “explanation”. (My negative curriculum would teach students to understand what “begging the question” means.)

One might proceed from here, in the mode of Sherlock Holmes, to the idea that eliminating all the false ideas necessarily leaves only true ones behind. As a matter of pure logic, that’s probably true. But as a matter of practical discovery — knowing when we’ve eliminated false ideas — it’s not valuable.

I do occasionally have dreams of model-theoretically limiting the space of permissible scientific models: we know that the truth can’t be X₁, and we know it can’t be X₂, and so forth, which means that the space of permissible models is constrained. I don’t know how much of a constraint all the known false models place on the models that are available to us; Cosma Shalizi recommended Patrick Suppes’s Representation and Invariance of Scientific Structures to me along these lines 3+ years ago, but I’ve not read it because I am a terrible person.

I should mention on here that I made it a New Year’s resolution to work out and lose some weight: I was developing a bit of a pudge, which doesn’t make anyone happy (least of all me, or … um … my girlfriend). I was hesitant for a while to say anything about it, just because New Year’s resolutions have a way of going unfulfilled. It’s been three full months now, though, and I’ve gone to the gym on average every 2.5 days or so. I’ve lost some weight and developed a lot of muscle. It’s gotten to the point that when I don’t go, I feel terrible. After a day spent at a desk without even the vaguest notion of sunlight or motion, I need to spend an hour or so beating the shit out of an elliptical machine and some weightlifting implements. Without that, my skin feels puffy from a day spent in dry, recycled air.

I’m particularly taken with this one exercise that’s mostly designed to work my triceps. You reach 18 inches or so above your head, grab a bar, and pull it down to chest level. Then you lower it as far as your arms will go and return it to chest level. By not returning it to its starting position, you’re maintaining some resistance throughout the exercise; the end of a rep does not mean the end of resistance. I both love doing this exercise and live in atavistic fear of it: it is really hard, and delivers the most concentrated body destruction of anything that I do at the gym. Often when I’m done with it, my muscles are too tired to lift my shirt over my head.

The great thing about that exercise, though, is that it’s surreptitiously working out my core muscles (abs and obliques, mostly). In order to do the exercise successfully, I’m forced to breathe properly, clench my abs as hard as I can, and keep my entire body apart from my arms perfectly motionless. I can’t exactly explain why doing these things is necessary, other than perhaps that moving those other muscles dissipates energy that ought to be going into my arms. [1] If I clench properly, my abs get a great workout. When this happens, I admit that I turn into a narcissist in the locker room afterward. I have broken several mirrors.

So if you see me and you notice that I need NEW PANTS, this will be why.

Having strengthened some of my core muscles, I’m hopeful that I’ll be able to do more yoga, and be able to balance better when I do. I need to be doing more yoga anyway, if I continue with this muscle-growing revolution; I don’t want my already limited flexibility to disappear entirely.

[1]: The other day when performing this exercise I was reminded of a section of The Physics of Baseball where Adair explores the “sweet spot” on a baseball bat. That spot is the place where the minimal energy is dissipated into the rest of the bat, hence where the maximal energy is conveyed into the ball. This happens to correspond to a node of the bat’s vibration; that is, if you hang the bat upside down from a long string and give it a strong tap, the sweet spot is the point on the bat around which it will oscillate. The worst place to hit the ball is an antinode of this oscillation; that’s where the most energy is dissipated into the bat. That’s why hitting a baseball improperly really hurts your hands, and why it often leads to broken bats.

(Attention conservation notice: 1100 words on Saul Bellow’s final novel, which was a meditation on philosophers, writers, and how to bridge the divide between their respective takes on the world. At the same time, it’s a beautiful story about two close friends. It’s in the pantheon of great Bellow novels.)

Ravelstein is a few things at once. It is Saul Bellow’s sweet farewell to his longtime friend Allan Bloom, and would be worth reading for that aspect alone; I’ve never read such a loving ode to a friend, nor such an erudite ode to the virtues of friendship itself. It is a continuing exploration of some of the ideas about modern life from Bellow’s earlier books. It may well be a look, à la Philip Roth, at the boundary between the artist and his narrator, though I’m not sure, and it’s not at all clear that it matters here as much as it matters in a Roth novel.

Ravelstein’s namesake, Abe Ravelstein, is a very thinly veiled Allan Bloom. He’s just written a very important, very popular book (a thinly veiled Closing of the American Mind) that has made him wealthy beyond his wildest dreams. Ravelstein is a disciple of a certain Felix Davarr, who’s a thinly veiled Leo Strauss (Leo for lion, Felix for cat; not sure about Strauss and Davarr). In fact the thin veiling of everything in this book may well be a hat tip to Strauss, who argued that the true meaning of many philosophical texts is hidden in plain sight, and that only initiates to the true way of reading the text could ever hope to piece it together.

In at least a couple books — Herzog and Mr. Sammler’s Planet — Bellow has given us a sad but ultimately redemptive picture of the United States: modern life is confused, people define themselves by their possessions, and we’re flooded with a million contradictory ideas that bring us no real understanding of the world. (The structure of society is given by the structure of production! No wait: society’s problems are caused by sexual repression! No wait: society’s problems will be fixed when unfettered capitalism allows individual talents to bloom! No wait…!) We’re rootless consumers of ideas. After a lifetime as a professor of philosophy, Moses Herzog’s life falls apart, and the best he can do to put some ground beneath his feet is to send letter upon letter to philosophers, living or dead. The irony of Herzog’s life is that his studies of philosophy, which were supposed to bring some order to the world, have not helped him at all when push came to shove.

Allan Bloom and Saul Bellow agree on this. Indeed, Bellow wrote the introduction to at least one edition of The Closing of the American Mind. Bloom’s critique of the university basically ran as follows: the university should be a monastery in which those rare souls capable of understanding true philosophy lived among their brethren. Since World War II at least, the American university has moved away from that priesthood; it has brought all the problems and confusion of the outer world within its walls. Where the outer world has given up trying to find any order in things, so the university has replaced thought with technology. Where the outer world venerates consumerism, so the university has invented the “business degree.” And so forth.

This can sound curmudgeonly rather than scholarly, and indeed that’s how The Closing of the American Mind sounded to me when I read it a few years back. With a few more books under my belt since then, Bloom sounds like Ruskin without the poetry. Having finished Ravelstein, though, I think I may need to go reread Bloom. Abe Ravelstein may not have experienced a dark day in his life. When he watches the world, he doesn’t see confusion; he sees the underlying order. When Ravelstein watches the narrator fall in love with one of Ravelstein’s students (again, a thin veiling of Bellow’s own life, in which his fifth wife [!] was a Bloom devotée), he sees it through Plato’s Symposium. Ravelstein’s life is one of pure joy and deep, continuous thought. His relationship with the narrator is one of constant, unflinching honesty. He lives the life of the philosopher.

Ravelstein spends most of the book living the life of the philosopher from a hospital bed. He is dying of AIDS. He urges the narrator to write a biography of him; this book is that biography, tinged with Bellow’s own thoughts on sex and death, and thick with notes on the difficulty of writing about such a brilliant man.

Both Bellow and Philip Roth have emphasized that the role of the writer is opposite to that of the philosopher or the political leader. The philosopher and the ideologue deal in abstractions; they cannot focus on the details. The novelist lives for the details; that’s what the novel is. In Bellow’s Nobel lecture, he talks about the novelist’s role in combating the modern malaise: we’re swallowed up by massive institutions, and we’ve become little more than numbers in a larger system. Some would say that this requires us to rethink who we are, and in particular abandon the novel: there are no characters anymore, they’d say, only statistical markers on a hard drive somewhere. How can the novel survive, they say, if it’s based on an outmoded view of the world we live in? Bellow insists to the contrary that if we’re going to save ourselves from the modern world, we have to do so by honoring those details more, by respecting characters more. Bellow’s own novels are testaments to that principle: amidst the whirlwind of modern urban life, his characters cannot escape the souls that they came in with. The characters of his characters will reveal themselves no matter what.

That’s why Ravelstein is such an interesting book, and Abe Ravelstein himself is such an interesting character. The book is a touching biography of Ravelstein, but it’s more about the friendship between Bellow and Bloom. It’s about the constant tension between the demands of the writer and those of the philosopher, and how those demands can exist peacefully side by side in the friendship of these two brilliant men. I think the message is that philosophy and the grubby details of existence must be kept in their proper balance. (I feel obliged here to quote from Doctor Zhivago : “I don’t like purely philosophical works. I think a little philosophy should added to life and art by way of seasoning, but to make it one’s specialty seems to me as strange as eating nothing but horseradish.”)

I found Ravelstein really touching. Bellow obviously loved Bloom dearly, and missed him a great deal after he’d parted. The last sentence in the book is “You don’t easily give up a creature like Ravelstein to death.” I feel the same way about Bellow.

P.S.: Copies of Ravelstein are available used on Amazon for a couple pennies. With shipping, I bought a paperback used copy the other day for $4.02. At that price, there’s really no reason not to buy it.

John “Torture Memo” Yoo is a law professor at Boalt. We now find that he wrote this sort of jaw-dropping stuff:

“If a government defendant were to harm an enemy combatant during an interrogation in a manner that might arguably violate a criminal prohibition, he would be doing so in order to prevent further attacks on the United States by the al Qaeda terrorist network,” Yoo wrote. “In that case, we believe that he could argue that the executive branch’s constitutional authority to protect the nation from attack justified his actions.”

Interrogators who harmed a prisoner would be protected by a “national and international version of the right to self-defense,” Yoo wrote. He also articulated a definition of illegal conduct in interrogations — that it must “shock the conscience” — that the Bush administration advocated for years.

“Whether conduct is conscience-shocking turns in part on whether it is without any justification,” Yoo wrote, explaining, for example, that it would have to be inspired by malice or sadism before it could be prosecuted.

This man teaches law? I’m horrified to know that this kind of teaching can make its way into any student’s mind. But I’m also just genuinely curious to know what sort of questions he gets asked, and whether he feels as though he can teach the kids infinitely flexible, infinitely callous sophistry. I would love to sit in on his class. This is the banality of evil, right at Berkeley. California taxpayers: your tax dollars subsidize it.

Anyone inclined to read this book should just ignore its subtitle — “How Judges, Bureaucrats, and Lawyers Put Innovators At Risk.” I can only imagine that was thrown in to make this book appeal to right-wingers. Which is really strange, come to think of it: the book is sedate, scholarly, reasoned, and exhaustive, and if anything it will appeal to readers who enjoy Larry Lessig and Richard Posner. In no way do Bessen and Meurer suggest that the system be scrapped and those “Judges, Bureaucrats and Lawyers” be thrown out. In the world Bessen and Meurer desire, “bureaucrats” will still examine patent claims; lawyers will still prepare patent applications; and judges will still hear infringement claims. Bessen and Meurer want to change policies here and there to make the system more efficient, but it will still be the same system by the time they’re done.

The dispute over patents has become rancorous in the era of the Internet, with lots of loud people asking that it be torn down, lots of other people claiming that the first group are unwitting pawns of big business, etc. Bessen and Meurer avoid almost all of that. Their claim is straightforward, and avoids all the parts of the debate that are unnecessary for establishing their point — which is simply this: that as a system of property, there are certain things we should expect of a patent system. Just as real property gives you the right to exclude ne’er-do-wells from your land, so a patent is a legal right to exclude people from developing a technology. And there are certain things that we should expect of patents, just as we would expect them from property law. First among these is notice: it should be easy to determine whether you’ve trespassed on my land, and just as easy to determine whether you’ve violated my patent.

In a system where notice is working perfectly, we’d expect very few patent-infringement lawsuits. In some fields, say Bessen and Meurer, that’s exactly what we see: the chemical and pharmaceutical industries have very low rates of infringement, because comparing your small molecule to a patented small molecule is easy. When the product gets more complex and abstract — as with a computer algorithm — deciding whether you’ve infringed gets more and more difficult, and the number of suits balloons.

Still, if lawsuits to defend against infringement claims were cheap enough, patents would be a net economic positive: the cost of lawyers when you accidentally infringe would be less than the money you bring in from non-infringing patented technologies. By this measure, Bessen and Meurer say that patents stopped being a net economic positive — outside of the chemical and pharmaceutical industries — in the mid-nineties.

In support of their conclusions, they use an arsenal of statistical methods based on the stock prices of public companies. Everyone knows the defects in using the stock market to infer knowledge, and Bessen and Meurer are not fools. Still, the most reliable data they could find were stock-price data, on which they conducted event studies to determine what value the market attached to particular patents and infringement lawsuits. Absent any data from lawyers or companies about how much they’re charging or paying, this is probably the best one could hope for.

Their second method of estimating value is to study rates of patent renewal. 20% of patents expire after four years; 21% expire after eight; 17% expire after 12; and 42% last the full 17-year term. Based on the cost of renewing, we can estimate how much the owner values the patent. Seeing that 58% of patents expire by their 12th year, and that $2,327 in renewal fees will have been due by then, we can estimate that the median patent is worth less than $2,500. Similar, but obviously more sophisticated, analyses are spread throughout Patent Failure; they constitute its bulk.

Bessen and Meurer blame the decreasing value of patents on a number of different factors. One is the rise of software and business-method patents, which are inherently abstract. They give a fascinating account of the Karmarkar patent, which covers a certain interior-point method for solving linear-programming problems. Years after Karmarkar obtained his patent, mathematicians discovered that the Karmarkar algorithm was formally identical to another that had been been in wide use for years. This sort of equivalence is much more likely for an abstract patent such as one governing a mathematical algorithm, and makes knowing ahead of time whether you’ve infringed that much more difficult. Abstraction harms the notice function of patents, hence harms their functioning as property.

They sidestep any number of other questions, such as the cost of “patenting around” — that is, developing new technologies only because existing technologies are all patented. Adding in these costs would likely only help their case, though one would have to crank through the numbers to be sure.

My only concern about this book is its subtitle. If I were to issue a second edition, I would drop the populist-sounding subtitle and rename it to what it really is: “A Calm Look At The Evidence.”