Author: Andrew Sullivan

An expat describes how she finally broke down and parted with her print-only loyalism:

I felt dirty watching the download bar turn blue, but once I began, I found I didn’t know how to stop. First came Alice Munro. The verisimilitude between Munro’s Southern Ontario and the Kabul expatriate community was striking. Both were a whorl of gossip. Both involved dinners parties made from canned provisions. Both were rife with what Munro called the “great shock of pleasure” that life affords you sometimes. And both were populated by the “wrecked survivors of the female life,” girls who grew up to be dissatisfied women, who found ways of negotiating with the rough terrain of an inherently male landscape. The small humiliations, the minutiae of everyday existence, the intensity of the social gaze—farm towns of rural Canada have much in common with this war zone capital.

Next came Junot Díaz’s short stories, then Zadie Smith’s novel, then George Saunders’ latest. I rationed out these texts as if they were wartime succor. Not only does downloading take a long time, spending $10 to $15 per book seemed slightly insane in a country where that amounted to an average worker’s weekly pay.

Once I tried explaining the concept of ebooks—that they cost more than three watermelons and cannot be lent to someone—to an Afghan friend and was swiftly ridiculed for my vanity. He never said as much, of course. But he did give me a look that Afghans often give me: a look that seemed to say, “Youuu eeeeediot.”

Frank Jacobs shares the story of François Élie Roudaire, a French Army captain who led the failed colonial campaign to create an inland sea in the Algerian desert. Roudaire proposed the project in 1874, after he came across upon a dry salt-lake bed, Chott el-Mehrir, that seemed strangely familiar:

Knowing his classics, Roudaire couldn’t help thinking that this submarine salt plain might once have been part of the fabled Bay of Triton. Described by Herodotus but unknown to modernity, the lake’s debatable existence and location constituted an Atlantis-type mystery popular among geographers. Could Chott el-Mehrir be contiguous with other chotts [dry lake beds] towards the Tunisian coast, forming the ghostly imprint of a former sea inlet? And could this semi-mythical body of water be resurrected?

With the blessing of Suez Canal engineer Ferdinand de Lesseps, Roudaire developed a plan to construct a 78-foot-deep, 3,100-square-mile inland sea. The project would cost a fraction of France’s Suez Canal venture, he assured investors:

The price tag [was] a mere 25 million francs. A small investment with a large return: the proof of France’s enlightened policies, progressive intentions and beneficial results in North Africa. “The Sahara is the cancer eating away at Africa,” Roudaire wrote. “We can’t cure it; therefore, we must drown it.”

But nothing came of the plan except dreams:

The London Times said that the plan “dazzles the imagination, yet it has a sufficiently substantial basis to satisfy several shrewd traders in African commerce and some distinguished engineers.” … This new sea “is sure to be frequented by trading vessels,” reported the British magazine All the Year Round, “to carry off the produce of its banks, which will eventually be dotted with groves of date and coconut palms. … Hotels, perhaps towns, will spring up on picturesque and eligible sites; luxurious house-boats will float in its most sheltered and shady creeks.”

(Image: Map of the planned sea. Note that Lake Geneva is offered for scale in lower right-hand corner.)

An expert on the subject writes in:

I hate that I have to partly defend Tony Bennett, but here goes. The education reform movement has sold politicians and the public what is, at best, the polite fiction of objectivity in school accountability. There’s some truth to Michael Petrilli’s flippant remark you quoted, describing these systems are “more akin to baking cookies than designing a computer,” and Bennett’s involvement is more the rule than the exception. I should know, I’ve worked in the kitchen.

The basic architectures of these systems are sketched out by elected officials, high-level education bureaucrats, and stakeholders (such as teachers’ unions or reform lobbyists). Discussions tend to focus on the traditional elementary-middle-high school model, and final designs are a compromise between what they’d like to measure and what’s actually available in the data.

Most of these systems are based on some combination of pass/fail percentage, graduation rate, attendance rate, and test score improvement or growth (Jeb Bush’s group has been particularly instrumental in pushing the latter as the solution to the snapshot-in-time “failure” of NCLB). The basic idea is that each measure has a value that’s converted to a score, and then all the scores are combined into a composite from which accountability identifications are made.

In practice, it’s not that simple.

There are well-known statistical problems with aggregated test scores from small populations, and every state sets minimum sizes for inclusion of a group. Different measures have different distributions: pass/fail rates almost always follow the classic bell curve, while attendance rates bunch at the high end with a long tail of “problem” schools. Combining measures requires making value judgments on balancing relative priorities.

Then states find schools that don’t have all the data to fit their designs, or don’t match the stereotypical model. What do they do with a K-12 school? How about a school with too few kids? Stretching the design to encompass more schools adds yet another layer of complexity and subjectivity. (This is where Indiana ran into trouble: they hadn’t accounted for charter high schools too new to have graduated any kids yet, so averaging what measures they had made this school’s score inherently worse than high schools with graduation rates, which often cluster toward the high end.)

Once states develop scores, they have to assign meaning to them. Specifying which scores fit what labels – either by defining ranges or translating scores to fit a particular rubric (like Indiana’s GPA) – is the most subjective and political part of the process. Decision-makers inevitably test designs against what they believe about particular schools. I guarantee that every state with this sort of school accountability system has an email chain like Bennett’s, with a top official questioning a favorite school’s score as too low.

Did Bennett cross an ethical line? He certainly sidled up to it, but whether or how far he actually stepped over depends on two things: First, how much Indiana’s system was modified to fit one school, as opposed to all schools with the same quirks. Second, whether there was a quid pro quo between the grade modification and campaign contributions Bennett received from the school’s founder (knowing how the reform movement operates, I suspect there was but doubt there’s any way to prove it).

So what does this mean? Major factions within both parties have bought into these accountability systems, even though most who champion them – including and especially Arne Duncan and his staff – don’t understand their complexity. Politicians and special interests misuse information that ought to identify trends and target improvements as ideological weapons. The media’s default position of blind credulity occasionally gives way to scandal-mongering, without any real attempt to understand the issue.

How’s that for a cheerful perspective on the future of American education?

Updates from readers originally posted in this post, in addition to an email from an Indiana school official, can be found here.

A shocking revelation:

The thing is: There’s not a lot of evidence to suggest that Proust really ate madeleines. Although his vivid memories of the delicate cookies from In Search of Lost Time have become iconic, early versions of the novel actually don’t include madeleines at all. Instead, we see Marcel biting into humble biscotte – a piece of dry toast.

What Proust really ate:

[A]s his illness worsened, his need to write began to subsume his desire to eat, and breakfast became Proust’s meal of choice. Instead of the madeleines and tea we known from his fiction, the real Marcel demanded croissants and cafe au lait, brought to him in bed while he read the paper and began his work. He would dunk his croissant in the coffee (just as his fictional self would mimic with a cup of tea) and ate little else for the rest of the day.

Cécile Albaret, Proust’s trusted servant, would later marvel at the writer’s ability to live on so little, after years of hedonistic eating. “The most extraordinary thing was how he could survive and work, ill as he was, … by living on the shadows of foods he’d known and loved in the past.” In the absence of beef and beer, Proust’s writing (and those morning croissants) was all that remained, evoking those tantalizing sense memories of meals gone by—his own personal madeleine moments.

(Image via James Brennan’s piece on Proust and madeleines)

Not so much:

This morning Procter & Gamble (PG), which rules the category with Mach-3-maker Gillette, said its razor sales are falling in developed markets. This followed yesterday’s announcement by Energizer (ENR) that unit sales of its Schick men’s razors have dropped 10 percent in the past year—a literal decimation.

Energizer blames the sales slide on aggressive promotions, specifically P&G’s. Meanwhile, P&G focused on its gains abroad and glossed over its losses in major markets. Euromonitor points to another culprit: “the vogue for stubble” and a “growing acceptance of the unshaven look in the workplace.” In other words: hairy dudes. And this is one market where China may not save the day; Euromonitor claims Chinese men are relatively “nonhairy.”