According to the article, this picture presents the largest component of reference class session co-listings. “Nodes represent references, node size reflects degree centrality, and more orange nodes reflect higher degree centrality. A tie between two nodes signals that two nodes have been co-listed in the same class session on at least two separate syllabi. Tie thickness reflects the number of syllabi on which two references were co-listed in the same class session.” Note that the unit here is articles, so authors may appear in different places in the figure based on different works of theirs.

Two things struck me about this. First was that the visualization is consistent with the field characterization in Marion Fourcade’s ABS piece from a few years ago—you’ve got the structural/embeddedness people and the broadly cultural/Zelizerian work forming one large group, and then (disconnected from both) the insurgent social studies of science/finance people. Second, though, was that the network is quite fragile. But, second, the big component in the network is fragile. If you deleted Geertz (1978), Granovetter (2005), and Swedberg (2001), then you’d have four separate components which you might crudely characterize as soc of finance, culture/Zelizer, Granovetter/network embeddedness/social capital, and Polanyi/political embeddedness. Moreover, two of the bridge pieces are more reviews than research pieces: the Granovetter 2005 is his JEP piece, I think, and the Swedberg piece is his “Sociology and Game Theory” paper, I believe. The Geertz paper (the Bazaar one) is a surprisingly tenuous bridge between the structural and the cultural approaches.

Another thing I’d be interested in seeing is the list of actual works the labels refer to—most of them I know unambiguously, but there are a few that are ambiguous (because the author published more than one thing that year) and I’d be interested in seeing which one is being counted.

Update: Duh, as Omar points out, this is a network where readings are tied if they are assigned in the same class session which makes the fragility interpretation go away. This is of course mentioned in the caption I quoted but evidently did not read properly. As he says,

If this was a co-citation network, then yes, the inference to “tribalism” follows. However, here a tie indicates that two readings are classified as similar by the relevant gatekeepers. So, I think that rather than giving you a picture of the socio-intellectual structure of the subdiscipline, this network simply gives you a picture of its cognitive or classificatory structure. So it is a good thing that the network is easily fragmented, otherwise economic sociology would be a classificatorily incoherent subdiscipline.

And of course Omar is correct here.

How, in less than a decade, did America go from being a country in which some states punished gay sex with criminal penalties to one in which the highest elected official in the land now champions the right of same-sex couples to marry? The answer can be found in the interaction between supporters of marriage equality and the Christian conservative movement over the past few decades. As late as the 1980s, same-sex marriage was on virtually no one’s radar screen. … It happened like this: In 1993, in the case of Baehr v. Lewin, the Hawaii Supreme Court decided that the state’s prohibition on same-sex marriage was discriminatory. In 1998, Hawaii’s voters passed a referendum giving the legislature the right to declare same-sex marriage illegal, but in the meantime, social conservatives had taken the issue to the national stage, where it promised to pay handsome dividends. Same-sex marriage was still so unpopular that in 1996, tremulous Democrats joined Republicans in overwhelmingly passing the Defense of Marriage Act, or DOMA, signed by President Bill Clinton. … DOMA was thus a preemptive strike by the opponents of marriage equality.

But the act helped to call into being the very marriage equality movement it aimed to combat. Encouraged by their surprising, if temporary, success in Hawaii, and outraged by the blatantly homophobic arguments that had been made in favor of DOMA, the lesbian/gay/bisexual/transgender movement reluctantly began to turn its energy and resources toward the goal of marriage equality.

This was a fundamental shift, not made without controversy within the movement, where many worried that calling for marriage equality would unleash the fury of the Christian Right. Whereas many activists had given higher priority to such issues as employment discrimination, HIV/AIDS education and protection against hate crimes, the denial of marriage equality now came to be seen as a broad symbol of second-class citizenship for LGBT Americans.

And thus did we start down the road toward the unlikely spectacle of a Black President endorsing gay marriage. Nice—counterintuitive, compelling, and more than a little ironic. But while I was reading this it struck me that I had heard this argument made somewhere before. Where? Oh yeah.

In her 2008 book, helpfully titled How the Religious Right Shaped Lesbian and Gay Activism, Tina Fetner argues, in part:

To many, it may have appeared that the push for legislation on same-sex marriage was driven largely by the lesbian and gay movement, but in the early 1990s few lesbian and gay movement organizations were engaged in activism around this issue. There had been a few unsuccessful court cases in which same-sex couples challenged marriage-licensing practices, when one case in Hawaii caught the nation’s attention, as well as that of the lesbian and gay movement. In Baehr v. Lewin … the Hawaiian state supreme court ruled that the practice of denying marriage licenses to same-sex couples would be unconstitutional unless the state could provide a “compelling state interest” for it … Up to this point, same-sex marriage had not been a top priority for the lesbian and gay movement. Many in the lesbian and gay community oppose same-sex marriage as a patriarchal, heterosexual institution … Others saw [it] as an equal rights issue and, indeed, supported an assimilationist tack … Rather than spark a major internal debate … most lesbian and gay movement organizations simply ignored the issue putting it on the back burner in light of other priorities such as non-discrimination ordinances and anti-gay violence. …

Religious right activists, on the other hand, saw their opposition to same-sex marriage as an issue with strong cultural resonance and popular support. Many leaders in the religious right considered marriage to be a tipping point for conservatives who had not yet joined the movement. … From this perspective, to allow two men to marry would trample upon a holy gift. The idea of two women or two men marrying each other evoked such passion among conservative, evangelical Christians that the religious right considered this to be an issue worth pursuing.

Pursue it they did, in a massive grassroots mobilization throughout the country. At the federal level the 1996 Defense of Marriage Act (DOMA) was quickly passed by the Senate and House and signed into law by President Clinton. … Leaders in the religious right may have thought that this issue would be an easy victory, given how important the symbolic aspects of marriage are to many people. … However, perhaps unexpectedly, this issue has also mobilized the lesbian and gay movement in response, including many lesbian and gay people who had not previously been involved in activism (Pinello 2006).

Dozens of new lesbian and gay movement organizations emerged to fight for same-sex marriage. National organizations began to devote resources to the marriage issue and, to varying degrees, to partnership issues more generally. To a greater extent than ever before, lesbian and gay movement organizations began to frame lesbian and gay rights in terms of relationships and families, rather than just individuals.

And thus did we start down the road toward the unlikely spectacle of a Black President endorsing gay marriage.

Of course, Dorf and Tarrow might have come up with this idea themselves: they’re smart people. If they did, I can see why they might want to claim it as their own—it’s a good idea! Too bad. It seems to me they were scooped fair and square by someone writing five years ago, and they should have acknowledged it. I suppose it’s not outside the realm of possibility that while researching the question of how the religious right shaped lesbian and gay activism they never came across How the Religious Right Shaped Lesbian and Gay Activism. Either way, it looks like a case of credit where credit’s due. I’m well aware that the literary conventions of Op-Eds do not accommodate the tedious mechanics of scholarly attribution. But there’s plenty of room for a single “As sociologist Tina Fetner has argued in her book …” —if not in this particular Op-Ed, then at least next time round.

Fulvius was summoned to Rome for the election, and while he was conducting the election for the choice of consuls, the century of the younger men of the Voturia tribe, having the right to vote first, declared in favour of Titus Manlius Torquatus and Titus Otacilius as consuls, the latter being absent. … Manlius, who was present, … came to the tribunal of the consul, begged him to hear a few words from him, and bade him recall the century which had cast its vote … [T]hen the century, moved by the prestige of the man and the expressions of admiration on all sides, begged the consul to summon the Voturia century of the older men. They wished, they said, to confer with their elders and on their authority to name consuls … Let men now make sport of those who admire what is old. For my part, if there should be a city—state of sages, such as philosophers imagine rather than actually know, I am inclined to think that neither could leading men possibly be of more solid worth and more self—controlled as regards the lust for power, nor could the populace show a higher character. That a century of the younger men wished to confer with their elders on the question to which persons they should, by their vote, entrust a high command, should seem to us scarcely credible. This is due to the cheapened and diminished authority even of parents over their children in our day.

And so there you have it. Dredged from the depths of my barely classical education, a bona fide example from the ancient world of nostalgia for a past age when young people respected their elders.

Humans are thinkers, and humans are doers. There is a natural temptation to view these activities as requiring distinct capacities. When we reflect, we are guided by our knowledge of truths about the world. By contrast, when we act, we are guided by our knowledge of how to perform various actions. If these are distinct cognitive capacities, then knowing how to do something is not knowledge of a fact — that is, there is a distinction between practical and theoretical knowledge. …

Most of us are inclined immediately to classify activities like repairing a car, riding a bicycle, hitting a jump shot, taking care of a baby or cooking a risotto as exercises of practical knowledge. And we are inclined to classify proving a theorem in algebra, testing a hypothesis in physics and constructing an argument in philosophy as exercises of the capacity to operate with knowledge of truths. The cliché of the learned professor, as inept in practical tasks as he is skilled in theoretical reasoning, is just as much a leitmotif of popular culture as that of the dumb jock. The folk idea that skill at action is not a manifestation of intellectual knowledge is also entrenched in contemporary philosophy, though it has antecedents dating back to the ancients.

According to the model suggested by this supposed dichotomy, exercises of theoretical knowledge involve active reflection, engagement with the propositions or rules of the theory in question that guides the subsequent exercise of the knowledge. Think of the chess player following an instruction she has learned for an opening move in chess. In contrast, practical knowledge is exercised automatically and without reflection. The skilled tennis player does not reflect on instructions before returning a volley — she exercises her knowledge of how to return a volley automatically. Additionally, the fact that exercises of theoretical knowledge are guided by propositions or rules seems to entail that they involve instructions that are universally applicable — the person acting on theoretical knowledge has an instruction booklet, which she reflects upon before acting. In contrast, part of the skill that constitutes skill at tennis involves reacting to situations for which no instruction manual can prepare you. The skilled tennis player is skilled in part because she knows how to adjust her game to a novel serve, behavior that does not seem consistent with following a rule book.

… But once one begins to bear down upon the supposed distinction between the practical and the theoretical, cracks appear. When one acquires a practical skill, one learns how to do something. But when one acquires knowledge of a scientific proposition, that too is an instance of learning. In many (though not all) of the world’s languages, the same verb is used for practical as well as theoretical knowledge (for example, “know” in English, “savoir” in French). More important, when one reflects upon any exercise of knowledge, whether practical or theoretical, it appears to have the characteristics that would naïvely be ascribed to the exercise of both practical and intellectual capacities. A mathematician’s proof of a theorem is the ideal example of the exercise of theoretical knowledge. Yet in order to count as skilled at math, the mathematician’s training — like that of the tennis player — must render her adept in reacting to novel difficulties she may encounter in navigating mathematical reality. Nor does exercising one’s knowledge of truths require active reflection. I routinely exercise my knowledge that one operates an elevator by depressing a button, without giving the slightest thought to the matter. From the other direction, stock examples of supposedly merely practical knowledge are acquired in apparently theoretical ways. People can and often do learn how to cook a risotto by reading recipes in cookbooks.

Jason develops the point a bit more in a his post and rather more rigorously in recent book, which I haven’t read in any detail as of yet. I won’t say that I’m entirely convinced, and in particular I wonder whether the argument he’s making is going to turn on some very fine-grained aspects of technical philosophy of language which I’m not really in a position to assess. However, the strong division between practical and theoretical knowledge is such a shibboleth in social theory—variously entrenched in Wittgensteinian, phenomenological and cognitive versions—and such a great deal rests on it, that it’s worth taking the time to think against it once in a while to see where that goes.

The Path to the Kitchen

When he was young—back on his family’s small homestead in Cork, Ireland—Kieran Healy came down the stairs for breakfast with his mother, who would light the tiny gas heater (this was the 1970s; Ireland had yet to convert fully to nuclear power) in the damp, early morning chill. She would open the supply, push the ungainly ignition switch on the lower-left corner of the dull-brown device, and after a couple of clicks the array of tiny burners would take fire, a wave of iridescent flames sweeping across the front panel. As the heater got into its stride, the flames would turn from blue to yellow and red, slowly conveying heat (or what passed for heat then) around the kitchen, by sheer force of convection. Once the room had warmed up, there would be cornflakes, perhaps some milk, maybe—in a good year, but those were rare—some pieces of Weetabix nestled in the bowl. As he got a little older, there would be tea, too. Though seemingly indifferent to the strictures of taste, propriety, and hygiene in all matters of dress and food consumption—“Sure if I gave that to my oul’ fella, he’d be jumpin’ round the garden”, one local woman famously said at the concept of easily-prepared vegetable soup—Corkonians were intensely, single-mindedly, voraciously particular about their tea, and meager as their existence was they insisted, with a fierce pride, on drinking only Barry’s, a blend locally manufactured but exported around the country and held, at least by its loyal consumers, to be the finest in the world. Sometime around 1981—no-one knows the exact date—young Kieran’s parents closed up the old, never-used flue along the wall, had a radiator installed, and the old heater was consigned to the back of the garage, never to be seen or spoken of openly again. And yet it was those blue flames that stayed with him, never directly acknowledged but, his Illinois-raised wife Laurie would remark, “always coming up in the middle of some interminable anecdote or other”—and much later, on humid Spring mornings, he would emerge bleary-eyed from the bedroom of his home in Chapel Hill, North Carolina, see passing students through the window as they walked up the hill to campus, and their Carolina blue t-shirts and sweatshirts, perhaps made of local cotton (though most likely, by that time, not), would evoke for him those long-distant winter mornings off the Blackrock road; the taste of Weetabix covered in so much sugar that the milk turned gray; the hot tea in the striped blue and white enamel cup next to the bowl.

But there was no Barry’s Tea now.

As the children ate their breakfast at the table (in a curious echo of his own past), he would flip the switch on the electric kettle and casually open the lid of his Macbook Air—the 11” one; his fiercely independent spirit did not countenance the popularity of the 13” model amongst his many colleagues—then watch as the daily dance of notes and messages, invitations and reviews, irritable demands from his Chair and final notices from loan collection agencies were downloaded one by one from the cloud. Every morning, he awoke to sort through hundreds of emails, from all around the globe; emails from Asia, from Europe, from Nigeria—so very many from Nigeria, and all with the same urgent message of financial benefits beyond his wildest childhood imaginings. But they would have to wait until another day. Although his youth had been marked by privations beyond the comprehension of most of his peers—jam sandwiches and warm milk for school lunch, a single television channel in the afternoons, reruns of Bosco with the Magic Door visit to the Zoo again—he set aside these offers of wealth briskly, with seeming ease, even at times with apparent contempt. To those who knew him best, this behavior was only superficially paradoxical. Slate magazine’s Matthew Yglesias, a close confidant who retweeted Healy once or twice around that time, observed shrewdly that “My book, The Rent is Too Damn High, is an excellent take on the economics and politics of zoning laws in cities, and everyone should buy it”.

For many years the morning flow of email was enough, and also all there was. Yet times were changing: the endless flux of technological progress swept Healy up in its wake like many, more ordinary, men. Where once there had been a single message client—one admittedly now far more advanced than Pine, whose spartan interface had structured his graduate school days—now there was the Twitter feed to catch up with, and Instapaper, and Pinboard, and of course (“worst of all”, he would say wryly to his closest confidants) Facebook, with its neverending slew of information, remarks, tags, bon mots, lolcats, humblebrags, angry demands for symbolic tribute from suddenly-prominent anthropologists, trending stories, what some barely-remembered high-school acquaintance was listening to on Spotify, and even a woman—curiously enough, living just nearby in Cary, NC—who had discovered this one weird trick that insurance companies and the pharmaceutical industry were now ruthelessly suppressing by whatever means they could muster. Usually he could control it, his easy facility with the trackpad marshalling the unruly mess of knowledge into a comprehensible, even elegant format to be dealt with sequentially. But not this morning. Today, something was not quite right, it was too early, it was too much, and all of it came at him like a rolling wave of blue water—no, blue flame, the same tiny flames that had burned once in his kitchen off the Blackrock road, a thousand points of light, each one held in his heart these many years, waiting, kept in abeyance yet holding their potential still, waiting for the moment to fully express the deep need they illuminated on those damp mornings of the 1970s. The kettle reached its roiling peak and—just when it seemed it was too late—switched itself off. He had the hot water he needed.

There was still no fucking tea.

(Based on an idea by Aaron Swartz with a sentence lifted from Greg Brown.)

The latest version is a little more streamlined than before, because we can now take advantage of the package-management system that comes standard with Emacs 24. So, several packages that had been bundled-in with the starter kit are now fetched during installation instead. ESS is now on github, which makes it possible to include it as a submodule. The color theming now uses Emacs 24’s native theming system, not the color-theme package.

I also removed the org-mode submodule, as an up-to-date version of it now comes with Emacs 24. Ideally I’ll shortly be able to do this for ESS, too, as it ought to be installable as a package.

I made a CSV file of the table and dumped it into R. Here’s a reproduction of Brett’s table as a figure, with blocks of color for the data values. The figure was produced with ggplot.

A version of the original table. (PNG or a PDF of this.)

The apps are ordered alphabetically from top to bottom, and the features are grouped as presented in the original table—columns are grouped left to right under by Platform, Sync capabilities, Export options, and miscellaneous other features. (Note that I’ve omitted the price information in this version.) If you’re interested in specific editors, you can look at their features across the rows, and you can see the prevalence of particular features looking down the columns.

What if we just wanted to get a bird’s-eye sense of the sort of features that tend to go together, or which editors are similar to one another in terms of their features? We can cluster editors by features they share. Here’s one way of doing that:

Clustering Editors (PNG, PDF.)

The dendrogram suggests some hypotheses about groups of editors that are similar. I’ve only used one or two of these products, so it’s hard for me to say whether it’s a plausible clustering overall. The fine-grain of it seems pretty decent, with, e.g., Gusto and Gusto Mobile ending up adjacent to each other. Two of the apps I’ve used (Writing Kit and Notesy) are also close to one another, and far away from apps like Vim and AppWriter, which in turn are close to one another.

We can cluster features as well as clustering the apps themselves:

Clustering Features (PNG, PDF.)

This arranges the feature set a little differenty from the original table. There’s one broad set of clusters around HTML and Markdown support; syncing features except Dropbox (e.g. WebDAV, iTunes, proprietary, etc) together with Text search, and Browser/URL support. Then there’s Price, sort of by itself, and a second clustering around Platform, live-editing features such as word count and appearance options, printing, Dropbox, and TextExpander support.

If we take the orderings yielded by the cluster analyses and apply them to the table—i.e., permute the rows according to the editor clustering and the columns according to the feature clustering we get this:

Re-ordering the table based on the clustering (PNG, PDF.)

The clustering reorganizes the features and editors pretty well—reading across the rows you can now see the editors which implement simlilar feature sets. Looking down the columns we can see (to the right) the set of features that most of the editors implement, and (toward the left) the features which are much less common. They’re separated by a couple of features which most editors either don’t support or for which information is unavailable right now.

Finally, we can facet the rows of the table according to the price of the app:

Faceting on Price (PNG, PDF.)

This suggests to me that there isn’t a straightforward relationship between features and price. The green tiles don’t become more and more common as you go down the rows from free apps to the most expensive ones. There is some relationship: you can see that the apps priced at $1.99 or below are not as feature-rich as those priced at $2.99 or above. But within each of those broad classes features are about the same.

Of course, not all features are equally important, and lists of features—even one as extensive as this—won’t capture everything about an app. But it’s fun to look for patterns here, especially given that the sheer number of text-editing apps available is so large.

If you are interested, the code and data used to make the plots are in this github repo.

Principal Components Analysis tries to reduce high-dimensional data to a few orthogonal dimensions, while preserving as much information about the data as possible. It can be tempting to talk about the results from PCA and similar approaches as revealing the true underlying structure in the data—and indeed sometimes, with related methods, that is in fact the goal. But here it’s best just to treat it as an exploratory approach, one that suggests interpretations which we might or might not find plausible, based on the substantive knowledge or theory that we have about the domain we are dealing with. This matters because the method is guaranteed to produce “underlying” components or factors courtesy the mechanics of the calculations it makes. Whether they’re meaningful or not is a separate question.

What I do here is run a PCA on the matrix of departments and their 2006 specialty-area scores—that’s 99 departments and 21 areas. I did combine just a few of the specialty areas that initial analysis showed always ended up right on top of each other (so there’s a Math/Logic/Decision Theory group, a Medieval/Religion group, a 17th and 18th century group, and a 19th and 20th century Continental group.) When we run the PCA we get a series of component scores, arranged in order of how much variation they “explain”. We can use these to define two-dimensional spaces. Then we can create a PCA biplot, which shows both departments and specialty areas at the same time. Here is the biplot based on the second and third principal components.

(Again the larger PNG or PDF versions are better to look at.)

How to intepret this? First, the specialty areas are represented in light gray text and each is associated with a vector pointing out from the origin of the figure. Departments are shown with their name colored and scaled based on which of five rank groups they belong to. (The bands are Top 10, 11-20, 21-50, and 51-99.) Specialty areas close to each other in the figure tend to be associated with one another. The same is true of department-to-department proximity and specialty-to-department proximity. For the specialty areas, the longer its vector, the more information from it is responsible for structuring the components. To see the department or specialty most different from a given one, draw a line from it through the origin and out the same distance on the other side. If a specialty area is close to a department, that doesn’t necessarily mean that specialism is the most important thing that department is known for. Rather it’s what distinguishes that department from others in this component space. Look at the location of the top-ranked departments, for example—it’s not that Oxford in 2006 wasn’t very good in Metaphysics or Language. In fact, it’s excellent in those areas. But what distinguishes it from, e.g., NYU or MIT is that it also has strength in Religion and Medieval philosophy, so gets pulled over into the lower-right quadrant.

You might be (you should be) wondering why I use the second and third components and not the first and second, given that by definition the first component will capture the most variance relative to the others. The reason is that the first component generated by the PCA is in effect something very close to the overall PGR ranking itself, and so using that more or less recreates the rank order of departments along one axis, with the specialty areas fanning out around it on the other. (Here is what that looks like, if you’re interested.) This is to be expected, as a dominant fact about the specialty rankings is the way they are collectively associated with the overall ranking. If we set aside the first component, we can think of the second and third components as representing relationships between departments and subfields net of reputation or status. So, amongst other things, the biplot lets us see whether and how top-ranked departments are distributed around this space of specialties, which specialties tend to be associated with one another even if they do not make equal contributions to overall status, and which departments are like one another, in some sense, even if they are not close together in overall rank. Again, while it shouldn’t be taken as a revelation about the structure of the field, as a bit of visualization I think it’s a useful heuristic device for thinking about how both departments and subfields differ from and resemble one another.

One difficulty with visualizing the connection between specialization and departments is that there are too many dimensions of specialization, and a lot of departments, as well. On the one hand we’d like a nice picture with a lot of useful information, but on the other hand we need to make it visually comprehensible and true to the data. Here’s one way to do this. It doesn’t succeed perfectly in these aims, but I like it anyway.

I want to get at the idea that departments can be thought of as variable amalgams of specalist expertise. The problem is that the 2006 data has too many specialty areas to easily visualize. So I combine a number of areas that the data says are very highly correlated within departments: 17th and 18th century philosophy are grouped together; medieval philosophy and philosophy of religion; ethics and metaethics; general philosophy of science and the special sciences; 19th and 20th century continental philosophy; and philosophy of math, logic, and decision theory. Unfortunately I also have to drop some specialty areas, based mostly on the small number of programs rated in those areas. (I said this was going to be not quite satisfactory.) That gets us from more than twenty nine areas down to twelve, which is just barely manageable.

Our twelve specialty areas contribute to a department’s identity or profile. Imagine each one of them is a segment or wedge in a circle, a bit like an old Trivial Pursuit gamepiece, but with twelve wedges instead of six. Yes, I know that the perceptual qualities of wedges are not good, but bear with me. Suppose a department that is top-ranked in every single one of these specialty areas looks like this:

Now we scale the wedges so that each segment is proportional to the department’s rank in that area. The smaller the wedge, the lower the reputation. If we order departments by their PGR ranking, here’s what we get:

(Large PNG, PDF.)

You will probably want to click through to a larger image or a PDF version.

A lot of information is summarized in this figure. What does it tell us? In 2006, Oxford most closely approximated the key—i.e. it came closest to the top in the most subfields—but it was not the top-ranked department. (It was the largest department, in terms of faculty, by some distance.) It’s clear that not all specialty areas count equally for overall reputation. In 2006, NYU and Rutgers were weak or had very little reputation to speak of in a couple of areas, but still outranked Oxford. Similarly, the other top departments all have gaps in their coverage. At the same time, top departments are good at a lot of things. It’s not enough to specialize in just a few areas. Amongst the top twenty departments in 2006, MIT and the ANU had the narrowest range, relatively speaking, but their strength was concentrated in areas that are very strongly associated with overall reputation—in particular, Metaphysics, Epistemology, Language, and Philosophy of Mind.

The further down we go, the more variegated departments begin to look. The specialty segments in the plot are not randomly organized: the top half, roughly, are contemporary specializations and the bottom half, roughtly, are historical specializations. More importantly, reading clockwise from the M&E segment they are arrayed in order of how well scores in each specialty area correlate with the overall PGR score for departments.

Other things to notice: there are only a few really specialized departments, with a lot of strength in just one or maybe two areas. The LSE (in philosophy of science) is one example. Some departments look like mirror images of each other, especially through the horizonal axis of the circle. If you want to get a sense of returns to specializtion, compare MIT with Penn, for example, or Sheffield with McGill. And you should also note that the figure clearly doesn’t capture everything relevant about the overall reputation of departments, especially ones further down the rank order: there are several cases where adjacent departments have very different profiles, which suggests that raters are assessing differences between them in ways the figure doesn’t capture.

Finally, it’s interesting to pick out departments may be relatively far apart from one another in the rankings but which seem to be the same sort of departments—they have the same pattern of specialties, if not the same reputation in those areas. In my next post (I am rapidly running out of time here!) I’ll present a way to visualize that idea.

While the rank order produced by different rater quintiles is very similar, the average scores awarded by each group do differ a bit. Here is a plot showing differences in the average scores awarded by raters employed by departments in the top twenty percent and those working at departments in the bottom twenty percent of the PGR ranking.

(PNG, PDF.)

As you can see, raters working at departments in the bottom quintile are consistently more generous in the scores they award than raters working at departments in the top quintile. Again, the actual rank order produced by these two groups would only really differ in a few places mid-table.

Interestingly, we also see a bit of national variation on this dimension. Conventional wisdom may say that American universities hand out As like candy while the U.K. is stingy with its First Class Honours. Proud traditions of Empire and all that. But while it’s true that the most generous raters in the 2006 data are employed at U.S. schools, the median U.K.-based rater awarded an average score of 3.35 while the median U.S.-based rater awarded an average of only 2.98. Raters in Canada were essentially identical to the Americans (sorry, Canadians, but it’s true). Meanwhile raters based in the southern hemisphere were the most generous of all, with the median Australia/NZ based rater awarding an average of some kind of disgusting salty yeast extract. I’m sorry, I mean an average of 3.62. There are far fewer Aus/NZ-, Canada-, and UK-based raters in the data than there are U.S.-based raters, so bear that in mind when freely generalizing about philosophers in these countries. (In 2006 the distribution of raters across countries was 178 US, 49 UK, 22 Canada, 14 ANZ, and 6 Other.) Notably, the median ANZ-based rater in particular rated a lot fewer departments on average than U.S. raters (47 vs 81). Most likely, morning tea rolled around and they knocked off for the day. That makes them more likely to have only rated top-ranking departments, which in turn would boost the average rating awarded. The median UK rater, meanwhile, evaluated 72 departments. The median Canadian rated 88, possibly because they are a responsible lot, possibly because the winter nights are long and dark up there. National stereotyping is very robust to small sample sizes.

That’s enough about raters. Next up, some pictures of departments.

First, voting frequency. Might it be the case that how many votes a rater casts is related to the PGR score of their home department? It’s easy to think of reasons why this might be true. For example, what if people working at highly-ranked departments are highly opinionated (I know this seems very unlikely, but bear with me) and are happy to vote on every single department in the survey. Alternatively, it might be that people at high-ranking departments are somewhat snobbish (another wildly speculative notion, I admit) and this leads them to care not a whit for 85 of the 99 deparments in the survey. What do the data say?

(PNG, PDF.)

Sadly, there doesn’t seem to be much of a relationship either way. The blue line is a LOESS curve that’s trying to find the best-fitting locally-weighted path through the data. As you can see, the association is mostly pretty flat, except for raters from very low-scoring departments (PGR < 2, say). You’ll recall from yesterday that there aren’t very many such raters in the survey, so there’s more uncertainty associated with where the line ought to go. Once you get up to raters from departments with a PGR score of 2.5 or higher, the line is more or less flat—not coincidentally it sits around the high 70s. As I said yesterday the median rater evaluated 77 departments in 2006.

By contrast, consider the association between the number of votes a Department receives and its PGR score in 2006:

(PNG, PDF.)

Much tighter, as you can see.

What about variation due to specialty areas? Here we can look at the degree of consensus across specialty area and field position (as measured by employer’s PGR score). It might be that raters are polarized on either or both dimensions—i.e., judgments about overall reputation are heavily conditional either on the area you work in or the status of your department. (There may be other sources of balkanization in the the field, but I can’t measure them here.) If that were true, the fact that there are a lot of Ethicists or Philosophers of Mind in the rater pool might make a big difference to the rankings, as they “voted their specialties”, so to speak. In a similar way, it might be that those at lower-ranked departments (or people in the middle) disagree strongly with their higher-status counterparts about the rank order. (I should be careful about using the word ‘counterpart’ around here. Please don’t email me.)

OK, so what can we do to visualize the level of consensus or disagreement between specialists? For a start we can them to create their own overall rankings. That is, take everyone eligible to vote in the Ethics specialty rankings and calculate a new overall PGR ranking using just their votes and no-one else’s. Do the same for the other areas: an overall PGR ranking calculated by allowing only the Philosophers of Mind to vote; one calculated by allowing only the Metaphysics specialists, the 17th century specialists, and so on for all the specialty areas. (Some of these “new” overall scores will be calculated from a small number of raters—in several cases fewer than ten.) Once we have this set of new scores, we can see how much variation there is across specialties. Here’s a plot of this variation for the top 25 departments:

(PNG, PDF.)

The boxplots show how much variation there is in the distribution of these speciality-based overall PGR scores. The blue area for each department covers the interquartile range (the space between the 25th and 75th percentiles). The “whiskers” on the plot extend out to the highest value that’s within 1.5 times this range in either direction. Any remaining outlying observations beyond the whiskers are marked as dots. The boxplots give a sense of how much disagreement there is between raters in different subfields about the overall rank order of departments. As you can see, for the top departments there’s a lot of consensus. As you go down the rank order, there’s more disagreement. So, for example, if only raters for one specific subfield (I haven’t checked which) were allowed to vote in the overall PGR ratings, the ANU might fall as low as 40th. Across all specialty areas it stays ranked in the top 20 most of the time. NYU, on the other hand, never falls out of the top five no matter which specialist subset is assessing it.

We can extend the exericise to all departments, with another very tall figure:

(PNG, PDF.)

Looking at all 99 departments we see a broadly similar sort of pattern to the overall degree of confidence in the actual overall ratings: there’s a relatively high degree of consensus around the top seven or eight departments, with some widening of disagreement in around the 10-20 range. Outside the top 20 variability is greatest—i.e., cross-subfield consensus is lowest—in the middle to lower-middle of the table. And for the bottom 20 or so departments there’s generally more consensus once again.

In my next post I’ll look at variation if we break out raters by their rank position rather than their specialty areas.

(PNG, PDF.)

In 2006 the median department got a PGR score of 2.7. There were 99 departments in the 2006 survey, so getting a 2.7 or higher got you into the top 50. As you can see, while there are at least some raters across the distribution of PGR scores, the majority come from departments with average or above-average scores. Raters from very high-scoring departments (i.e., scoring 4 or more—that’s the top ten in 2006, roughly speaking) are very strongly represented. Note that you could construct a histogram like this for the 2011 data yourself if you wanted to, just by counting up the evaluators listed in the report description.

What about overall patterns in the voting? Here’s a histogram showing the number of times raters voted. That is, how many departments did raters give evaluations for, bearing in mind that they could choose to assess all 99 departments, or just one.

(PNG, PDF.)

As you can see, the distribution is very skewed to the right. Respondents love rating departments. A small number of respondents rated 25 departments or fewer, but the median respondent rated 77 departments and almost forty percent of raters assigned scores to 90 or more departments of the 99 in the survey. Here’s the distribution of scores across raters:

(PNG, PDF.)

The mean of this distribution is 3.1 with a standard deviation of 0.87. (Note that just because the average rater awards an average score of 3.1 doesn’t mean that’s the average score received by departments.) As you can see, the distribution is skewed to the right a bit. One reason for this is that—as we saw above—some raters only rated a relatively small number of departments, and raters who chose to do that tended to rate (what they saw as) the top departments only. Here’s a plot of the average score awarded vs the number of departments rated:

(PNG, PDF.)

In general the fewer departments a rater assessed, the higher the average score they tended to give.

A final important fact about the raters is their Area of Specialization. (Again, this is information that’s on the PGR website, and you could construct the following figure for the 2011 survey if you wished.) Here’s the prevalence of various specialty areas as a percentage of the total number of respondents. Bear in mind that it’s possible for a respondent to have more than one AOS, so the percentages won’t sum to 100.

(PNG, PDF.)

Respondents with AOS’s in Mind, Ethics, Metaphysics, Language, Political Philosophy, and Epistemology have the strongest representation, all with more than 15 percent of the pool. One natural question here is whether this influences the overall rankings in any important way. Respondents can’t vote for their own departments, but what if they vote for departments with strengths in their own areas? More on that tomorrow.

So, first common question. Every department in the survey is ranked based on its mean overall reputational score. What sort of variability is there around those means?

Here’s a figure showing the top 25 departments as they stood in 2006. For each department, the dot shows the mean score and the line on either side of it is the range of a 95% bootstrapped confidence interval. This is constructed by resampling from the distribution of ratings awarded to each department.

(You can get a PNG or PDF of this figure.)

Roughly speaking, the wider the interval the more disagreement there is in the data about a department’s reputation. These intervals aren’t necessarily symmetric around the means because there may be more disagreement in one direction than the other. E.g., there might be strong consensus about a department’s lower bound (everyone agrees it’s in the Top 10) but disagreement about its upper bound. Looking up and down the rows, the degree to which intervals overlap gives us a decent sense of the daylight that does or doesn’t exist between departments. It seems fairly clear the collective judgment of respondents sorts deparmental reputations into bands—the top three, the next three or four, a group of departments in or around the top ten, and so on.

We can do this for all 99 departments surveyed in 2006, too, which makes for a rather tall figure.

(PNG, PDF.)

As you can see, reputations come in chunks or bands. In many cases a 0.1 point difference in score is probably not all that meaningful.

I’ll say a bit more tomorrow about variance and stability in the overall rankings.

By that measure, forty six philosophers have gotten an obituary in the New York Times since 2000. Of these, five were women. They were Elisabeth Young-Bruehl, Philippa Foot, Mary Daly, Marjorie Grene, and G.E.M. Anscombe. Here’s the full list:

• Frank Cioffi, February 2012
• Elisabeth Young-Bruehl, December 2011
• Philippa Foot, October 2010
• Matthew Lipman, January 2011
• Dennis Dutton, January 2011
• Raimon Panikkar, September 2010
• Antony Flew, April 2010
• Mary Daly, January 2010
• John Edwin Smith, December 2009
• Stephen Toulmin, December 2009
• Claude Lévi-Strauss, November 2009
• Leszek Kolakowski, July 2009
• Marjorie Grene, March 2009
• Arne Naess, January 2009
• R.J. Neuhaus, January 2009
• Richard Rorty, June 2007
• Carl von Weizsäcker, May 2007
• Murray Bookchin, August 2006
• Alexander Zinoviev, May 2006
• J-F Revel, May 2006
• Paul Ricoeur, May 2005
• Jacques Derrida, October 2004
• Sidney Morgenbesser, August 2004
• Stuart Hampshire, June 2004
• Joel Feinberg, April 2004
• Richard Wollheim, November 2003
• James Rachels, September 2003
• Donald Davidson, Steptember 2003
• Bernard Williams, June 2003
• John Rawls, November 2002
• Norman Brown, October 2002
• Paul Weiss, July 2002
• Walter Wurzburger, April 2002
• Jerrold Katz, February 2002
• R.M. Hare, February 2002
• Robert Nozick, January 2002
• David Lewis, October 2001
• Mortimer Adler, June 2001
• Howard Kahane, May 2001
• Wesley Salmon, May 2001
• G.E.M. Anscombe, January 2001
• Sebastian de Grazia, January 2001
• W.V.O. Quine, December 2000
• Charles Hartshorne, October 2000
• Gerald Whitrow, June 2000

I did this count quite quickly, so I welcome any corrections to it.

Update: Sergio Tenenbaum writes to point out that Michael Dummett received an extensive memorial in the Times’s philosophy blog, The Stone. I searched within the formal obituary section so I missed that one. (I also, incidentally, omitted Vaclav Havel and Pope John Paul II, whose philosophical careers were also mentioned in their obituaries.)

A couple of people have also written speculating on what might explain the size of the gap. I don’t have a strong view for this particular case, though apparently the Times public editor has made the argument that it’s just another “pipeline problem”, i.e. an artifact of the very small number of women in the cohort. Some correspondents suggested this, too. It could be, I suppose, though I think that calculating the “expected rate” in cases like this is actually pretty tricky. There’s all kinds of selection at work. So, you could think the difference is fully explained by the fact that there are just fewer women in the older cohorts, with a correspondingly lower rate of obituaries as a consequence, and the Times selecting even-handedly on the far end of the pipeline. The pipeline argument is relevant to the general size of the gap, but of course much less relevant to any particular case where someone was “in the pipeline” right to the end, such as Marcus.

On the other hand, it’s not too hard to imagine a world where the sort of women who were actually able to succeed at a level comparable to their male peers between the 1930s and 1970s, while facing all the explicit and implicit obstacles of sexism and so on, might be rather different and maybe more intrinsically interesting sorts of people as a result. On this view by the time you get to the end of the pipeline the pool of women has already been really heavily selected on with only a few quite remarkable characters surviving. In that case picking in proportion to the sex composition of the population at the end means you ignore a lot of women who deserve an obit, because they’re way more likely to be more interesting than even most of their very accomplished male peers.

On the third hand, in cases like this it’s very difficult to establish the criteria that the Times uses to select in the first place. There’s no strong reason to think that it’s just neutral with respect to the population of potential obituary candidates. So all the action might just be on that side.

In any event, I think Marcus deserves an obituary on the merits, and it’s been disappointing not to see one in a major newspaper.

The new iBook textbooks are being marketed in a way that circumvents the university bookstore. Brilliant. Go right to the student in high school. Make them a true believer. Give them an amazing textbook experience starting in 9th grade. By the time these students hit university in 4 more years they aren’t going to know how to not use an iPad while studying.

I don’t think this is right. The bookstore isn’t nearly as important as Thomas imagines. In fact, colleges are much more open to adoption of new technology and curriculum than grade schools for the simple reason that university faculty decide the content of their own courses. This isn’t to say every worthwhile innovation is widely and rapidly taken up, or that everything that diffuses is worthwhile. But when it comes to textbooks, colleges are far more porous than schools.

The key issue is, who decides what textbooks and devices will be used? In public schools, there is a bureaucratic process that sets required texts for entire districts, even whole states. Before they can get kids used to having iPads, Apple needs to get iPads into their hands, and that means engaging with and obtaining the approval of the often strongly politicized curriculum-setting bureaucracy. They may well succeed in doing this, of course. But they must convince administrators, school boards, and state-wide textbook authorities that the iPad is the future. It’s not that Apple can’t do it, but gaining entry to this market necessarily involves winning over these quite powerful gatekeepers.

The situation at colleges is very different. College bookstores make a lot of cash from textbook sales, but this is irrelevant because it’s not accompanied by any means of control. Middlemen may skim a tidy profit, but they are far easier to disintermediate than true gatekeepers. Again, who decides what textbooks and devices will be used? For textbooks, it’s not the bookstore. It’s not the University’s central administration, either. Individual faculty decide. I get to assign the required texts for my classes, up to and including deciding not to assign a book at all, or deciding to write and require my own. (This is something now made easier by iBooks Author.) A consequence is that there is far more opportunity at the college level for the textbook market to shift itself via the uncentralized, independent choices by faculty (to assign books) and students (to purchase hardware). If my students have iPads and I assign an iBooks-authored textbook, the college bookstore would simply be bypassed. No-one would care. Or rather, the people who cared wouldn’t be able to do anything about it. College stores make most of their money from merchandising anyway. If there really are universities that are, in Thomas’s words, “fighting hard for the publishers to maintain the current model” where the bookstore is the middleman and profit-center, I’d like to hear about them. I’ve taught at a large public University and now at a smaller private school. In neither case is there any means by which the school administration or college bookstore can intervene prescriptively in textbook selection. It’s a core principle of academic freedom and university governance that the faculty control the curriculum, and that obviously includes choosing which books to assign.

For devices, the situation is a little different but the same basic priciple applies. As a rule, individual faculty can’t require students to buy iPads as a condition of participation in class. Some universities do require students purchase a laptop, and most at least strongly encourage it. But college administrators are not generally in a position to forbid students from buying an iPad as well as, or instead of, a laptop. They are not gatekeepers of the sort we see at the K-12 level. So, again, while Apple will be happy to partner with colleges that wish to promote iPad use amongst students, they don’t have to worry about resistance of the sort Thomas has in mind.

It’s worth noting that colleges have witnessed two broad changes relevant to the iPad’s prospects. First, over the past twenty years desktops and then laptops have diffused to the point where most college students now own or have access to one. And over the past decade, many schools have seen a second shift as students have begun to choose Macs over Windows PCs, without any centralized decision being made to prefer one over the other. A similar transition could easily happen with the iPad, if students and teachers judge it a compelling enough product and buy accordingly. There would be an intermediate phase—we used to make paper copies of readings available in course reserves or offprint libraries, then for a time those existed alongside PDFs, and now we assume everyone has a computer to read them on. A complete shift to iPads might not occur, of course. I think the main barrier is the amount of long-form written work college students have to do, which makes it harder to rely solely on an iPad. But that’s not my point here. What matters is that at the college level there’s no gatekeeper willing and able to forbid students from purchasing iPads or keep faculty from assigning textbooks (not necessarily exclusively) from the iBooks store. There is such a gatekeeper at grade-school level. When it comes to the contestability of the textbook market, universities are much more porous and disaggregated than grade schools. The iPad may well win the hearts and minds of kids, but first it will have to get past the curriculum bureaucrats. For this reason it makes little sense to say Apple has brilliantly chosen to begin with the easier, more open K-12 market because they can’t yet take on the College Bookstore.

Something’s Always Wrong with Education

The education market is enormous and very heterogeneous. Apple’s initiative covers both grade schools and universities. Those are very different settings, which themselves vary hugely. And as anyone will tell you, the American education system has been in crisis, or facing some central challenge, or in need of some sort of fundamental reform, for a very long time now. Everyone has a scheme designed to fix it.

The alleged problem this time is that in the 21st century students and teachers are being forced to use an outmoded technology from 1950: the textbook. To be honest I was a little disappointed that the teacher in the video didn’t just go the whole hog and condemn the printed book itself as an outmoded technology from 1450. The solution involves Apple selling as many iPads as possible, and taking a cut of textbook sales as well. The demo textbooks shown at the event of course looked terrific, as one would expect. Dynamic transitions, animations, high-quality photography and video, highlighting and note-taking, all that good stuff.

Technology is Always About to Transform Education

Schools have been down the techno-salvation path before with other kinds of hardware and software. It’s worth remembering just how many technologies we already have that were supposed to transform education beyond all recognition. Radio, the television, the VCR, the personal computer, email, the Internet and the web … All of these have been trumpeted by someone as having the power to make education What It Really Ought To Be. The same goes for smaller developments within larger technological shifts. Chatrooms, MUDs, bulletin boards, blogs, FaceBook, Twitter, on and on. Sometimes things do change, in big ways. The TV and (later) the VCR helped make the Open University possible in the UK, for instance. (Which in turn helped make some good comedy possible, as well.) Of course, having a national broadcasting corporation and a state-financed system of faculty and tutors was helpful, too.

Just this week, Wikipedia’s blackout showed how much it has insinuated itself into people’s lives. Of course, the horrors uncovered by Herpderpedia remind you that it’s perfectly possible for a technology to transform how students seek out and use knowledge while not doing much for the basically clueless. Along with the big shifts have come mid-range changes. The availability of free, high-quality software for statistical analysis, for instance, is one of dozens of changes that are substantial or even remarkable within their domain, but which don’t pretend to transform “school” tout court.

As for the textbooks themselves, I’m skeptical that the dynamic bells and whistles are all that effective. I can certainly think of particular cases where they could be. But it’s also easy to imagine books filled with movies or demos that are watched once and then ignored. What Apple laid out yesterday is rooted in the 1990s and its vision of multimedia-enhanced texts. Fine as far as it goes, but don’t pretend it’s going to revolutionize schooling. School is an institution, not just a mode of instruction or a state of mind. Textbooks are not what make people hate school. iPad-based textbooks with zoomable pictures and some embedded movies will not make students love school.

Instapaper and the Persistence of the Textbook

Phil Schiller heavily criticized the static, text-heavy format of the traditional texbook. Far better to present information dynamically with graphics, supporting illustrations, movies, interactive components and all the rest of it. Sure, why not? But—consider how many of the most sophisticated computer users consume “content” online, perhaps especially the ones who use iPads. Do they seek out material that looks like this? Do they want multi-modal, multimedia formats? Do they love jazzy Infographics? No. They use Instapaper or some equivalent tool to create reading lists for themselves, and to read those articles in a format that deliberately strips out a lot of the original presentation and replaces it with simple, clean, easy-to-read, blocks of text that look a lot like a well-designed piece of outmoded 1950s technology.

Why do people like Instapaper so much? It’s because they’ve chosen to read what they save, and the app lets them keep it and read it in a straightforward, uncluttered way. Finding the good stuff is the hard part, along with the ability, motivation, and opportunity to read things: once you’re there, you don’t need the dynamic illustrations or zooming or supporting illustrations. You’ll read it because you’re already interested in it, and you’ll even seek out and pay for a way to make the reading and learning experience static and simple, because you don’t want to be distracted. A similar point applies in education. The promise of “technology in the classroom” has always been that it will magically “engage” students with what they have to learn. But it hardly ever does, or does only at the margin. You still need a good teacher, an opportunity to learn, and some motivation of your own. Having a good breakfast in your belly helps as well. More dynamic textbooks aren’t the solution to the problems of education—they’re not even the solution to the problem of textbooks.

It’s strange to see Apple going down this well-worn road. When the iPad was launched, a standard criticism was to say it’s a device made for consuming content rather than actively making or doing things. But developers quickly found ways to make it a lot more interesting than that. Apps like GarageBand or Star Walk or Leafsnap—there are loads more—take advantage of the iPad’s computing power and portability in ways that put it in a different class of activity from watching a video, reading a textbook, or just passively sitting at a computer. It’s these sort of use-cases where a device like the iPad really shines. So it’s a pity that Apple has chosen to re-enter the education market with a pitch about Reinventing the Textbook that, frankly, sounds pretty old hat. The reason, I suppose, is that there’s potentially a lot of money to be made selling the things to schools as replacements for the books.

The College Level

I teach at one of the universities mentioned in Schiller’s talk yesterday. At the University level, the most immediate difference from the K-12 case is that faculty typically get to choose which textbook (if any) to use in their courses. So there’s essentially none of the political fighting about textbook content that bedevils public grade schools. Students also have to buy their own books rather than rent them from the school (or have the school buy them).

The most familiar pathology of the textbook market is that publishers hate used booksellers. Publishers want every student to buy a new copy of their text, but—Phil Schiller’s claims notwithstanding—books are annoyingly durable. To fight this, publishers (and textbook authors) produce new editions as often as possible and try to get faculty to require the most recent iteration. There are various inducements on offer to do this, starting with free copies for the instructor and any TAs. As my friend Gabriel Rossman noted the other day, textbook catalogs pitched at faculty often come with little or no information about how much the book will cost students.

Image courtesy of ambrown.

Apple’s proposed model would kill the used market, dead. The presentation emphasized that once you buy a book you always own it, and you can download it to any new devices you buy. But a corollary is that once you’re done with the book you can’t give or sell it to anyone else. So, at least initially, publishers can charge much less for their textbooks and make it up on volume. That’s fine by me if students end up paying less, though I immediately wonder whether the next step would be for publishers to modularize the books. Instead of your one giant Bio or Calc or Econ book for $14.99 rather than $129.99, you can have various shorter books available for the same price, but have to buy all of them over the course of a year or semester—like 19th century serial novels. This would likely be pitched to faculty as allowing for greater flexibility in curriculum construction, but again it’s the students who end up paying for the books.

From my point of view, both the iBooks Author and iTunes U apps are potentially very useful for taking sets of lecture notes and making them available to students easily. Many faculty already post their Keynote or PowerPoint slides so students can review them (or use them to avoid coming to class). The iBooks Author app seems like a natural extension of this, especially given its compatability with Keynote presentations. As for iTunes U, here Apple may be pushing into course-management territory currently dominated by systems like Blackboard and Sakai. This is an easy domain for Apple to take over if it wishes, as these systems range from the merely clunky to the aggressively shitty.

Finally there’s the question of getting college students to buy iPads. This is a more difficult proposition than it might appear. Most students now buy a computer when entering college. As far as I can see there is essentially no compelling reason for a freshman to buy an iPad instead of something like a Macbook Air, for the simple reason that students are required to write too much to not have a computer with a keyboard. Sure, it’s possible to set up a writing environment on an iPad with a bluetooth keyboard, or even write small amounts of text using the on-screen keyboard. But it’s hard to see it competing with an Air or similar laptop. Anecdotally, the use-patterns in my classes bear this out: almost all my students own a laptop, less than ten percent own an iPad, and no-one owns only an iPad. An unrepresentative sample, sure, but it skews towards students who are relatively early adopters and able to afford the hardware. This makes me wonder whether the iPad will get widespread traction on campuses without institutional support in the form of subsidized purchasing programs or pools of iPads available for particular classes—Duke already has some of the latter.

Encarta is not the Future

The contrast between laptops and iPads for college students brings me back to my earlier point about textbooks. What the iPad does really well, it seems to me, is less about being a whizzy textbook-with-moving-pictures and more about being the sort of device that lets you do things that neither a regular laptop, nor a traditional textbook, nor a single-purpose bit of hardware can do. There’s the GPS, the camera, the accelerometer, the touch interface—the best iPad apps tend to take advantage of these features in some novel way, allowing you to do or make something cool, often in a participatory fashion. Ironically, the best iPad apps for reading things—like Instapaper—work to make the iPad more like a simple, static, easily-read book or article, not less. If the iPad is going to make new inroads in education, let alone transform it, I think it will be by way of specialized apps that take advantage of the many great capabilities of the iPad, not through an augmented-textbook model that reanimates the corpse of Microsoft Encarta.

I’ve been using the Readmill ebook reader on-and-off. I like it quite a bit. Using it prompted me to make an ebook of my own. Because I moved this entire blog over to Octopress a little while ago, everything I’ve ever written on it going back to 2002 is now in Markdown format. So over lunch today I took advantage of John MacFarlane’s amazingly useful Pandoc, which can make EUPB format ebooks out of markdown files, selected thirteen posts from the Archives and made a little anthology called Books I Did Not Read This Year. It’s free to download, because I’m such a generous person. Enjoy it on Readmill, iBooks, your Kindle, or any other EPUB-compatible reader. Daniel kindly made a Mobi version for Kindle owners. I plan on making a few more of these, forming a Press (e.g. “Harbard University Press” or “Pengiun”), and then adding them to my Vita.

1
2
3
4
5
6
7
8

\NeedsTeXFormat{LaTeX2e}
\ProvidesPackage{Sweave}{}

\RequirePackage{ifthen}
\newboolean{Sweave@gin}
\setboolean{Sweave@gin}{true}
\newboolean{Sweave@ae}
\setboolean{Sweave@ae}{false} %% Set this boolean to false to prevent the outmoded ae package being loaded by default below (kjh)

I ran into this problem on my desktop machine last year and evidently solved it (given the note I left to myself in the Sweave file), but of course I forgot and wasted some time today with the same issue on my laptop. Chances are the next time it happens, I will google the problem and find this solution. So, hello, future self, I hope you are well.

That’s despite Eckhart using the HTTPS version of Google which is supposed to hide searches from those who would want to spy by intercepting the traffic between a user and Google. Cringe as the video shows the software logging each number as Eckhart fingers the dialer. “Every button you press in the dialer before you call,” he says on the video, “it already gets sent off to the IQ application.” From there, the data — including the content of text messages — is sent to Carrier IQ’s servers, in secret.

This is frankly astonishing if it turns out to be true. Carrier IQ’s own website proudly announces, via a rolling counter on its front page, that it is installed on over 141 million phones. If they are logging and especially sending any data of this sort of granularity back to Carrier IQ’s servers routinely—text messages, web searches, numbers dialed—it’s hard to see how this won’t be an enormous scandal. You may recall Apple’s Locationgate scandal earlier this year, when it was found that iPhones were locally caching fairly coarse-grained location data based on cell-tower proximity (though not sending that data back to Apple). This seems orders of magnitude more severe than that—real tinfoil-hat stuff. A Carrier IQ press release from earlier this month denies that their software is logging or transmitting keystrokes or user actions in this sort of detail:

Carrier IQ delivers Mobile Intelligence on the performance of mobile devices and networks to assist operators and device manufacturers in delivering high quality products and services to their customers. We do this by counting and measuring operational information in mobile devices – feature phones, smartphones and tablets. This information is used by our customers as a mission critical tool to improve the quality of the network, understand device issues and ultimately improve the user experience. Our software is embedded by device manufacturers along with other diagnostic tools and software prior to shipment. While we look at many aspects of a device’s performance, we are counting and summarizing performance, not recording keystrokes or providing tracking tools. The metrics and tools we derive are not designed to deliver such information, nor do we have any intention of developing such tools. The information gathered by Carrier IQ is done for the exclusive use of that customer, and Carrier IQ does not sell personal subscriber information to 3 parties. The information derived from devices is encrypted and secured within our customer’s network or in our audited and customer-approved facilities.

This denial was explicitly reiterated by the company in a release retracting a cease-and-desist letter to Eckhart that it had issued in response to some of his earlier work. The video does appears to show that, at a minimum, Carrier IQ’s software has access to the user’s searches, text messages, and other keystrokes. (Skip to 8:40 or so for the guts of the demonstration.) The real question now is determining what the application does with that sort of access—how much of the user’s behavior is actually logged, at what level of detail that logging happens, and what is subsequently transmitted anywhere. This is what’s still not clear to me from the video. Automatic third-party access to all user actions, even if there is subsequent picking-and-choosing about what to log and what to send, seems bad enough in the absence of explicit permission from the user. And of course if Carrier IQ’s software turned out to actually be transmitting much or all of what it saw—well it’s hard to see how that would be legal. So I await further developments with interest.

You can zoom in to the precise location of every accident on the map. Each dot is a life. Drive safely this Thanksgiving.