not just my opinion

21 Sep 2011

The real class warfare...

Seventy-one percent of current U.S. debt was accumulated during Republican presidential terms.
Two-thirds of debt-ceiling elevations since 1960 have been signed into law by Republican presidents.
In 1961 the percentage of corporate profits paid in taxes was nearly forty-one; now it is less than eleven.
Seventy-five percent of the increase in corportate profit margins since 2001 has come from depressed wages.
The value of government subsidies that will go to oil and gas industries between now and 2015 is estimated at $78,155,000,000.

Information from Harper's Index, courtesy of frontporchrepublic.com

9 Sep 2011

The Sparrow, by Mary Doria Russell
I found this on some random list of recommended speculative fiction (can't remember the source of the list). It's a great premise, as far as scifi goes: scientists discover a radio signal coming from a star system in the vicinity of Alpha Centauri, which turns out to be music. A mission is sent to the planet, and 40 years later one survivor returns to tell the tale. Pretty promising premise. I had two major quibbles with this book, however:

The writing was sub-par. It wasn't grammatically questionable, but it just felt amateurish throughout. There was too much use of dialogue as plot exposition, and that kind of thing. Also, the characters felt like they were designed for maximum effect rather than existing as fully faceted entities, even from a fictional standpoint; problem one may have been a side effect of the other major problem;
In the end, I felt that the whole book was mainly an excuse to explore the theological problem of evil. Ostensibly, the question of the existence of God was also under consideration, but it became pretty obvious that this wasn't really up for grabs, and that the real question was how to integrate evil into belief. The author suggests that her original topic of interest was the anniversary of Columbus's voyage, and the various implications of one culture "discovering" another, but I felt like that topic didn't get nearly as much attention as the theological question. Since that's not a theological question that interests me, I got tired of that aspect quickly.

There definitely were some interesting aspects to the book. The intelligent alien species MDR imagines, for example, are quite fascinating. There are two intelligent species with a disturbing symbiotic relationship. MDR is actually quite good at putting cultural trappings around a different conception of intelligent lifeforms in ways that make sense. At least part of my disappointment was that of a missed opportunity: because she is quite good at certain aspects of expanding on an interesting premise, I found it that much more disappointing that her main interest seemed to be the theological questions.

Breaking Bad

I finally got around to watching Breaking Bad, starting from the beginning. The fourth season is ongoing, apparently, though I'm still working my way through Season 2. I must say, first of all, that I'm once again very impressed by the quality of original programming being put out by AMC. Mad Men is amazing of course (if uneven), and I really enjoyed the first season of The Walking Dead. Like those, Breaking Bad is very high quality work.

The basic premise: a high school chemistry teacher, Walt, finds out that he has inoperable lung cancer, and this shock to his system leads to his decision to start manufacturing crystal meth with Jesse, a former student of his who has since become a drug dealer. There are a lot of directions you could go with this story, and the creators/writers have chosen (so far anyway) a gritty, realistic approach. Unlike Weeds, say, which immediately swerved towards cartoonish characters and straight up gallows humor, Breaking Bad is trying to stay on the straight and narrow of something that's at least plausible (not that it doesn't have its share of funny moments, mind you).

The story is anchored/constrained by Walt's pregnant wife, Skyler, and disabled son, Walt, Jr., who are constantly forcing Walt back towards evaluation of his ongoing transformation from high school chemistry teacher to would-be drug kingpin. The characters are interesting and solid, and the storylines have been very engaging. There's both episode and major story arc foreshadowing (fragments of action yet to come that doesn't make sense at first, but becomes clearer as the story goes along), from time to time, and this also is a great way of ratcheting up the tension. I'm nearing the Season 2 finale, and will report more as I go.

2666, by Roberto Bolaño

Unfortunately, I don't have the time to write a review of this book that does it justice. Suffice to say that I think it's somewhere up in the masterpiece category (flawed masterpiece, perhaps, but still masterpiece). Which doesn't mean that it's a compelling page-turner, mind you. This is not reading candy. It's work, and in some places a full-on slog, but ultimately well worth the effort. I'm not sure it's even worth trying to summarize the plot, except to say that the work is divided into five sections, each of which could stand alone (and in fact, Bolaño, who died before 2666 was published, left instructions to publish the five sections independently, presumably because it would be easier to sell them individually), but which definitely are interwoven in ways that make them connected (there's a bit of a Mobius strip aspect to the way they're connected, though not in the tidy, clever way of something like Memento or Pulp Fiction).

The best way to describe the work of Bolaño generally, and this work in particular, is that I believe them to be near perfect examples of what Roland Barthes calls a "writerly text," which Barthes describes thus:

"[a] writerly text is ourselves writing, before the infinite play of the world is traversed, intersected, stopped, plasticized by some singular system (Ideology, Genus, Criticism) which reduces the plurality of entrances, the opening of networks, the infinity of languages..."

Or, to put it another way, a writerly text does not constantly struggle to reduce the possibilities of meaning, but instead glories in the interpretive space created by (inherent to, according to Barthes) all texts. I think I can safely say that 2666 does not (nor try to) shut down the interpretive possibilities it opens up. If all of that sounds like deeply annoying gobbledygook to you (and I'm sympathetic if it does), then 2666 is probably not for you. If you're even a little curious or intrigued, I recommend the experience.

0 7 Sep 2011

Models and Meaning, Part 2

I wrote an earlier post about meta-modeling, or meta-analysis. A friend made some helpful and cogent remarks, including this:

Part of the problem for many modelers is that they're trying to hit moving targets. You can pick all the relevant drivers, the functional forms that fit the data, and describe much of the variation over a given historical sample only to see a regime shift as soon as you apply the model out-of-sample. Ugh. It's better if you can build an all-weather model that knows how and when to switch to the new state of the world. Of course, that's inherently difficult since you probably would have needed at least some episode in the past that covers and correlates with the present situation in order to train the model to account for it. That's more of a challenge in some disciplines more than others. The hard sciences seem to have all these wonderful immutable laws that make there jobs, if not easier, at least more robust to the discoveries they make.

Related to that, we often see rich and complicated models that are overfit. By that I mean, there were so many different variables tried as potential explanators, that some will be statistically signficant by chance alone. Five out of every 100 variables will show up as significant at the 95% confidence level, after all, even if they're not truly related. Of course, they do a crappy job out-of-sample when their true lack of relationship shows itself. Fortunately, honest statisticians can account for this.

He also recommended an out-of-print book that looks very promising, called Specification Searches, which does directly address questions of meta-statistics.

I'm kind of surprised that there isn't a larger meta-modeling "community of practice," if you will. I would think that there would be commonalities of approach/outcome that would warrant that type of community. But maybe the mathematics of statistics is in fact that community, and the conversation is primarily mathspeak.

But even there, I would expect to find sources that had zeroed in not just on mathematical modeling, but on the relevance/success of the outcomes: which types of modeling turn out to be most suitable for particular types of situations? I would expect there to be an organized endeavor devoted to after-the-fact model analysis, to try to assess effectiveness. I realize, of course, that there are no "right answers," as such, and that, even after the fact, not everyone agrees on the outcomes. And I suspect that after-the-fact modeling is not much less subject to bias than before-the-fact modeling, but that fact could be (and should be) part of the modeling analysis as well.

As with most things, though, I assume that bright people somewhere are pursuing this and better questions, and that I'm just not aware of their work. Or maybe this is a subject that necessarily contains too much stultifying recursion, sort of like epistemology (am I simply describing micro-epistemology, instead of macro-epistemology?), so it causes rational people to thrown up their hands in disgust.

My admittedly haphazard Google search on this subject reveals a discipline called "meta-statistics" which is primarily concerned with second-order statistical analysis: studies that aggregate other studies, essentially. I can certainly see the usefulness of that effort, but that's nowhere near the same thing as the kind of meta-statistics that I'm talking about.

Incidentally, I do know of, and highly recommend, one high profile "meta-statistician": Nate Silver, who writes the excellent blog Five Thirty Eight, now hosted by the New York Times. Silver is very good at reshuffling multiple pieces of data (in 538, he focuses on political polling) and commenting on the outcomes in ways that laypersons can understand. In fact, in retrospect, some amount of my curiosity about meta-analysis maybe be precisely due to the kind of writing Silver does on 538, which does ask, "how predictive was model x compared to model y?", and similar questions. Maybe another way of asking my question is: why isn't there more analysis/writing like Silver's out there in the world? Especially in a world that is growing ever more data-centric.

0 17 Aug 2011

remembrance of blogs past

Occasionally I go to write a post and I think, I've written this up before, and probably better, because I had more time and energy when I was younger and childless. The blog where I used to post things, however, is now defunct, though I still have all the content. In that spirit (aka, the spirit of laziness), I'm going to re-post a few things I liked from my old blog here. Greatest Hits, if you will. Today's RoBP:

"Ithaca or Bust" was our rallying cry that spring. Someone had told one of us that Ithaca, New York, home of Cornell University, was the hippest place on the planet, and we had bought into this notion without bringing to bear our usual finely tuned skepticism. It did sound great--it was back East, which has the mysterious allure of all things fine and proper (practically Europe, for all intents and purposes), and it was Ivy League, but minor Ivy League (Princeton or Harvard is just plain overkill), and it had a name that was Greek or Roman or something. Plus it was within striking distance of New York City, which we were starting to realize was the cultural center of at least the Americas, if not the Western Hemisphere.
The more serious students, we figured, could pursue their graduate educations at Cornell, while I would struggle through my ragtag undergraduate program at some college there (not Cornell, of course; it takes decent grades to get into even a minor Ivy League school). There was bound to be much discussion of very serious matters: ethics and aesthetics, the fine nuances of our personalities, and what this revealed about human nature in general. Clever and fascinating people were sure to enter our lives. We would be fine specimens of verve and vim and vigor, sipping our hot chocolate in coffeeshops and sparring verbally.
I don't remember whether anyone really made any serious inquiries into any part of this plan. Was any admissions application ever requested, or course catalog perused? Probably not. But it was a great plan.
Barring something that required so much extra work (requesting admissions applications and perusing course catalogs and the like), we soon decided that we needed to at least form a commune and move in together. This was a very big deal in those days, three of us having penises and one of us having a vagina. The powers that be took a very dim view of penises and vaginas living under the same roof. (What these powers didn't know, and what I didn't know until much later, was that penises could have just as much fun with other penises, and vaginas with vaginas, and that they did so with some frequency in that part of the world. Had they thought this problem all the way through, they probably would have insisted that every sexual organ live under its own roof, and that no one be allowed to visit anyone else's habitation. At least that way the greatest sin possible would be masturbation.) We had heard tales of people managing to pull a fast one on the Housing Office.
Like Ithaca, forming a commune also sounded great. It was vaguely sixties, which was cool, and very rebellious, which was extra cool. Plus we all had a good time together, so living together sounded downright fun. Those stay up late conversations about whether lampposts and washing machines were art could go on that much longer when our bedrooms were just down the hall.
This time, action was taken. The rental sections of newspapers were scanned for several days running. Some speculation about how to pull the wool over the Housing Office was undertaken. Could it really be that renting a house costs THAT much? Who has money like that? Discouragement quickly set in. We were people of very limited means, and it seemed likely that many landlords didn't have the kind of vision necessary to give us a reduced rate for being so forward thinking.
During one of the paper scanning sessions, however, someone noticed a mobile home for sale for one thousand dollars. Could that be a misprint? Unenlightened landlords were asking for nearly a thousand dollars every month, and here someone was selling a dwelling outright for the paltry sum of one thousand dollars. This seemed too good to be true. An investigatory committee was sent to Springville to determine whether we might have finally found a home for our young commune. The committee came back with a mixed report: yes, the mobile home was indeed for sale outright for a thousand dollars, but it wasn't in the best shape. Also, there was a monthly fee for parking our commune on a little strip of land in Springville. But what's a little plumbing and a leaky roof to a group of enterprising young people such as ourselves? How hard could it be? And, once having paid the initial thousand dollars, coming up with $110 every month would be cake. The matter was put to a vote, and the commune was officially born.
There was the small matter of the thousand dollars. As I remember, most of this sum came from Glen, though I don't remember exactly how he came by so much money. I have a vague memory that it may have come from the payoff on Glen's mother's yellow Mustang, which I totalled. I don't want to overstate my role in this matter, but I did sustain a very big bump on the head wrecking this car. Sacrifice is an easy thing, however, when the cause is noble.

1 15 Aug 2011

models and meaning

Models are a form of interpretive compression. Their value lies somewhere on a continuum between abstraction and specificity. Obviously, one end of the spectrum is not, per se, better than the other, but simply more useful, depending on one's purpose. I'm curious whether someone has done some analysis of modeling generally. Questions that leap to mind:

Can modeling be grouped? Are there standard model "types" (by which I mean, are there productive, albeit artificial, modeling categories)? Are some types more useful in particular contexts, or for particular purposes? It seems highly likely that our conceptual approach to our surroundings would in some way reflect lower order perceptual experiences, and that this, at the least, would suggest some ideas about modeling categories.
Are there second and third order levels of compression? In effect, are certain models more efficient, in that they cover more "space" (on the abstract/specific continuum) with less content? If so, then that quality, rather than simply being closer to the abstract end of the spectrum, would seem to be the best definition of elegance.
Is "model" simply too abstract a concept to allow for that kind of analysis? Would certain principles of formality (efforts and defining and investigating completeness and consistency) be required before you could seriously investigate modeling? Or, to put it another way, do you have to go all the way to axiomatic systems to perform useful analysis of modeling systems? Is anything less simply an exercise in aesthetics?
I assume that compression, in modeling terms, is always lossy. Could that possibly be false?
Are these just ordinary questions of theoretical mathematics?

Anyone know anything about this? Know any good books you could point me towards?

0 5 Aug 2011

Cory Doctorow's wonderful defense of public education

Education is a public good. It is best supplied and paid for by the group as a whole, because no individual or small collective can produce the overall social benefit that the nation can provision collectively.
Education doesn't respond well to market forces because many of the social goods that arise from education -- socialization, a grounding in civics, historical context, rational and systematic reasoning -- are not goods or services demanded by a market, but rather they are the underlying substrate that allows people to intelligently conduct transactions in a marketplace as well as establishing and maintaining good governance.
There is a long and wide body of evidence that people with wide, solid educational foundations that transcend mere vocational skills produce societies that are more prosperous, more transparent, healthier, more democratic -- that attain, in short, all the things we hope markets will attain for us.
There are innumerable "market based" educational options. We call them private schools. A few of them are very good, and are generally priced out of the reach of average people. Many of them are very poor indeed. But functional democracies require that all people -- not just those who are already wealthy -- are given the foundational knowledge that allows them to prosper and participate in the full range of social activities that make nations great.

-Cory Doctorow, in the comment section of one of his own BoingBoing posts (via

Read the rest of this post »

0 31 Jul 2011

maddening

There are so many things that are maddening about the Obama presidency that it's hard to know where to start. But this latest "deal" exemplifies the big trend: Obama, in the interests of being moderate, starts out every bargaining situation by splitting the difference between Democratic and Republican positions (i.e., gives up a great deal of his leverage from the get-go). Republicans gasp and declare that Obama is an evil socialist and they move their bargaining position significantly to the right. Rinse, repeat, ad nauseum. Eventually, Obama moves to the right of the Republicans' ORIGINAL bargaining position, with Republicans still harrumphing and talking about impeachment. Eventually a terrible deal gets cut, with Democratic legislators now having their hands tied lest they look like the obstructionists. The end result: Republicans get what they want, but without having to take responsibility.

Even though the anti-Keynsian, pro-business viewpoint of the Republicans will do still more damage to our already gasping economy (with most of the damage concentrated on the middle class and the poor), you can bet that during the 2012 election, Republicans will run firmly on the notion that it's the Democrats who have ruined the economy. And what can the Democrats say in response? That they held the presidency and the Senate yet managed to pass some of the most right-leaning legislation since the Hoover era? They got nothing. While I would absolutely positively NOT vote for a Republican in the next election, if they win, perhaps it will lead to at least one positive thing: their disastrous policies will at least be enacted in their own name, instead of in the name of the Hope/Change Democrat.

1 31 Jul 2011

Worst. Deal. Ever.

I never thought I'd say this, but I really think Obama just lost my vote in 2012. His ridiculous "centrism" amounts to a willingness to cave in to greedy, corporate interests (read: Republicans) at every turn. This unholy alliance between Obama and the Republicans is government of the rich, by the rich, for the rich.

0 22 Jul 2011

all your jobs are belong to us

Very interesting piece about the relationship between unemployment in the U.S. and general economic trends related to globalization (the NYT piece is a summary of a report prepared by the Council on Foreign Relations). The short version: globalization of the economy (the ability to produce goods in pieces across several countries) is causing less skilled American workers to be in greater competition with workers from emerging economies (i.e., China, India, Brazil, etc.), who are generally willing to work for considerably less. There is a segment of the economy that is immune to those kind of pressures, which the authors of the CFR report call the "nontradeable" sector, and most of the recent employment gains in the U.S. come from that sector. Unfortunately, two of the largest segments of the "nontradeable" sector are government and healthcare. In other words, the very things that are high profile targets for both short-term and long-term budget cuts.

Intellectually speaking, I can't fault anything about this trend. For a long time, economic powers like the United States have been willing to increase their standard of living at the expense of the third world (particularly in the forms of ecologically damaging resource extraction and labor exploitation). Who are we to squawk about the downward wage pressures that result from those same countries becoming more viable players in the international economy?

What I find disturbing, though, is that this report (to the extent that it's an accurate forecast) seems to augur a continuing wealth gap in the United States. The very well off, people on an executive pay scale, will continue to reap the benefits of global expansion. The downward wage and employment pressure is mainly on the unskilled and less skilled labor segments (though certain kinds of skilled labor are far from immune; witness the increase in software engineering being outsourced to India, for example). Or, to put it in bleaker terms: the pain being felt by the American middle class isn't likely to go away anytime soon, even if the economy does rebound.

http://economix.blogs.nytimes.com/2011/07/22/is-there-hope-for-the-unemployed/?hp

0 13 Jul 2011

my new tolerance system for reviewing media

Not that I write that many reviews, but if I did...here's the system I would use henceforth:

Albums. Judged on a scale of 50 stars, where this is the basis for judgment: The number of stars an album earns depends on the number of times I'm able to listen to and enjoy a particular album. If I can't make it all the way through an album and enjoy it, it gets zero stars. To get 50 stars, I must be able to listen to an album 50 times and enjoy it every single time. They don't have to be consecutive listens. Thus, an album can grow in stature over time. Maybe there should be another scale for density, or a time limit? How many times does an album hold up in the course of one year, let's say.

Movies. How many times does a movie hold up? I think that my rough tolerance for a movie, most movies anyway, is around 10 viewings. So this one will be a 10-star system, where 10 stars go only to movies that both invite and withstand 10 pleasurable viewings (where pleasure is a complex metric for me that can include things like being deeply disturbed, but in a way that provokes me to further contemplation). Or maybe the number of stars should be open-ended, and include life time viewings. Thus, there could be 50-star movies, but only for people with way too much time on their hands. There have been movies I've watched over a dozen times. I haven't kept track exactly. I'll have to ponder that one. I think, though, that this system is biased against "difficult" movies. I might have watched Tarkovsky's Stalker only twice, while having seen The Matrix 10 times, but that doesn't make The Matrix a 10-star movie and Stalker a 2-star movie. Or does it? Is my estimation of Stalker belied by my unwillingness to watch it 10 times? Perhaps this is the great equalizer that undermines the pretentiousness factor. Because I have watched Citizen Kane quite a few times, and enjoyed it every time.

TV Shows. A TV show should be judged on the basis of a season, I think, rather than simply by episode. Since this is a considerably bigger time commitment, depending on the series, this one should be judged on a 5-star scale, where each star represents watching an entire season. A 5-star TV series can withstand five complete viewings of an entire season. Granted, this does tend to favor series with shorter seasons, such as the 12- and 13-episode seasons used by HBO and Showtime and, most recently, AMC. As with the movies, I'm leaning towards leaving this open-ended. Of course, many, many TV series wouldn't earn even one star under this system, but that seems perfectly fair.

Books. We're moving here towards the long and difficult end of this scale, and perhaps this kind of system doesn't work so well for books. In principle, I think it does. Good books invite and withstand multiple readings without running out of steam. The tricky part is that the investment required to read a book can vary so wildly. Reading The Little Prince is a somewhat different experience from reading Infinite Jest. So I'm thinking that with this one, rather than gauging stars on a times read basis, perhaps we should lean more towards a flat number of hours devoted to reading a particular book. Thus, if you can stand to read The Red Pony 100 times and enjoy it every time, this could be equal to having tackled Remembrance of Things Past only once, but having made it all the way through. One star per hour seems a bit excessive here, however. Perhaps one star per five hours of pleasurable reading. Thus, a book that's short can still accumulate stars by inviting multiple readings. Similar to movies, this system suffers a bit from the bias towards "easy" books. And perhaps, like movies, it undoes some of the factor of pretentiousness. Though it seems entirely possible that I make my way through some books mostly to be able to say that I've done so, more than because I'm really, truly drawn in by them. Mostly, though, in those cases, I'm unable to force myself all the way through. Thus I have long believed that I should read (and enjoy) Ulysses, whilst never having finished it once.

The general idea here is that "good" is synonymous with interested/engaged, and that interested/engaged can be judged somewhat by investment. Of course, a system like this could never work for some people who would never watch the same movie or read the same book twice, but that isn't me. I relish art that calls for repeated engagement, that has enough depth to keep giving with repeated engagement.

What do you think, reader? What's your system for measuring aesthetic value?

not just my opinion

0

21 Sep 2011

The real class warfare...

0

9 Sep 2011

reviews

0

7 Sep 2011

Models and Meaning, Part 2

0

17 Aug 2011

remembrance of blogs past

1

15 Aug 2011

models and meaning

0

5 Aug 2011

Cory Doctorow's wonderful defense of public education

0

31 Jul 2011

maddening

1

31 Jul 2011

Worst. Deal. Ever.

0

22 Jul 2011

all your jobs are belong to us

0

13 Jul 2011

my new tolerance system for reviewing media

Search

Archive

Contributors

not just my opinion

Get Updates