breaking culture

I missed this terrific essay on writing pedagogy (and machine scoring) Marc Bousquet wrote back in April.  He builds a very insightful argument, first pointing out that human scorers aren’t all that more “human” in their grading process.  On this count, I was glad to see him cite Dan DiMaggio’s 2010 Monthly Review essay which gives a vivid account of their Turneresque factory logic, e.g.

Scoring is particularly rushed when scorers are paid by piece-rate, as is the case when you are scoring from home, where a growing part of the industry’s work is done. At 30 to 70 cents per paper, depending on the test, the incentive, especially for a home worker, is to score as quickly as possible in order to earn any money: at 30 cents per paper, you have to score forty papers an hour to make $12 an hour, and test scoring requires a lot of mental breaks. Presumably, the score-from-home model is more profitable for testing companies than setting up an office, especially since it avoids the prospect of overtime pay, the bane of existence for companies operating on tight deadlines. But overtime pay is a gift from heaven for impoverished test scorers; on one project, I worked in an office for twenty-three days straight, including numerous nine-hour days operating on four to five hours sleep—such was my excitement about overtime. 

40 papers/hour.  That means a little more than 90 seconds reading a paper and deciding what grade it should get.  If a machine can do it in 5, would it really be that significant a loss to the “humanity” of the process? Both DiMaggio and Bousquet also cite Todd Farley’s series of revelations about his time as a test scorer.  

The more important question - as the quote above points out - is whether the real problem lies in the way we are teaching writing because of these tests, and the effect this has on students’ conception of their abilities.  The last half of Bousquet’s essay draws this problem out using Rebecca Moore Howard’s research on students’ citation use in college research papers - assignments that are supposedly far removed from the mechanistic high school writing exams DiMaggio and Farley were scoring.  Howard’s research finds that students generally “patchwrite” their papers, meaning that they simply string together a set of badly paraphrased abstracts or excerpts from other sources without clearly understanding that source material.  

As she and her co-author explained to the folks at Project Information Literacy last August, the project began as a way to look at growing concerns about plagiarism in the writing of college students.  But as they worked through the writing samples they were analyzing (using, they note, human coders) they found something even more surprising:

most of the citations were for material that was either quoted or paraphrased. If your focus is on procedure and correct format, these papers are a great success. But if you look at this another way and remember that for most of us, “research” is about the discovery of new information and ideas, and the synthesis of those ideas into deeper understanding, the majority of the papers failed. Only 6% of the citations are to summarized material. It is in summary that writers demonstrate comprehension of the larger arguments of a text, working from ideas rather than sentences. And in the papers we studied, students are not doing that. Further, 46% of the citations are from the first page of the source in question. Yes, that really is 46%, and a full 70% come from somewhere in the first two pages (1,328 citations from a total of 1,911 that we coded). The majority of the sources are cited only once, and only a handful of the papers cite any source in a way that suggests the student was engaging with the entire text. 

Their work is evolving, but I agree with Bousquet’s recommendation that a serious interaction with the field is essential to any academic writing - not only would it help students create better, more relevant summaries and arguments, but it would also help them see the reason for the entire enterprise.  On this count, the real punchline of Bousquet’s piece is that it isn’t just underfunded schools, overworked teachers, and robot graded tests that are stunting our student’s writing and critical thinking skills, the entire culture writes and thinks in this “smash and grab” soundbite way (e.g. newspaper editorials, cable TV) and the political economy of higher ed virtually ensures that the people responsible for giving students a basic education in doing something otherwise are as overworked, underfunded, and inexperienced as the system will allow.  He ends with this targeted critique of the critique of robot grading, one fitting with his understanding of How the University Works.

If we are to do more with writing classes and writing assignments, we need to put aside the hysteria about machine grading and devote our attention to the mechanical teaching and learning environment in which we daily, all but universally, immerse our writing faculty.  We need to change the kind of writing we ask them to teach. We need to enable writing faculty to actually do the kind of academic writing they should be teaching–which means changing our assumptions about how they’re appointed, supported, evaluated and rewarded. You want to be a machine-breaker and fix writing pedagogy? Great. Start with with your professional responsibility to address the working circumstances of your colleagues serving on teaching-only and teaching-intensive appointment.

This statement appeared in a Business Week article about Amazon’s challenge to mainstream publishing and the role of print-on-demand (POD) in that strategy.  While it has the breezy know-it-all tone of the business press, I don’t find their analysis of why big publishers aren’t doing POD to be compelling.  I don’t have the numbers myself, but to make their case, they would have to do something like compare publisher profits to the number of books they remainder or destroy from big print runs.  Otherwise, the story just comes off as what it is: empty, tech-fetishizing cheerleading for Amazon.

In any case, it misses the much larger elephant in the room, focusing only on the production of commodities rather than the production of the means of production of commodities.  In other words, the article focuses solely on the circulation of printed books rather than the extensive process necessary for a printed book to enter circulation - including but not limited to the very large machines that make it possible to sell books at wholesale prices and still make a profit.  I don’t know much about these machines, I admit, but part of what annoys me about the article is that it doesn’t provide any comparison between the productivity of the machines used in POD and that of traditional presses.  Instead it opts for looking at POD as the cure for all that ails mainstream publishing - a cure they are refusing to take because it would threaten their power.  This may be true in some way, but it may also be true that the POD option isn’t economical for mainstream publishing in the first place.

I think it is far more likely that big publishers don’t want to use POD because, on the one hand, the machines they use are likely far more efficient at printing large runs and; on the other hand, they have invested a lot in these machines and the distribution network that makes them profitable.  Big publishers don’t want to make a switch to small batch, POD publishing because they are, by definition, big publishers.  

In general, POD puzzles me.  It enters the debate about the future of publishing in an odd place.  Most people gunning for innovation in book publishing focus their energy on digital only, e-books and other e-reading platforms that will replace the printed codex as the dominant delivery mechanism for long form content.  POD seems to say all that is true, but people will likely continue to want printed books for whatever reason so we should still have a way to print books.  And for small runs or books that have a really small and/or dispersed market, it makes a lot of sense.  For the next Stephen King pulp title, not so much.

I’ve never seen any analysis about publishing that says their biggest overhead is in overprinting, warehousing, and distribution, but if these were the primary problems (or if the current solutions weren’t working) I can’t imagine they wouldn’t be doing more about it.  On the other hand, the article itself opens by noting that Amazon remains pretty invested in the warehousing and distribution of books.  In this sense, POD appears less as a model of efficiency than as one of Amazon’s strategies for distributing “the long tail.”  It is unlikely, for instance, for it to be efficient for Amazon to do POD for big bestsellers that would require huge print runs in a limited amount of time.  

I’m a little fuzzy on the larger scale POD operations mentioned in the article CreateSpace is an Amazon affiliate that will distribute independent authors’ via Kindle or POD print books.  It also aids indie musicians and moviemakers in online distribution via Amazon. Lightning Source specializes in printing small runs of books for small presses, and appears to have some of the enormous presses that likely sit in the warehouses of major publishers.  In each of these cases, the POD is only part of the bargain.  They also help some with marketing (i.e. it appears on Amazon’s webpage), distribution, and in some cases design.  As major publishers would also point out, they don’t offer anything in the way of editorial guidance or real direct marketing to bookstores or libraries, but they are still rather large scale and “full service” as publishing operations.

This is especially true when the article put these in the same category as the Espresso Book Machine (EBM), which is the perennial hero of POD stories.  It was launched almost a decade ago, though the actual machine didn’t emerge until 2007, when it was named by Time as one of the best inventions of that year. It’s is now manufactured (and serviced) by Xerox.  They started out at about $50k a pop (according to the Time article above) but now run about $100k. They will produce a book - complete with binding and full cover color - from a digital file in a little under 5 minutes, depending on the size of the book.  Video of the machine in action here.

It’s impressive as a machine, but has taken a long time to catch on.  It had only sold about 30 of the machines by 2009 and it’s parent company at the time, On Demand Books, had hoped to sell another 60 that year, with sales boosted by a partnership with Google’s public domain book search.  According to Rick Anderson, writing last August, there were only 45 machines; other sources claim there are about 70, but the website for the company currently shows it remains under 60.  Anderson claims that the low number of titles available through digital vendors limits its uptake, but he also cites a lot of technical problems they’ve had with the machine at the University of Utah’s Marriott Library, where Anderson works.  Alongside static electricity problems that the company worked out when Brigham Young University bought one, he mentions that,

It takes about five minutes to print a 300-page book — as long as the machine is warmed up. If it isn’t, you’ve got to let the glue melt, which will take 45 minutes to an hour. [… .] we had issues with balky and leaking ink jets, malfunctioning sensors, and recalcitrant cover feeds, all of which have been fixed or mostly fixed at this point. And we’re still waiting for a color text-block printer that will communicate effectively with the machine, despite having been offered that option initially. For now we’re still making do with black-and-white (we can print covers in color without any problem) 

Many of these bugs will get worked out to some degree the more people adopt them.  On the other hand, as anyone with access to an office copier can tell you, machines like these often require a lot of maintenance to keep up with the use.   Still, it doesn’t make sense to speak about the EBM in the same breath as industrial size initiatives like Lightning Source.  And it is hard to see either as viable alternatives to the scale of major printing operations of mainstream publishers.  Maybe I’m wrong, but these sound like apples and oranges.  

Certainly for some publications, for instance academic monographs, it makes a good deal of sense if it is also paired with digital distribution.  Adding currently out of print titles would also be a fantastic gift - as the Utah Library found when they were able to print out a 300 year old German manuscript for a faculty member there.  In other words, anything that couldn’t be efficiently printed at scale.  But almost by definition, none of these are the purview of the major houses Business Week claims are challenged by this model.  

While it is just one story, it is of a type in that it looks for disruptive technology as the leading edge of cultural and social change, rather than thinking of it as a part of a total strategy that would make little sense outside of that strategy.  At the same time, it ignores far more significant social and cultural disruptions.  For instance, is it more important that Amazon offers print on demand or that Amazon offers authors a 70% royalty rate for Kindle books?  One could argue that they aren’t all that distinct.  If an author was reticent about signing onto Amazon’s platform for that price - say, for instance, because they also want their book in print - having the print on demand is a nice backup.  If your ultimate goal is to get out of the pulp business altogether - and slash those labor costs for warehousing and shipping - then POD is useful to entice more first time authors to the platform.  On the other hand, if your business is the pulp business, it hardly makes sense to try retrofitting your infrastructure after the fact.

None of this points to the real disruptions that these emergent models suggest: namely,  is there still need for editorial input in the publishing process? And what are the archival and historical consequences of a model like Amazon’s POD taking hold? Both of these are questions with widespread implications, ranging from mainstream publishing that is the subject of this article, to scholarly communication, which is under increasing pressure to be open source.  Having a big stack of print books may be less and less important for bookstores, libraries, and publishers in the future, but this will not eliminate the extensive human input that goes into creating, curating, and preserving content.  

Right now, the argument remains superficially focused on technologies that will likely never catch on completely.  It should be focused on instances where Amazon - or other publishers - help to create alternative institutions that fulfill these more essential functions - functions that often cost a lot more than amortize in the short term and often continue to cost money long afterwards.  The longer we think about this as a decision to be made by clashing industrial titans instead of broad social questions requiring deliberation and concerted action, the less the answer we get will reflect anything approaching a thoughtful solution for the future.