Humane/Anti-Humane

Though he doesn’t state it directly, Timothy Burke recognizes that humanistic inquiry circa 2013 is at risk of being subsumed—enfolded into—techno-scientific discourse and scrutiny. In many venues, as he notes, it has already been subsumed:

I would not call such views inhumane: more anti-humane: they do not believe that a humane approach to the problems of a technologically advanced global society is effective or fair, that we need rules and instruments and systems of knowing that overrule intersubjective, experiential perspectives and slippery rhetorical and cultural ways of communicating what we know about the world.

The anti-humane is in play:

–When someone works to make an algorithm to grade essays

–When an IRB adopts inflexible rules derived from the governance of biomedical research and applies them to cultural anthropology

–When law enforcement and public culture work together to create a highly typified, abstracted profile of a psychological type prone to commit certain crimes and then attempt to surveil or control everyone falling within that parameter

–When quantitative social science pursues elaborate methodologies to isolate a single causal variable as having slightly more statistically significant weight than thousands of other variables rather than just craft a rhetorically persuasive interpretation of the importance of that factor

–When public officials build testing and evaluation systems intended to automate and massify the work of assessing the performance of employees or students

At these and many other moments across a wide scale of contemporary societies we set out to bracket off or excise the human element , to eliminate our reliance on intersubjective judgment. We are in these moments, as James Scott has put it of “high modernism”, working to make human beings legible and fixed for the sake of systems that require them to be so.

That humans can be quantified and their behaviors inserted into mechanistic or, more recently, statistical models is an idea as old as Comte and Spencer. Humanists of all stripes, religious and secular, have long denounced this idea, but with each passing decade, their denunciations have been met with more and more techno-scientific intrusions into the venues of humanistic inquiry. Researchers in China are currently attempting to map the genetic architecture of human intelligence itself; natural language processors are attempting to teach computers to learn human languages; and researchers across a wide array of disciplines continue to produce research which suggests that all life—human life included—is essentially “a mixture of genes, environment, and the accidents of history.“

E.O. Wilson’s Sociobiology may have been problematic in tone and too over-confident, but its underlying idea—that everything human and humane will eventually be describable in techno-scientific terms—is as valid as ever. Kurzweil is far too optimistic about the speed of transhuman advancement, but if history has shown us one thing, it’s not to bet against scientific advancement. And if we do build a sentient machine or use machines to amplify the abilities of humans (aren’t we doing this already?), then what else can we conclude but that humanity can be meaningfully reduced to techno-scientific terms?

Timothy Burke writes that “a humane knowledge accepts that human beings and their works are contingent to interpretation,” but some interpretations are more productive than others. Western science may be a reductive way-of-knowing, and techno-scientific reductionism may be a tough interpretation for humanists to find value in, but in the post-Enlightenment marketplace of ideas, anti-humane knowledge has always been and will continue to be the driver of discourse. Why? It produces. Its material applications are powerful.

Most of Burke’s essay is nuanced and generative, but he concludes with a bit of a rhetorical flourish that undermines the continuing material productivity of science:

We might, in fact, begin to argue that most academic disciplines need to move towards what I’ve described as humane because all of the problems and phenomena best described or managed in other approaches have already been understood and managed. The 20th Century picked all the low-hanging fruit. All the problems that could be solved by anti-humane thinking, all the solutions that could be achieved through technocratic management, are complete.

If by “anti-humane thinking” Burke means a purely mechanistic view of humanity, then he’s probably right. However, no one holds such a view anymore, if they ever did. For example, machine learning is modeled on statistical probabilities, and genetic research is looking at complex polygenic and epigenetic effects that are not reducible to single gene tinkering. The techno-scientific lens is no longer mechanistic or averse to complexity, but I still get the feeling that it remains “anti-humane” in the eyes of many humanists.

The worst thing the humanities can do is to continue theorizing about how its subject matter simply cannot be subsumed by techno-scientific practice while its subject matter continues to be subsumed by techno-scientific practice. We need to stop talking about, say, “the social construction of gender and sexuality” as though representation and discourse were more important for understanding gender and sexuality than hormone therapy or the biology of same-sex reproduction. Too often, humanists confuse ethical critique with epistemology. In my opinion, instead of assuming that our areas of inquiry are by definition off-limits to the techno-scientific lens, we need to recognize that the humanities are indeed “incomplete” without recourse to the knowledge of science. We should cross the border into the Land of Techno-Science more often . . . . for this sets up an encounter in which the sciences will recognize that they, too, are “incomplete” without recourse to the knowledge of humane inquiry. Every discipline has its deflections.

True incommensurability is rare. I’m confident that reading, e.g., E.O. Wilson and Donna Haraway together could be productive—so long as neither work remains unchanged in the encounter. The encounter would not be about whose knowledge gets to be the base of the other’s, or whose knowledge anchors the other’s. Rather, the point of such a humane/anti-humane encounter would be to give birth to an epistemological offspring comprised of elements from both but resembling neither.

Robo-Graders

I was wrong about the mechanization of student writing. I had assumed another year or two would pass before MOOCs began utilizing essay grading software. Turns out it’s happening now. EdX, founded by Harvard and probably the most prestigious online course program, has anounced that it will implement its own assessment software to grade student writing.

Marc Bousquet’s essay successfully mines the reasons why humanities profs are anxious about algorithmic scoring. The reality is, across many disciplines, the writing we ask our students to do is “already mechanized.” The five-paragraph essay, the research paper, the literature review . . . these are all written genres with well-defined parameters and expectations. And if you have parameters and expectations for a text, it’s quite easy to write algorithms to check whether the parameters were followed and the expectations met.

The only way to ensure that a written product cannot be machine graded is to ensure that it has ill-defined parameters and vague or subjective expectations. For example, the expectations for fiction and poetry are highly subjective—dependent, ultimately, on individual authors and the myriad reasons why people enjoy those authors. It might be possible to machine grade a Stephen King novel on its Stephen-King-ness (based on the expected style and form of a Stephen King novel), but otherwise, it will remain forever impossible to quantitatively ‘score’ novels qua novels or poems qua poems, and there’s no market for doing that anyway. Publishers will never replace their front-line readers and agents with robots who can differentiate good fiction from bad fiction.

However, when we talk about student writing in an academic context, we’re not talking about fiction or poetry. We’re talking about texts that are highly formulaic and designed to follow certain patterns, templates, and standardized rhetorical moves. This description might sound like fingernails on a chalkboard to some, but look, in the academic world, written standards and expectations are necessary to optimize for the clearest possible communication of ideas. The purpose of lower division writing requirements is to enculturate students into the various modes of written communication they are expected to follow as psychologists, historians, literary critics, or whatever.

Each discourse community, each discipline, has its own way of writing, but the differences aren’t anywhere near incommensurable (the major differences exist across the supra-disciplines: hard sciences, soft sciences, social sciences, humanities). No matter the discipline, however, there is a standard way that members of that discipline are expected to write and communicate—in other words, texts in academia will always need to conform to well-defined parameters and expectations. Don’t believe it? One of the most popular handbooks for student writers, They Say/I Say, is a hundred pages of templates. And they work.

So what’s my point? My point is that it’s very possible to machine-grade academic writing in a fair and useful way because academic writing by definition will have surface markers that can be checked with algorithms. Clearly, the one-size-fits-all software programs, like the ones ETS uses, are problematic and too general. Well, all that means is that any day now, a company will start offering essay-grading software tailor-made for your own university’s writing program, or psychology department, or history department, or Writing Across the Curriculum program, or whatever—software designed to score the kind of writing expected in those programs. Never bet against technology and free enterprise.

And that’s another major point—there’s not a market for robot readers at publishing firms, but there certainly is a market for software that can grade student writing. And wherever there’s a need or a want or some other exigence, technology will fill the void. The exigence in academia is that there are more students than ever and less money to pay for full-time faculty to teach these students. Of course, this state of affairs isn’t an exigence for the Ivy League, major state flagships, or other elite institutions—these campuses are not designed for the masses. The undergraduate population at Yale hasn’t changed since 1978. A few years ago, a generous alumnus announced his plans to fund an increase in MIT’s undergraduate body—by a whopping 250 students. Such institutions will continue to be what they are: boutique experiences for the future elite. I imagine that Human-Graded Writing will continue to be a mainstay at these boutique campuses, kind of like Grown Local stickers are a mainstay of Whole Foods.

For the vast majority of undergraduates—those at smaller state colleges, online universities, or those trying to graduate in 4 years by taking courses through EdX—machine-grading will be an inevitable reality. Why? It fulfills both exigencies I mentioned above. It allows colleges to cut costs while simultaneously making it easier to get more students in and out of the door. Instead of employing ten adjuncts or teaching associates to grade papers, you just need a single tenure-track professor who posts lectures and uploads essays with a few clicks.

So, the question for teachers of writing (the question for any professors who value writing in their courses) is not “How can we stop machine-grading from infiltrating the university?” It’s here. It’s available. Rather, the question should be, “How can we best use it?”

Off the top of my head . . .

Grammar, mechanics, and formatting. Unless we’re teaching ESL writing or remedial English, these aspects tend to get downplayed. I know I rarely talk about participial clauses or the accusative case. I overlook errors all the time, focusing instead on higher-order concerns—say, whether or not a secondary source was really put to use or just quoted to fill a requirement. However, I don’t think it’s a good thing that we overlook these errors. We do so because there are only so many minutes in a class or a meeting. With essay-grading software, we can bring sentence-level issues to students’ attention without taking time away from higher-order concerns.

Quicker response times for ESL students, and, perhaps, more detailed responses than a single instructor could provide, especially if she’s teaching half-a-dozen courses. Anyone who has tried to learn a second language knows that waiting a week or two for teacher feedback on your writing is a drag. In my German courses, I always wished I could get quick feedback on a certain turn of phrase or sentence construction, lest something wrong or awkward get imprinted in my developing grammar.

So, I guess my final point is that there are valid uses for essay-grading software, even for those of us teaching at institutions that won’t ever demand its use en masse. Rather than condemn it wholesale, we–and by we, I mean every college, program, professor, and lecturer–should figure out how to adapt to it and use it to our advantage.

Building a Chinese Room

Chomsky isn’t a fan of statistical machine learning. However, this video (via Steve Hsu) suggests that using Really Big Corpora is the best way to get machines to figure out how language works, both structurally and–as the video shows–phonetically and acoustically.

Around six minutes in, the demonstration begins. The speaker’s words are translated almost instantaneously into Chinese, and the auditory output sounds somewhat similar to the speaker’s actual voice. There are obviously Chinese speakers in the audience, and their response suggests that the demo was successful.

This video is a good example of the ways that computer scientists (and I include researchers in natural language processing in that category) are operating squarely in the realm of the humanities–what’s more humanistic than language translation? There have been tomes and manifestos written unto its spiritual, social, epistemological, and theoretical nature. And now computers are getting the hang of it. We humanists ignore their successes at our peril.

Research Soundtrack 3

Digital maps and social data

In the 80s and 90s, critical cartographers, such as J. B. Harley, reminded us that the map is not the territory. A map is always a representation, a construction, designed by humans to show certain things and to not show other things. The critique was elementary. Fifty years earlier, Borges had acknowledged much more creatively the map/territory distinction in his story “Of Exactitude in Science”:

. . . In that Empire, the craft of Cartography attained such Perfection that the Map of a Single province covered the space of an entire City, and the Map of the Empire itself an entire Province. In the course of Time, these Extensive maps were found somewhat wanting, and so the College of Cartographers evolved a Map of the Empire that was the same Scale as the Empire and that coincided with it point for point. Less attentive to the Study of Cartography, succeeding Generations came to judge a map of such Magnitude cumbersome . . .

A map that fully represents the territory is simply the territory re-created. How else would maps work without selecting certain elements of a territory (elevation countours, for instance) and deflecting everything else?

As with most things ‘critical,’ critical cartography was really not an epistemological critique but more of an ethical and political critique of mapping. It asked us to consider who and what gets ‘left off’ officially sanctioned maps, and warned us not to believe that maps provide a “view from nowhere,” to use Donna Haraway’s terminology, a god’s-eye-view that provides a totalizing and completely objective picture of a material space. What gets deflected in cartographic practice is often the economic, social, and political realities on the ground, and these are important elements of a space.

With this critique in mind, let’s turn attention to Google Maps and Google Earth. While a topographic map is obviously ‘not the territory,’ Google Earth provides real pictures of the territory from varying heights—not the territory precisely, but a lot closer to it than a topographic map. Digital maps also provide more than a static representation. While the selections (and thus deflections) of a paper map are set in stone, so to speak, Google Maps and Google Earth allow for fluid and ever-expanding selections, so that the deflections inherent in one selection can themselves be selected, and thus made visible, with the click of a mouse or a simple download. The only limit to what can be selected and visualized is the availability of data—and data abounds.

By design, paper maps leave out many elements of a material space; by design, Google Earth and Google Maps can return many elements to the map, including social, political, economic, climatological, biologic, and many other elements. Their interfaces are designed for Maximum Selection. Each selection is finite and incomplete, of course, but digital mapping interfaces allow multiple finite selections to be layered onto one another, compared, contrasted, and mined for patterns. Each selection is a partial perspective—similar to the “situated knowledge” discussed by Haraway—but by bringing together the partial perspectives with the help of a digital interface, one can move closer to an objective view of a material space, not a “view from nowhere” but a view from many different places and perspectives all at once.

The selected elements of a topographical map of a city show us many things—roads, toponyms, elevations, city and county boundaries. However, they also deflect many more things that might be interesting to visualize on a map—for example, class boundaries or concentrations of poverty.

To map something like poverty requires not only a traditional map interface, which represents territory, but also data about the socio-economic conditions within that territory. A poverty map will always, therefore, be a mashup of spatial and non-spatial data (as will any map that attempts to bring together situated knowledge of or multiple perspectives on a material space). The precise socio-economic data one uses to define poverty may or may not change the nature of the map, but the utility of digital mapping interfaces is that they allow users to deploy multiple data sets and compare them with relative quickness and ease.

This simple map shows states in different shades of red, corresponding to percentages of individuals living below the federal poverty line, supplemented with information about statewide median income and percentages of individuals lacking health insurance.

Here’s a map that shows the same thing at a more granular scale, tracking poverty percentages within state boundaries as well as between them.

And here’s a map of London socio-economic data—based on the UK’s multiple deprivation index—that also plots where riots occurred in 2011. The map shows a clear correlation (one that brings together spatial and non-spatial data) between the location of riots and the location of deprived communities.

The ways we might map poverty (or any other non-spatial data typically deflected by traditional cartography) are nearly endless. A lack of data is our only barrier, but data in the digital age is rarely in a state of lack.

Here’s another way, then, to map poverty and class boundaries—I haven’t seen anyone do this before, so it’s a good example of the endless ways to mash up data in a digital map interface.

Using Zillow, I gathered the addresses of the dozen or so most expensive homes for sale in two different cities: Los Angeles and New Orleans. Then I gathered the addresses of public housing units in both cities. Using Google Maps, I plotted all the addresses to see where they existed in proximal space. What I found were two antithetical visions of American class boundaries. The blue points are the most expensive homes; the pink points are the housing projects.

View Larger Map

View Larger Map

In New Orleans, the expensive homes stand near Tulane University and stretch in a U-shape from there along St. Charles Avenue. However, these homes are still relatively close to public housing sites. (One multimillion dollar home stands a few blocks away from a notoriously violent housing project.) Only two projects are far-flung from the expensive homes, and even then, the distances pale in comparison to the distances found on the Los Angeles map.

In L.A., it appears, millionaires gather in the northwest, into which the road grid does not extend, miles from the start of the urban sprawl and a good drive from the eastern and southern ends of the city, where the housing projects cluster. Bringing up Google Earth, we can zoom downward to this rich area to achieve a more topographical and thematic view . . .

Find some non-spatial data. Plot it onto a digital mapping interface. Interesting results? Rinse and repeat. Start building a theory.

For example, I think the above maps suggest two types of class boundaries, which I’ll call the Grid Boundary and the Fortress Boundary. New Orleans represents the first type: the separation of the very poor and the very rich occurs within the urban grid. Spatial distances are much smaller than distances of income and wealth. In these situations, you have ‘no-go zones,’ neighborhoods and streets where You Just Don’t Travel After Dark, and crime is a function of street rather than zip code or town. Within the Grid Boundary, poverty remains visible to the rich, and poverty’s effects will hit closer to home because, quite literally, they are closer to home (even though not felt or experienced directly).

Los Angeles, on the other hand, represents a Fortress Boundary: the separation of the very rich and the very poor occurs beyond the urban grid. Spatial distances are as large as, if not larger than, distances of income and wealth. In these situations, members of the upper class need not ever be physically near members of the lower classes. Crime is a function of zip code or town, so it never hits close to home for the rich because crime occurrs nowhere near their homes. Within the Fortress Boundary, poverty is largely invisible to the rich; they pass over it on freeways or avoid it altogether.

So where do we go from here with this nascent theory? Again, the only obstacle is a lack of data, and there’s rarely a lack of data. For one, we could use the same method (plotting public housing and most expensive homes for sale) in different American cities to see whether most correspond to a Grid Boundary or a Fortress Boundary or something else altogether. Do America’s upper classes more often gather into fortresses cut off from the worst levels of crime and poverty, or do they live in and amongst crime and poverty? We could also plot homes for sale at median levels in order to see where the middle classes fit within these boundary structures.

With these mashed-up maps in hand, we could start adding layers of data that we have hitherto been deflecting—for example, voting patterns among the upper classes who live in Fortress Boundaries versus those who live in Grid Boundaries. The L.A. Times, fortunately, has already provided a granular visualization of L.A. county voting patterns in the 2012 presidential election. Here are some screenshots of areas corresponding to the two housing clusters (most expensive and public) seen above:

Voting patterns in the area of most expensive homes

Voting patterns in the public housing cluster of South L.A.

The Fortress voted either for Romney or for Obama by a close margin. The areas corresponding to the cluster of public housing in South Los Angeles voted overwhelmingly—and I mean overwhelmingly—for Obama.

But people vote for other things besides presidents. How about adding another layer of data designed to target particular social views, such as acceptance of gay marriage. Here’s a map of voting patterns for Proposition 8, the California gay marriage ban:

An interesting reversal of what we might have expected based on the presidential vote . . .

What we’re doing here is not trying to make a political point; rather, we’re mashing up non-spatial data on digital mapping interfaces in order to bring together multiple pieces of situated knowlede about and to get diverse perspectives on a particular space. Too often, digital maps are treated like paper maps: they select one or two elements and then deflect everything else, which completely undermines the utility of these interfaces and the plethora of data available online. Mapping socio-economic factors shows us one thing; mapping presidential voting patterns shows us another; mapping proposition voting shows us something else. These just scratch the surface. Individually, digital maps are valuable, but together, they construct a much richer and more robust view of a place than they do individually.

(Here’s an excellent study that brings together data on poverty and many environmental factors—rainfall, distribution of livestock, prevalence of disease carrying flies—in order to build a more robust and predictive model of poverty in Uganda.)

MOOCs and the robot economy

From the Associated Press: The robot economy emergent. How long until the Rise of the Neo-Luddites? I’d be willing to wager $100 that within 5 years, we will see at least one case of someone’s vandalizing a robotics company headquarters (or some other place of high-tech employ).

In academia, the MOOC represents the first significant intrusion of technology into the academy’s stagnant, bloated labor structure. Colleges want to cut costs; students want to pay less for courses. MOOCs score on both accounts, so how can they be resisted? They won’t be. But from whence do the collegiate cost cuts come? In great part, from adjunct labor. MOOCs won’t hurt the tenured members of the academy. The Neo-Luddites in this sector will rise from the adjunct class, the class that will be affected by MOOCs. The California State University is now allowing students to enroll in MOOCs for their basic mathematics courses. If half of Cal State students opt to take College Algebra online, that spells doom–not for professors, who don’t teach general education courses–but for the hundreds of adjuncts who drive from San Bernardino to Dominguez Hills three times a week in order to teach five sections of College Algebra at a few grand per section. What once required thirty adjuncts now requires a single MOOC.

As the AP article makes clear, once these jobs are cut, they won’t return.

And I’m glad they won’t. The adjunct labor system is abhorrent and exploitative. It’s grown well beyond the bounds of its original design; it’s like a prison meant to house 20,000 low-level offenders housing 100,000 Predators hyped up on speed. It needs to go away. Though the MOOC means serious life-altering consequences for many adjuncts–ending in Neo-Luddite reaction–these consequences are merely penultimate. The fade-to-black is the end of (or, at least, the serious curtailing of) the adjunct labor system full stop. No adjunct jobs, then no more lop-sided supply of MAs and PhDs into disciplines that cannot absorb the supply. Students in these disciplines will transfer their creative potential to other, more productive sectors instead of being led on by low wages and the dangling carrot of a tenure track position. Academic departments will contract in accordance with demand for their product–fully employable MAs and PhDs–instead of in accordance with demand for laborers in an exploitative labor system.

Mathematics suffers first because mathematics is pure quantifiable content. It can be delivered en masse and student work can be graded easily with a few algorithms. However, because academic writing is highly structured and “already-mechanized,” it’s very possible for robots to grade student writing, as well. English departments won’t suffer second or even third, but they will suffer eventually. Writing instruction can surely be delivered en masse, so the only way for adjuncts to save their jobs (the only way for writing studies to contract modestly instead of totally) is for the discipline to make its content impossible to quantify.

Technology and the empirical study of writing

A materialist theory of literary form will ultimately have to concern itself with the organic processes of reading and composition, but the way to do this is through empirical study of readers and writers, not more interpretation of texts, or armchair ruminations.  –Cozma Shalizi in a response to Franco Moretti’s Graps, Maps, Trees (128)

Janet Emig initiated the writing process movement when she published The Composing Processes of Twelfth Graders, an attempt to study writing in an empirical way (lower case e; no Lockean baggage implied) by closely observing and polling several high school seniors as they wrote essays. Today, the shortcomings of her study are obvious—the sample size was small, and she had no way to track granular textual changes as they were made in real time. However, despite its limitations, Emig’s work introduced an important assumption to the field of writing studies:

Writing is a natural, organic phenomenon that can be studied empirically.

Unfortunately for Emig and the process movement, writing studies was and is situated in an academic context that requires a sharp pedagogical focus, and the empirical study of writing has little to no educational value. Studying writers can tell us how people write, but it doesn’t necessarily tell us how to teach writing, academic or creative or otherwise.

Intimately tied to the pedagogical critique of the process movement was the political critique. In early studies of writing processes, certain contextual elements (read: race, gender, class) were ignored. Emig, for example, did not deeply address the racial differences of her subjects. Critics claimed that the study of writing processes would not pay enough attention to relevant cultural factors that affect how, where, and why people write. This critique was weak, however, because all empirical pursuits must by design bracket out certain contextual elements in its early stages. As the pursuit progresses and gathers knowledge, the causes and effects (if any) of various contextual factors can be coded and controlled for. Race, class, and gender are such factors—important ones at that—but we needn’t stop there (cross-linguistic differences would be first and foremost on my mind). Dozens of material factors must be taken into consideration when studying writing. Had the process movement not been abandoned, researchers would have gotten around to controlling for all of them.

Then there was the philosophical critique. The goal of studying writing is to build evidence-based theories about this unique human practice from a variety of angles—stylistic, material, cognitive, neuronal, linguistic—and eventually to see how these levels interact. (E.g., What areas of the brain are operational at various stages of writing and re-writing? Are small stylistic changes or large organizational changes more often influential of a text’s shape?  How does textual cohesion emerge? What roles do vision and memory play in the way writers work with their texts on word processors? Do writers across languages and writing systems have completely different processes?)  However, like many humanities disciplines, writing studies has been influenced by postmodernism and is thus adverse to data-driven, quantitative, empirical methods, and not interested in questions—like the ones above—that require these methods. Gary Olson typifies the philosophical critique when he writes that the process movement attempts to “systematize something [writing] that simply is not susceptible to systematization.”

Of course, no evidence is provided for this claim—but then, none is needed. It isn’t a claim at all. It is an a priori assumption, a statement of faith, designed to obviate any empirical work on the subject.

Ironically, the most valid critique of the process movement in writing studies was one that no one ever made: the technological critique. In the 1980s, when the process movement was jostling for academic legitimacy, researchers simply did not have access to the technology that could enable a more robust inquiry into the material, organic processes of writing.

Today, we do have access to that technology. What’s more, the postmodern zeitgeist has waned in its influence, and the political critique was never quite valid. The time seems right for a return, not necessarily to process theory, but to the assumptions it made about data-driven, quantitative, empirical studies of the way humans compose.

Scholars like Chris Anson, Richard Haswell, and Chuck Bazerman are leading the way. Haswell’s call for “RAD” research—replicable, aggregable, data-supported—is essentially a call for empiricism in writing studies. And Anson’s recent report on the use of eye-tracking devices to study writers at work demonstrates how new technologies will enable and benefit this empirical endeavor. Anson’s line of research could lead to major insights into the ways writers access their ‘textual memory’ in order to manage the many semantic strands that comprise any written text. Indeed, this is a perfect example of how technology and RAD methods can test old ideas in writing studies, confirm or complicate those ideas, and fill them in with data-driven details. In 1999, for example, Christina Haas wrote about the way writers manage their texts in Writing Technology: Studies in the Materiality of Literacy:

Clearly, writers interact constantly, closely, and in complex ways with their own texts. Through these interactions, they develop some understanding—some representation—of the text they have created or are creating . . . As the text gets longer and more ideas are introduced and developed, it becomes more difficult to hold an adequate representation in memory of that text, which is out of sight. (117, 121, qtd in Brooke’s Lingua Fracta)

In 2012, enter the eye-tracking software, which can show us where writers look, physically, to develop representations of their texts as they are constructed: What kinds of words or phrases do writers reference most often, as ‘anchors’ for their intellectual wanderings? What are the outside limits of textual vision? Where do writers focus their vision at different stages in the writing process—choosing words, writing sentences, organizing paragraphs, et cetera? Do accomplished writers use their eyes differently than novice writers? Do high-IQ individuals use their eyes less or more while writing; are their visual memories more robust, requiring less visual tracking to make sense of their texts’ cohesion?

Without eye-tracking devices and empirical methodologies, researchers could never hope to answer these and other questions. They would never even think to formulate them.

Outside the field of writing studies, researchers are already using technology to capture and study authorial processes. An IBM study used an application called history flow to study contributions in Wikipedia articles and how numerous contributions endure, change, or are phased out entirely over time. Ben Fry built an amazing visualization of the multiple changes Darwin made to On the Origin of Species across six editions. And on a lesser scale, Timothy Weninger created a time-lapse video that shows the writing of a research paper in various stages (I’m in the process of figuring out how to do something similar, using track changes in Word).

The interest in organic authorial processes extends beyond writing studies, so it boggles my mind that writing studies scholars aren’t at the forefront of this research, which, I grant, is in its early stages. Luckily, things are looking up for RAD, empirical research in writing studies. Now that we can start grounding our theories of composition in real data, it’s only a matter of time before we start gaining empirical insight into this strange, relatively recent human behavior that we call ‘writing’.