A popular Tumblr site for graduate students in the humanities last year was MLAJobs, which satirized postings in the Modern Language Association’s job list. One faux job ad captured one of the many frustrations faced by those in the humanities:
“Digital Humanities: Asst prof in American or British literature, 16th-20th century, with interest in the problematic of digital humanities. We prefer candidates who can tell us, What is digital humanities? Some familiarity with MSWord expected. Send vita and letters to Cornell University, Ithaca NY 14853.”
The problem, of course, is not that no one knows what the “digital humanities” are. Many scholars work in what might be described as digital humanities and have done so since long before the buzz term appeared. And I share their enthusiasm for the many ways in which digital projects can make our objects of study more accessible or open new ways of seeing and understanding them.
The problem is that “digital humanities” now appears to trump plain ol’ “humanities,” particularly among those who hold the purse strings. I fear that what now matters, what legitimizes the humanities in the eyes of many “stakeholders,” is that modifier: digital. Too many of us, beaten down by the relentless insistence on the supposed practicality of STEM degrees—and, thus, in an argumentative leap, the greater value of the STEM fields—are willing to accept this state of affairs. But if we do, we put at risk much more than job lines or funding. We enable the wave of utopianism that the digital pioneer Jaron Lanier has described and criticized in You Are Not a Gadget (Knopf, 2010).
In the minds of digital enthusiasts, the application of new technologies necessarily extends, rather than narrows, our capacities and vision. Perhaps we are all in the process of plugging into a larger meta-consciousness as we connect to one another through the web. Or, at a less heady level, maybe we just need more data to understand our relationship to one another and the world around us. The big-data approach, inspired by corporate data-mining and finding its way into humanities scholarship, presumes the latter. But Lanier questions the idea that “quantity not only turns into quality at some extreme of scale, but also does so according to principles we already understand. … A trope from the early days of computer science comes to mind: garbage in, garbage out.”
The computer doesn’t make meaning. We do. And most of us in the humanities are not sophisticated computer engineers; we require assistance to understand and use the algorithms required to get the next “level of description,” to use the language of complex-systems analysis. When we pass the buck to programmers, the algorithms, and, in turn, to the models they generate, we cede a major part of the meaning-making process. If we wade into the digital humanities, we need to understand, and continue to question, the digital part as well as the humanities part. We can’t allow interpretation to be hidden in a mathematical black box. We need to remember that these digital methods are based, at least initially, on human inputs.
Even our wisest scholars of digital humanities, while continuing to argue for the importance of humanities scholarship, are at risk of obscuring that point. N. Katherine Hayles, in her article “Cognition Everywhere: The Rise of the Cognitive Nonconscious and the Costs of Consciousness,” speaks of the need for the humanities to recognize “nonconscious cognition"—the interpretive function performed by nonconscious entities and systems—in order to avoid the “isolation of the humanities from the sciences and engineering” and to participate in collaborative intellectual work. Only by acknowledging that “interpretation is … pervasive in natural and built environments,” she argues, can “the humanities … make important contributions to such fields as architecture, electrical and mechanical engineering, computer science, industrial design, and many other fields.”
But many of the forms of nonconscious cognition that Hayles identifies, like stock-trading algorithms and networked smart devices, are part of a human-built environment. They do not spring fully formed from the head of Zeus. What she acknowledges as “the sophisticated methods that the humanities have developed for analyzing different kinds of interpretations and their ecological relationships with each other” need to be at the front end of our development of these forms of “nonconscious cognition,” such as shaping the interpretive assumptions that feed into the algorithms of many digital-humanities projects, rather than meekly following behind them.
The risk I see to the humanities is not in using algorithms or, indeed, in any individual digital-humanities projects. The risk is, to quote Lanier, in “digital reification,” in “lock-in,” which “removes ideas that do not fit into the winning digital representation scheme.” If it doesn’t fit, we ignore it, or change the definitions (of texts, of musical notes, of the humanities, of consciousness) to fit the scheme. We dumb down the object of representation in order to make our design look better; and, in turn, the object of representation changes to match the design.
Not only do texts change when digitally filtered, but we change when we observe and interpret them. As Lanier, Hayles, Sherry Turkle, and others have noted, mainstream cybernetics presumes that the human mind works in the same way that a computer does, and that “information is more essential than matter or energy,” as Hayles puts it. Turkle, in Alone Together: Why We Expect More From Technology and Less From Each Other (Basic Books, 2011), has suggested that we are in a “robotic moment,” allowing our technologies to shape our notion of the human rather than the other way around. Her research points out that our sense of the human is changing, and not necessarily for the better. We should perhaps ask ourselves the humanities questions that are being raised by neuroscientists: What happens to our consciousness when we let ourselves be raised by robots and screens, however benign? How do those mirror neurons fire when presented with animated emojis instead of human faces? Are we increasing human diversity, or are we constraining it?
As Lanier quips, “People degrade themselves in order to make machines seem smart all the time.” We become “restricted in practice,” he says, “to what can be represented in a computer.” Our complex, ambiguous selves are at risk from this digital pressure toward conformity. To accept this state of affairs is a kind of madness, a madness of the sort that G.K. Chesterton described when he pointed out that “a small circle is quite as infinite as a large circle; but, though it is quite as infinite, it is not so large. In the same way the insane explanation is quite as complete as the sane one, but it is not so large.”
The humanities, like humans ourselves, are large. But we reduce ourselves more and more often to our tools, describing ourselves as computational systems, soft machines, wired meat. Those are only metaphors, but metaphors that profoundly shape us. In How We Became Posthuman (University of Chicago Press, 1999), Hayles challenged us to uncover the “profound normative and moral implications” of our technologies and examine how those technologies affect how we think and live.
The answer isn’t simply for humanities scholars to learn how to program or to work with programmers. The answer is also for programmers to heed the humanities in thinking through the implications of their decisions, to make those decisions visible, and for all of us to recognize that digital technologies are only some of the tools at our disposal. We needn’t be limited to them—or by them. The humanities aren’t “extra,” nor can they be subsumed into a more scientific or technological worldview. The"soft” perspective of the humanities, just like the human itself, cannot be adequately represented or processed by the digital. It is no failure to say so, any more than it is the saxophone’s failure that its music cannot be satisfactorily represented by a limited and limiting computerized version of an instrumental “note,” to use one of Lanier’s examples.
As we look to the future, then, humanities scholars need to think about how best to make use of our technologies without trying to emulate them. Digital technologies have excellent applications but aren’t a good fit for all projects, nor are the projects that do fit necessarily better than the ones that don’t. We must resist the temptation to jam our square pegs into round holes and to treat the products of digital-humanities scholarship as more valuable simply because they are digital.
We must also recognize, as Hayles has pointed out, that humanities scholars have leaned heavily on human exceptionalism as the basis for our approach, and that digital scholarship might mitigate that sense of species self-centeredness. But if that sense of exceptionalism falls away as we continue to explore our place in complex world systems, that should be neither the end of the human or the end of the humanities. We needn’t fall into misanthropic despair, like some gaggle of Gullivers. Rather, a new humility with respect to our human limitations can help us to develop not only a wiser humanities but also wiser science and wiser technology, built on the realization that what we are capable of knowing, even with the help of our tools, is not all there is. That wiser scholarship might help us to understand our failures as well as our successes, to recognize our limits and ill-founded assumptions.
Read that as failure if you must. I prefer to think of it as an opportunity for wonder. And wonder, as Plato’s Socrates suggests, is the beginning of wisdom.
