UC Santa Barbara UC Santa Barbara Previously Published Works Title The Meaning of the Digital Humanities Permalink https://escholarship.org/uc/item/5gc857tw Journal PMLA/Publications of the Modern Language Association of America, 128(2) ISSN 0030-8129 Author Liu, Alan Publication Date 2013-03-01 DOI 10.1632/pmla.2013.128.2.409 Peer reviewed eScholarship.org Powered by the California Digital Library University of California https://escholarship.org/uc/item/5gc857tw https://escholarship.org http://www.cdlib.org/ WHAT IS THE MEANING OF THE DIGITAL HUMANITIES TO THE HU- MANITIES? his question of disciplinary meaning—which I ask from the viewpoint of the humanities generally—is larger than the question of disciplinary identity now preoccupying “DH” itself, as insiders call it. Having reached a critical mass of participants, publications, conferences, grant competitions, institutionalization (centers, pro- grams, and advertised jobs), and general visibility, the ield is vigor- ously forming an identity.1 Recent debates about whether the digital humanities are a “big tent” (Jockers and Worthey), “who’s in and who’s out?” (Ramsay), whether “you have to know how to code [or be a builder]” (Ramsay, “On Building”), the need for “more hack, less yack” (Cecire, “When Digital Humanities”; Koh), and “who you calling untheoretical?” (Bauer) witness a dialectics of inclusion and exclusion not unlike that of past emergent ields.2 An ethnographer of the ield, indeed, might take a page from Claude Lévi- Strauss and chart the current digital humanities as something like a grid of aili- ations and diferences between neighboring tribes. Exaggerating the diferences somewhat, as when a tribe boasts its uniqueness, we can thus say that the digital humanities—much of which ailiates with older humanities disciplines such as literature, history, classics, and the languages; with the remediation of older media such as books and libraries; and ultimately with the value of the old itself (history, ar- chives, the curatorial mission)—are not the tribe of “new media stud- ies,” under the sway of the design, visual, and media arts; Continental theory; cultural criticism; and the avant- garde new.3 Similarly, despite signiicant trends toward networked and multimodal work spanning social, visual, aural, and haptic media, much of the digital humanities focuses on documents and texts in a way that distinguishes the ield’s work from digital research in media studies, communication studies, information studies, and sociology. And the digital humanities are exploring new repertoires of interpretive or expressive “algorithmic criticism” (the “second wave” of the digital humanities proclaimed ALAN LIU , professor in the English depart- ment at the University of California, Santa Barbara, is the author of Wordsworth: The Sense of History (Stanford UP, 1989), The Laws of Cool: Knowledge Work and the Culture of Information (U of Chicago P, 2004), and Local Transcendence: Essays on Postmodern Historicism and the Database (U of Chicago P, 2008). He started Voice of the Shuttle, a Web site for humanities re- search, in 1994. Recent projects he has di- rected include the University of California Transliteracies Project, on online reading, and RoSE (Research-Oriented Social Environ- ment), a software project. Liu is a coleader of the 4Humanities advocacy initiative. the changing profession The Meaning of the Digital Humanities alan liu 1 2 8 . 2 ] © 2013 alan liu PMLA 128.2 (2013), published by the Modern Language Association of America 409 in “The Digital Humanities Manifesto 2.0” [3] ) in a way that makes the ield not even its earlier self, “humanities computing,” alleged to have had narrower technical and service- oriented aims.4 Recently, the digital humani- ties’ limited engagement with identity and social- justice issues has also been seen to be a differentiating trait—for example, by the vibrant #transformDH collective, which wor- ries that the digital humanities (unlike some areas of new media studies) are dominantly not concerned with race, gender, alternative sexualities, or disability.5 Of course, there are overlaps of people and methods between tribes—to the point that, taking another page from Lévi- Strauss, we would want to mention trickster figures embodying both sides of digital humanities’ diferences. One such trickster is the ield of book history. Proliic in its revisionary studies of matrial texts, ephemera, marginalia, social reading, publishing history, and so on, book history now partly parallels post- McLuhan media studies. Peter Stallybrass argues, for in- stance, that “the codex and the printed book were the indexical computers that Christianity adopted” as its “technology of discontinuity” (74, 73). Book history therefore overlaps on its new- media- studies side with media archaeol- ogy (especially in the German lineage well rep- resented in this context by Cornelia Vismann’s Files: Law and Media Technology) and on its digital humanities side with approaches to digital texts rooted in revisionary textual edit- ing, bibliography as the sociology of texts, and the materiality of the digital.6 A similar trick- ster is science and technology studies (STS), which are curiously underrepresented in both new media studies and the digital humanities even as they are clearly relevant in a way repre- sented by such scholars as N. Katherine Hayles and Tim Lenoir. So, too, historical sociology— which now applies social- network analysis to document corpora—is a trickster splitting the diference between the social sciences and the digital humanities, the latter of which also in- creasingly use social- network analysis to study plays, novels, literary reception, history, and so on.7 And of the communities of people who overlap, trickster- like, between the digital hu- manities and new media studies, the HASTAC collaboratory—which has had enormous suc- cess tapping into the energies especially of graduate students (“HASTAC Scholars”)—is perhaps the most coyote. Yet even if we were to complete our hy- pothetical ethnographer’s chart, it would not adequately explain the digital humanities. We would be leaving unexplained the relation of the digital humanities to the humanities gen- erally. My thesis is that an understanding of the digital humanities can only rise to the level of an explanation if we see that the un- derlying issue is the disciplinary identity not of the digital humanities but of the humani- ties themselves. For the humanities, the digi- tal humanities exceed (though they include) the functional role of instrument or service, the pioneer role of innovator, the ensemble role of an “additional field,” and even such faux- political roles assigned to new ields as challenger, reformer, and (less positively) ith column. his is because the digital humani- ties also have a symbolic role. In both their promise and their threat, the digital humani- ties serve as a shadow play for a future form of the humanities that wishes to include what contemporary society values about the digi- tal without losing its soul to other domains of knowledge work that have gone digital to stake their claim to that society. Or, precisely because the digital humanities are both func- tional and symbolic, a better metaphor would be something like the register in a computer’s central processor unit, where values stored in memory are loaded for rapid shuff ling, manipulation, and testing—in this case, to try out new humanistic disciplinary identi- ties evolved for today’s broader contention of knowledges and knowledge workers. he question of the meaning of the digi- tal humanities best opens such an argument to 410 The Meaning of the Digital Humanities [ P M L A t h e c h a n g i n g p r o f e s s i o n view because it registers both a speciic problem in the digital humanities and the larger crisis of the meaningfulness of today’s humanities. Meaning is clearly a metavalue and also metaproblem for the digital humanities. To unpack this meaning problem, I will spotlight a recent work of digital literary scholarship by two beginning scholars that is state- of- the- art and representative of major trends in the digital humanities—a tactic that has the addi- tional advantage of providing outsiders to the ield with an end- to- end look at an example of research by digital humanists. he work is Ryan Heuser and Long Le- Khac’s A Quanti- tative Literary History of 2,958 Nineteenth- Century British Novels: he Semantic Cohort Method (2012), the fourth of the inf luential digital pamphlets issued by the Stanford Liter- ary Lab.8 Heuser and Le- Khac report on their innovations in the methods of “distant read- ing” and text mining that are the signatures of the lab (where they worked with Matthew L. Jockers, Franco Moretti, and others), and they do so with a methodological self- awareness that puts the meaning problem front and cen- ter. hey relect near their opening: [W] hat is the meaning of changes in word us- age frequencies? What do we do with such data? With much current research drawing on word frequencies and other quantiiable aspects of culture, these are big questions. We can see now that the greatest challenge of de- veloping digital humanities methods may not be how to cull data from humanistic objects, but how to analyze that data in meaning fully interpretable ways. (4; my emphases) And they add in their concluding section: he general methodological problem of the dig- ital humanities can be bluntly stated: How do we get from numbers to meaning? he objects being tracked, the evidence collected, the ways they’re analyzed—all of these are quantitative. How to move from this kind of evidence and object to qualitative arguments and insights about humanistic subjects—culture, literature, art, etc.—is not clear. (46; my emphasis) Throughout, Heuser and Le- Khac give meaning a gravity that indicates that in the digital humanities the meaning problem has roughly the same weight as the “saving the phenomena” problem in the philosophy of science—the problem, that is, of relating em- pirically observed phenomena to explainable, theorizable, or predictable (“saved”) phenom- ena. In essence, Heuser and Le- Khac are out to “save the data” by making it meaningful.9 he form of data they wish to save is quantitative. Of course, no single example can typify all aspects of the booming ield of digital hu- manities. So before we examine the speciics of the meaning problem in this instance, it will be useful to make four observations that relate Heuser and Le- Khac’s research to other work in the digital humanities. he meaning problem may not be as central everywhere, but its frame of analysis is convertible to other frames so that we can see that many parts of the ield link up to the same cluster of issues. he following is a kind of conversion table for relating Heuser and Le-Khac’s work to other digital humanities areas. First, we note that Heuser and Le- Khac select their research material from already digitized texts (2,958 British novels from 1785 to 1895, all from The Internet Archive except for 250 from Chadw yck- Healey).10 his means that their work belongs not in the orbit of digitizing, text encoding, publish- ing, or archiving (activities characteristic of many projects in digital editing, collecting, and curating) but in that of processing and analyzing already built digital repositories (in the manner of projects in text analysis, social- network analysis, visualization, spatial history, etc.). Still, the digital humanities are young enough that these two broad modes of work are not fully specialized. On the one hand, leading text- encoding and digita l- 1 2 8 . 2 ] Alan Liu 411 t h e c h a n g i n g p r o f e s s i o n archiving projects ind it necessary to create their own analytic, processing, and visualiza- tion tools to present materials. And, on the other hand, text- analysis, visualization, and other processing projects oten have to go to great lengths to select, clean, and prepare pre- existent digital materials as a usable corpus. Second, while Heuser and Le- Khac for- mulate their meaning problem in terms of quantification, that is not the only possible formulation. Quantification is indeed a key digital humanities issue, especially when linked to sheer quantity in the wake of recent “big data” funding competitions in the ield. But digital humanists also have nonquanti- tative ways of putting the meaning problem. One is the idea of models (and modeling pro- cesses) as richly developed, for example, by Willard McCarty. Models reveal meaning (recognized in patterns, trends, forms) only by reducing the dimensions and features of meaning. Diagrammatic models, especially the visualizations proliferating in the digital humanities (including Heuser and Le- Khac’s essay), are a case in point, since large dis- course networks (visualized through such tools as Gephi) are comprehensible when their scope or detail is kept low but otherwise grow into beautifully mystifying galaxies of nodes and links. Another case in point is the kind of textual model (instantiated in document- type deinitions [DTDs] and XML schemas) that enables text encoding. A well- known de- bate in the digital humanities thus concerns whether the principle of an ordered hierarchy of content objects (OHCO) underlying such models makes texts machine- readable only by disallowing the full range of what Jerome Mc- Gann calls human- readable “overlapping” and “recursive structures” (“Position Statement”). hird, even if we concentrate on quantii- cation as the key meaning problem, it makes a diference which disciplinary branch of the digital humanities we are dealing with, since digital history and digital literary studies, to take two major branches, arrive at quantiica- tion through diferent routes and at diferent stages. Digital history emerged in a discipline that had already forcefully experienced nu- merical method—for example, the quantita- tive side of the Annales school and cliometrics. Perhaps as a result, digital history is farther along than most digital humanities branches in molding quantitative work to a related heritage of Annales method: spatial and car- tographic history. he state of the art in digital history, as it were, is Fernand Braudel plus sat- ellite mapping (Seefeldt and homas; Owens). Fourth, Heuser and Le- Khac’s research required a combination of skills in program- ming and interpretation, thus addressing the “do you have to be a builder?” question that has recently bedeviled the digital humanities by answering, in efect, “you have to be both a builder and an interpreter.” Less interesting than this question itself, which is based on an increasingly obsolete notion of solo work in which one is either a builder or interpreter, is the way Heuser and Le- Khac are both builders and interpreters: through rich collaboration. Here we reach the outer limits of the frame of the meaning problem, where it converts into a coextensive frame. Just as meaning is both a metavalue and a metaproblem, so is collabo- ration as it bears on such urgent issues in the digital humanities as coauthorship, collec- tive project building, multigraph books, open peer review, social media, crowdsourcing, and the hiring and promotion implications of all these. Rather than explore the collabo- ration problem in its own frame here, I note only that it is fundamentally convertible to the meaning problem. For example, the ques- tion of what kind of knowledge is produced by “the wisdom of the crowd,” “collective in- telligence,” “the long tail,” “the hive mind,” “folksonomy,” and so on (dominant memes of Web 2.0) is essentially a question about the meaning of the social version of big data, the big crowd. he mind, or mindlessness, of that crowd has been a core problem of modernity since at least the French Revolution. 412 The Meaning of the Digital Humanities [ P M L A t h e c h a n g i n g p r o f e s s i o n With these observations as a guide for re- lating Heuser and Le- Khac’s work to other em- phases in the digital humanities, we can now look closely at their pamphlet. he meaning problem comes to the fore there as a conse- quence of Heuser and Le- Khac’s key research innovation. hey report on their use of a tool they created called Correlator, which, when fed “seed words” suggested in part by “exist- ing literary scholarship” (11), processed their long- nineteenth- centur y corpus of novels (with the aid of a database designed by Jock- ers that tabulated the number of occurrences of each word [Heuser and Le- Khac 6]) to ind other words that were statistically correlated with the seed words and whose frequency trends, measured longitudinally across the century in decade intervals, closely followed the frequency trends of the seed words. In other words, Correlator finds what Heuser and Le- Khac call “word cohorts” in the cor- pus, consisting of words that kept company with one another and behaved similarly over time, waxing or waning in frequency together and standing out from other companies of words. he inal result—which I jump to while skipping the details of Heuser and Le- Khac’s algorithmic method as well as, temporarily, an important adjustment step in their method— is the identification in the word cohorts of “rich, consistent semantic f ields” that are “both semantically and culturally legible” in historical trends. An example is the 136 words such as gentle, sensible, vanity, elegant, deli- cacy, reserve, mild, and restraint that they label the “social restraint ield” in the novels (8). Using this method, Heuser and Le- Khac made two principal discoveries. First, they used the seed words integrity, modesty, sen- sibility, and reason (suited to scholarship on novels of the period) to ind a strikingly large cohort of “abstract, socially norma- tive, evaluative, and highly polarized words” whose frequencies declined dramatically over the century. his cohort they subcategorized into semantic ields labeled “social restraint,” “moral valuation,” “sentiment,” and “partial- ity” (11–19). hen, serendipitously (instead of under the sway of preexisting scholarship), they fed the unlikely seed word hard into Cor- relator to discover a strikingly distinct, very large cohort of “concrete description words of a direct, everyday kind” whose frequencies rose dramatically over time. his cohort they subcategorized into semantic ields for action verbs, body parts, colors, numbers, locational and directional adjectives and prepositions, and physical adjectives (19–27). he strongly inverse correlation between these two large and diferentiated word cohorts, which they further statistically corroborated (28–29), started them on the path of interpretation. he “abstract” cohort, they concluded, con- sists of words whose usage, while unanchored in speciics, was “monitored and tightly con- strained ” by the traditional smaller, rural communities represented more or less ear- lier in the century in evangelical, gothic, and village novels and the novels of Jane Austen, Walter Scott, and George Eliot (Eliot is a chronological outlier), while the “hard” co- hort is populated by words whose stand- alone referentiality and alienation from larger con- texts correlated later in the century with the “wider, less constrained social spaces” of the urban centers represented in city, industrial, adventure, fantasy, science iction, and chil- dren’s novels (30–34). This insight finally led Heuser and Le- Khac at the highest level of interpretation to follow Raymond Williams’s Culture and Society and he Country and the City in sug- gesting that the inverse “abstract” and “hard” trends reveal something signiicant about the “social space of the novel.” hey argue that the “values of conduct and social norms” in “knowable communities” (a phrase from Wil- liams) declined in the face of “urbanization, industrialization, and new stages of capital- ism” (33, 35–36). he computational veriica- tion of this previously known thesis, coupled with the discovery of precise word cohorts 1 2 8 . 2 ] Alan Liu 413 t h e c h a n g i n g p r o f e s s i o n giving genuinely fresh insight into the thesis, enables them in later sections of their pam- phlet to ofer more recognizably normative literary and cultural criticism, touching on action, setting, and character (37–45). Here they closely read texts (including, for the irst time, block quotations), match aggregate trends to “units understandable and famil- iar to us as readers and literary scholars[:] the actual novels, genres, and authors” (31), and generalize about sweeping changes in cultural history, but with the important dis- tinction that their reading is based not on the usual anecdotal, faux- empirical, or unique- case observations of literary criticism (e.g., noticing that a word appears “oten” or in an “important” location) but on lines of inter- pretation generated by machine observation. Here we reach the crux of the meaning problem in the digital humanities. While Heuser and Le- Khac mix “supervised and unsupervised procedures” (28), the newest, boldest, and most interesting part of their met hodolog y is unsuper v ised. They help advance an important, general digital hu- manities goal that might be called tabula rasa interpretation—the initiation of interpreta- tion through the hypothesis- free discovery of phenomena. Also shared by such digital humanities methods as topic modeling (a mathematics- based way to discern differ- entiated clusters of words that Heuser and Le- Khac use to corroborate their findings), the ideal in its purest form is what Heuser and Le- Khac call “an unsupervised method that generates topics without subjective in- put from users” (28).11 That is, a computer should be able to read texts algorithmically and discover word cohorts or clusters lead- ing to themes without acting on an initial concept from an interpreter looking to con- firm a particular theme. Of course, Heuser and Le- Khac assume that there are preexist- ing themes to be found in the word cohorts of primary materials and also that the main mission is to discover them. But tabula rasa interpretation is equally a goal of the more postmodern side of the digital humanities, which argues that critics should use algo- rithmic methods to play with texts experi- mentally, generatively, or “deformatively” to discover alternative ways of meaning that are not so much true to preexisting signals as rifs on those signals. he common goal is to ban- ish, or at least crucially delay, human ideation at the formative onset of interpretation. However, tabula rasa interpretation puts in question Heuser and Le- Khac’s ultimate goal, which is to get from numbers to human- istic meaning (“qualitative arguments and insights about humanistic subjects—culture, literature, art, etc.”). It is not clear epistemo- logically, cognitively, or socially how human beings can take a signal discovered by ma- chine and develop an interpretation leading to a humanly understandable concept unless that signal (in order to be recognized as a sig- nal at all) contains a coeval conceptual ori- gin that is knowable in principle because, at a minimum, the human interpreter has known its form or position (the slot or approximate locus in the semantic system where its mean- ing, or at least its membership in the system, is expected to come clear).12 If the machine can discover word cohorts triggered by seed words, in other words, then what seed con- cepts—which is to say seed semantics (using “semantics,” for the moment, as overlapping with “concepts” in a manner consistent with Heuser and Le- Khac’s usage)—lurk in the background as a latent, classificatory form of relational semantic positions able to make word cohorts into “proto- semantic ields” (7)? hus the immense importance of the ad- justment step in Heuser and Le- Khac’s method that I earlier elided. In fact, Heuser and Le- Khac used Correlator by itself to produce only initial word cohorts and not inished semantic ields because they realized that they needed to ensure that their cohorts had a semantic consistency that quantitative correlation alone could not ofer. Some word cohorts discovered 414 The Meaning of the Digital Humanities [ P M L A t h e c h a n g i n g p r o f e s s i o n by machine, for example, suggested only fuzzy semantic ields that seemed blurred by extra- neous words or, inversely, to lack words that should have been there. In other words, word cohorts had to be iltered and illed out in ways that made sense. Heuser and Le- Khac thus re- alized that they needed not just seed words but seed semantic concepts suspended precisely in what I above called a classiicatory form of re- lational semantic positions, endowing cohorts with a sense of what is and is not proximate in meaning. In short, Heuser and Le- Khac needed a thesaurus, and not just any thesaurus but one ofering a historical semantics matched to the longitudinal dimension of their word cohorts. Ater they had already begun using Correlator, they pulled a rabbit out of the hat. hey turned to the remarkable Historical hesaurus of the Oxford En glish Dictionary (2009; hereafter HTOED), which had just been published, and borrowed its historical semantic classiications through what they call “a dialogic method that drew on both quantitative historical data and qualitative semantic rubrics to construct semantic ields with precision and nuance”: Having moved through an empirically and his- torically focused stage of semantic ield devel- opment, we needed to return to the semantic focus in order to make such purely empirical word cohorts interpretable and meaningful. Our initial approach was to ilter through these words for groups that seemed semantically co- herent, but this proved too loose and subjective. . . . Finally we turned to the [HT]OED. . . . It’s nearly exhaustive, its categories are nuanced and specific, and it’s truly organized around meaning. We used this powerful taxonomy to do two things: irst, be more speciic in identi- fying the semantic categories that constituted our word cohorts; second, to expand these word cohorts with many more words. (7)13 Created at the University of Glasgow be- ginning in 1965, the HTOED taxonomizes the En glish language into three master semantic concepts, “the external world,” “the mind,” and “society,” and then, descending its clas- siicatory tree by stages, into myriad ramiica- tions. On the lower branches of the taxonomy, synonyms appear in the chronological order in which they entered the language.14 At irst glance, therefore, the HTOED is the perfect concept- hunting guide for Heuser and Le- Khac’s word- cohort- hunting machine. his seems even more apparent when we realize just how human- powered the semantic in- terpretation involved in making the HTOED was. As documented in reports by Christian Kay and Irené Wotherspoon, among its chief editors, and in a detailed e-mail to me from the current HTOED associate director, Marc Alexander, the editors’ sorting of words by meaning and chronology recapitulated the famous use of paper slips in the compilation of the OED itself.15 Started before humani- ties computing was practical, the HTOED required human beings over decades to write down individual words from the OED on paper slips with meanings, usage dates, and sparse metadata, then to sort, bundle, and ile the slips in conceptual groupings and hierar- chies. When computation entered the picture, it did so originally in a secondary capacity (to drive the print run of the work). In its forma- tive state, the HTOED was a human labor of semantic ordering. B y i n s t a l l i n g t h e H T OE D a s w h a t amounts to a plug- in for Correlator, Heuser and Le- Khac sowed their hermeneutical pro- cess with a coseed of human semantic inter- pretation. They thus “solved” the meaning problem only by deftly turning the aporia between tabula rasa quantitative interpreta- tion and humanly meaningful qualitative in- terpretation into its own apparent solution: a “dialogic approach that oscillates between the historical and the semantic, between empiri- cal word frequencies that reveal the histori- cal trends of words and semantic taxonomies that help us identify the meaning and con- tent of those trends” (9). hey add, “Strictly 1 2 8 . 2 ] Alan Liu 415 t h e c h a n g i n g p r o f e s s i o n speaking, the methods developed here are not looking at word cohorts, which have histori- cal consistency but may lack semantic coher- ence, or semantic ields, which have semantic coherence but may have an ahistorical rela- tionship. he real object of study is a hybrid one that satisies both requirements” (9–10). Such hybridity is a prevalent feature of digital humanities method.16 How might such hybrid method be better grounded theoretically? I return to my earlier diferentiation of the digital humanities from neighboring ields. Two of the deiciencies in the digital humanities revealed by that ethno- graphic map of ields are relevant. One i s de sig n t he or y a nd pr ac t ic e , which Anne Burdick, Johanna Drucker, Pe- ter Lunenfeld, Todd Presner, and Jef frey Schnapp in their recent Digital_ Humanities declare with manifesto- like boldness to be central to the ield (esp. 12–16, 117–19). Cur- rently, it seems to me, the union between the digital humanities and new- media- studies communities that would be needed for full realization of this vision is more a goal than a reality, existing in speciic projects and not programmatically. But the book’s coauthors, expressing a West Coast view of the digital humanities rooted in media and design arts, are right to aim for design as a principle of knowledge discovery and generation rather than (more typical in the digital humanities) as an ater- the- fact rendering of data in scat- ter plots, social- network graphs, and other stale visualizations or, equally tired, book- like or blog- like publication interfaces. As the coauthors put it, when “used to pose and frame questions about knowledge,” design is “an intellectual method,” an “embodiment of a project’s argument and methodology,” “an act of thinking,” and a “new foundation for the conceptualization and production of knowledge” (13, 14, 15, 117). Interactive, mul- timodal, dynamic, and participatory design in the digital age is a method not just of pat- tern recognition but of pattern understand- ing. Seeing design in data is a method for knowing meaning in the digital humanities. he other deiciency I refer to is science and technology studies (STS), which digital humanists oten occlude even as they speak of “digital technology,” “media technology,” and so on, as if technology were an indivis- ible part of the digital and media without its own history, philosophy, sociology, poli- tics, economy, and aesthetics all tangled up with, yet also distinct from, science. I invoke especially the postmodern branch of STS (e.g., Feyerabend; Latour; Pickering), whose “against method” view of science (especially in its weird relations with technology) is that any quest for stable method in understanding how knowledge is generated by human beings using machines founders on the initial fallacy that there are immaculately separate human and machinic orders, each with an ontologi- cal, epistemological, and pragmatic purity that allows it to be brought into a knowable methodological relation with the other— whether a relation of master and slave, cause and efect, agent and instrument, or another. What could we learn from STS if we took the Stanford Literary Lab and other digital hu- manities centers and programs at their word and studied them as labs, much as Andrew Pickering studied the “hunting of the quark” in a physics lab (68–112)? he answer is likely that digital humanities method—converg- ing with, but also sometimes diverging from, scientif ic method—consists in repeatedly coadjusting human concepts and machine technologies until (as in Pickering’s t he- sis about “the mangle of practice”) the two stabilize each other in temporary postures of truth that neither by itself could sustain. Knowledge is an ice- skater’s dance on a slip- pery epistemic surface, on which neither the human nor the machine—the dancer nor the skates—alone can stand. STS, in other words, is another method for knowing mean- ing in the digital humanities. In fact, it can be thought of as complementing the method 416 The Meaning of the Digital Humanities [ P M L A t h e c h a n g i n g p r o f e s s i o n of design. he diference between a mangle of practice and a dance, ater all, is design. Indeed, an STS approach opens a fasci- nating chapter in our reading of Heuser and Le- Khac’s work that I can only briely relate.17 While I have not conducted STS- style eth- nographic and documentary research on the Stanford Literary Lab (except through infor- mal observations during a daylong visit at the invitation of the lab),18 I have looked from a distance into the “lab” of the making of the HTOED—that is, the human- semantics lab that Heuser and Le- Khac position as the part- ner to their computational- analysis lab. When we view the HTOED in this way, we realize that Heuser and Le- Khac pulled not just any rabbit out of a hat but a special rabbit much like the one that led Alice down the rabbit hole. he HTOED is less a solution to the meaning prob- lem than a recursive, looking- glass version of the very same problem. At least ive aspects of its making are relevant in this regard: 1. While the creation of the HTOED was essentially precomputational, it was not pretechnological. Requiring human be- ings to write words, meanings, and meta- data on paper slips and then to sort the slips in drawers (and later in databases), the HTOED originated as a thoroughly en- tangled human- technological, rather than simply human, semantic act. 2. Following Heuser and Le- Khac, I have so far treated “semantics” and “concepts” as coincident. But in relecting on the HTOED’s method, Kay separates out lexical semantics (meaning relations among words) and conceptual semantics (meanings linked to external- world referents). A focus on the former, she says, was “essentially the procedure adopted in HTOED, and is what we mean by saying that the classiication should ‘emerge’ from the data,” meaning the purely lexical data in their source text, the OED (“Classiication” 265–66). 3. he phrase “‘emerge’ from the data” is the cue for an important tenet of the HTOED. As Kay argues, “Our theoretical position on HTOED has always been that the classiica- tion at whatever level should develop from the data rather than be imposed upon it using some predetermined schema” (“Clas- siication” 258). he underlying authority of the HTOED, it turns out, is “data” (the record of the En glish language observed through the instrument of the OED). hat is, the HTOED itself adhered to the princi- ple of the tabula rasa discovery of phenom- ena and initiation of interpretation (in this case, taxonomic interpretation). 4. It was thus ater surveying their lexical data that the HTOED editors contravened the ordering scheme of their canonical predecessor, Roget’s hesaurus, which had put “abstract relations” irst. he HTOED puts the “external world” irst in its tri- nary taxonomic structure. hat’s because concrete and near- to- hand word senses relating to the external world entered the language earlier (Kay, “Classiication” 257–62). he implications for our read- ing of Heuser and Le- Khac are startling. he long historical trend identiied by the HTOED, in which concrete words precede abstract ones, is the reverse of the long- nineteenth- century novelistic trend that Heuser and Le- Khac identify, in which “abstract” words dominate earlier and “hard” words later. Of course, this by no means contradicts their thesis, since in more fully stated form their argument may well be that, because of urbanization, the nineteenth century (and the novel form) was exceptional, or at least matched the trend of only a few other centuries (and forms) in history. But a commitment to reading nineteenth- century novels in con- junction with the HTOED’s larger corpus would require further testing to see how exceptional the nineteenth- century trend really was and thus whether additional sociocultural or other phenomena must be factored in to explain its speciicity.19 5. Finally, as a grace note, I add that while computation was an aterthought in mak- ing the HTOED, today it is crucial for advanced uses of the work. he HTOED 1 2 8 . 2 ] Alan Liu 417 t h e c h a n g i n g p r o f e s s i o n editors eventually migrated their content into a succession of relational databases be- cause they realized that computation might support advanced, real- time querying (Wotherspoon, “Historical hesaurus” 218). Even more dramatically, a recent essay by Alexander shows that querying the HTOED for insight into the history of the language can itself be a form of the digital humanities (“‘Various Forms’”). Alexander conducts quantitative computational analysis of the HTOED to generate visualizations leading to hypotheses about language changes (see the igure below). Connecting the STS and design approaches I outlined earlier, we might even say that at this point using the HTOED becomes an experiment in digital design. In the end, the HTOED is not the “other” that Heuser and Le- Khac need to help make the work of their Correlator meaningful; it is the precursor of Correlator. But going down a rabbit hole, while nec- essary in pursuing any case of the digital hu- manities to its methodological foundation, is not how we must conclude. I thus climb out of the speciic purview of my example, and even of the digital humanities ield, to open my argument to its most general extent. Ap- ropos is the following insight from N. Kath- erine Hayles: he further one goes along the spectrum that ends with “machine reading,” the more one implicitly accepts the belief that large- scale multicausal events are caused by conluences that include a multitude of forces interacting simultaneously, many of which are nonhu- man. . . . If events occur at a magnitude far ex- ceeding individual actors and far surpassing the ability of humans to absorb the relevant in- formation, however, “machine reading” might be a irst pass toward making visible patterns that human reading could then interpret. (29) It is not accidental, I can now reveal, that at the beginning of this essay I alluded to Lévi- Strauss and structural anthropology. Structur- alism is a midpoint on the long modern path toward understanding the world as system FIG. 1 Marc Alexander’s tree- map visualiza- tion of present- day En glish in the HTOED. Each dot represents a word, and the shade of the dot corresponds to when the word entered the lan- guage (darker dots show earlier words). Words are arranged by semantic prox- imity as indicated in the labels. 418 The Meaning of the Digital Humanities [ P M L A (e.g., as modes of production; Weberian bu- reaucracy; Saussurean language; mass, media, and corporate society; neoliberalism; and so on) that has forced the progressive side of the humanities to split of from earlier humanities of the human spirit (Geist) and human self to adopt a worldview in which, as Hayles says, “large- scale multicausal events are caused by conluences that include a multitude of forces . . . many of which are nonhuman.” his is the backdrop against which we can see how the meaning problem in the digital humanities registers today’s general crisis of the mean- ingfulness of the humanities. The general crisis is that humanistic meaning, with its residual yearnings for spirit, humanity, and self—or, as we now say, identity and subjectiv- ity—must compete in the world system with social, economic, science- engineering, work- place, and popular- culture knowledges that do not necessarily value meaning or, even more threatening, value meaning but frame it systemically in ways that alienate or co- opt humanistic meaning. Humanistic knowledge today is thus increasingly assimilated to what humanists themselves call research, evidence, analysis, method, productivity, and “impact” (as this term is institutionalized in “research assessment exercises” in British universities), with no unfilled time and space left for any old ghosts in the machine—unless, as I have argued in Laws of Cool, there remains a yearn- ing, nowhere keener than among our students, caught in the educational mangle, to be cool. Cool people say nix to today’s knowledge- work system even as they walk into the cubicles. Of course, if this were only a problem of research methodology, then I would be extrav- agant to call it a “crisis” in the meaningfulness of the humanities. But “crisis” is appropriate when we realize that the meaning problem also afects pedagogy and jobs in the wake of economic recession, which brings the problem cruelly to bear on individual humanists in training or seeking jobs (not to mention on the humanities programs that nurture and employ them). Here opens a set of topics that I cannot deal with in present limits but that more fully demonstrate how the digital humanities regis- ter the larger issues of the humanities. One topic is the way digital pedagogy— as witnessed in current controversies over massive open online courses (MOOCs), all- digital “campuses” in public universities, and so on—registers the possibility of gigantic changes in the aims, practices, audiences, in- frastructure, and staf of humanities teaching. he bluf is now called on decades of defen- sively legitimating the humanities as thinking and language “skills” added on top of tradi- tionally meaningful humanistic knowledge. If the humanities it that mold in part, then maybe—some administrators and legislators think—they should it it entirely so that their content can be “delivered” modularly through the Internet in the manner of the MOOCs or Khan Academy courses in science, technol- ogy, engineering, and mathematics that have made the biggest public impression. Another topic is the job market for digital humanists, especially in regard to tenurable faculty lines versus “alt- ac” (alternative aca- demic career) adjunct, staff, support, and cultural- institution positions. he turmoil and uncertainty in the nature of digital humani- ties jobs register the larger uncertainty of em- ployment in the humanities as “meaningful” jobs transition away from tenure and toward a corporatized ideal of reconigurable and re- placeable professional- managerial knowledge workers perpetually threatened with restruc- turing layofs in favor of even more exploited “permatemp” and outsourced labor. In all these ways, the digital humanities register the crisis of the humanities. For that reason, I and others started the 4Humanities advocacy initiative, “powered by the digital humanities community,” so that the digital humanities can try to advocate for the hu- manities and not just register their crisis. I do not know how much diference that ini- tiative and others like it will make in the 1 2 8 . 2 ] Alan Liu 419 t h e c h a n g i n g p r o f e s s i o n meaningfulness of the humanities to the world. But I do know that such an effort— dedicating the digital humanities to the soul of the humanities—is what is meaningful for a humanist, digital or otherwise, now.20 NOTES 1. Among recent surveys and overviews of the digi­ ta l hu ma nit ies a re Svensson’s a r t icles; Hayles; K ir­ schen baum, “Digital Humanities” and “What Is Digital Humanities”; Liu, “State” and “Where Is Cultural Criti­ c i sm”; Bu rd ic k , Dr uc ker, Lu nen feld , P re sner, a nd Schnapp; and “Digital Humanities Manifesto 2.0.” Re­ cent or forthcoming essay collections about the digital humanities include Schreibman, Siemens, and Unsworth; Gold; Price and Siemens; and Goldberg and Svensson. On whether the digital humanities are a ield and on the linguistic usage of the phrase digital humanities, see Liu, “Is Digital Humanities a Field?” 2. These expressions in recent conferences, blogs, tweets, discussion threads, papers, etc., have become im­ portant “memes” shaping the digital humanities commu­ nity. On the issues of theory, building, and hack versus yack, see also Cecire, “Introduction.” 3. Since the question of who is included in the digi­ tal humanities can be a sensitive one, I should clarify my understanding of the scope of the ield. While I focus in this essay on digital literary studies and one concrete re­ search example, the digital humanities are much broader. I thus take care to relate my example to shared methods and problems across the ield (including, e.g., digital his­ tory). However, while my goal is to address a fairly broad notion of the digital humanities as they are commonly practiced and discussed, I do not try to make the tent so big—by, for instance, covering what should or could be part of it but so far is not commonly recognized to be so (e.g., the design ield)—that the term digital humanities becomes formless or aspirational. Also, I do not specif ically discuss digital work in such near humanities as the branches of the social sci­ ences, archaeology, and anthropology that have joined the humanities in the so­ called linguistic and cultural turns and related trends. Whether such work is consid­ ered digital humanities depends on the prior issue of whether, digital methods aside, it is humanistic (itting, e.g., the loose deinition in the 1965 National Foundation on the Arts and Humanities Act: “those aspects of social sciences which have humanistic content and employ hu­ manistic methods” [National Endowment 1]). Caveats a lso apply to my genera l i z at ions about new media studies and to the gross simpliication here (given nuance below) by which I separate them from the digital humanities. For helping me think about these inclusion issues, I am grateful to members of the audiences at talks where I de­ livered versions of this essay who pointed out that they do not entirely recognize their ield or work in my description of the digital humanities. Ultimately, of course, the “who is in the digital humanities?” issue will be adjusted on the ground through normal professional processes of adjudica­ tion—where one gets a job or places one’s students, where one publishes in print or online, which forums or blogs one posts on, which Twitter hashtags one is associated with, which conferences one goes to, which grants one gets, etc. 4. On algorithmic criticism, see Ramsay’s essay by that title. 5. On #transformDH, see Phillips. he collective now appears through its hashtag on Twitter, has a Tumblr page, and is a HASTAC group. 6. Examples of the textual­ editing, bibliography­ as­ sociolog y­ of­ texts, and materiality­ of­ the­ digital ap­ proaches in the digital humanities include, respectively, McGann, Radiant Textuality; the Text Encoding Initia­ tive’s TEI: P5 Guidelines, with its attention to prosopog­ raphy and social relations (“Names, Dates, People, and Places”); and Kirschenbaum, Mechanisms. 7. Examples of the historical sociology I refer to in­ clude Franzosi; Bearman and Stovel; and Mohr and Duquenne. An example of social­ network analysis in dig­ ital literary studies is Moretti, Network heory. (On such analysis in general, see my “From Reading.”) An example of social­ network analysis in digital history is Lemercier and Rosental. (For other historians using the methods and tools of social­ network analysis, see “Bibliography.”) 8. See also Heuser and Le­ Khac’s “Learning to Read Data,” which summarizes the research reported in their pamphlet. 9. For a discussion of “saving the phenomena” that bears centrally on the issue of data, see Bogen and Woodward. 10. he full list of the novels in Heuser and Le­ Khac’s corpus can be found in their “Online Companion.” In­ formation on the source of the digitized texts is from an e­mail from Heuser to the author. 11. Goldstone and Under wood of fer an explana­ tion and example of topic modeling of special interest to readers of PMLA. Separately, Underwood more fully explains the methodology. For a succinct discussion of the diference between supervised and unsupervised data mining—a distinction originating in the ield of machine learning—see “Analytics.” 12. Sculley and Pasanek consider the problem of “cir­ cularity” (and other issues) in humanities data mining. 13. For a fuller description of the way Heuser and Le­ Khac used the HTOED, see appendix C in their Quantita- tive Literary History. 14. For this essay, I have consulted the online HTOED, which now appears as part of the online OED. On the 420 The Meaning of the Digital Humanities [ P M L A t h e c h a n g i n g p r o f e s s i o n HTOED’s classiication system, see “Structure of the Histor- ical hesaurus.” For additional general information about the work, see Kay, “What Is the Historical hesaurus.” 15. For information about the making of the HTOED referred to below, I have consulted Kay, “Classiication” and “What Is the Historical hesaurus”; Kay and Chase; and Wotherspoon, “Historical hesaurus” and “Making”; as well as Alexander’s e‑mail. For photos of the paper slips, iling drawers, and computers used, see Historical hesaurus Photo Gallery. My special thanks to Alexander for helping me gather resources and for his e‑mail illed with details and relections on the manual and techno‑ logical processes used in making the HTOED. 16. See, e.g., Gibbs and Cohen’s discussion of their hybrid method (70, 76). 17. I abbrev iate here a longer d isc u ssion of t he HTOED to be included in a version of this essay for my book in progress on the digital humanities. My thanks to Clare Birchall, who, in a conversation with me ater I presented an early version of this essay at King’s Col‑ lege, asked a question that made me start looking into the technological dimensions of the HTOED. 18. My thanks to the Stanford Literary Lab for invit‑ ing me to visit on 21 May 2012. On the genesis and prin‑ ciples of the lab, see Jockers. 19. In their statistical analysis of diction in En glish lit‑ erary works over a longer period (the eighteenth and nine‑ teenth centuries), Underwood and Sellers make a related point when comparing their indings with Heuser and Le‑ Khac’s. hey track a rise in the nineteenth century in the proportional incidence of earlier‑ vintage, oten Anglo‑ Saxon words in literary genres (including prose iction) that is “largely consubstantial” with the rising incidence of “hard” words in novels of that century found by Heuser and Le‑ Khac. However, they add that their longer histori‑ cal baseline shows that the trajectory of such words “had recently reversed direction” as part of a new way of be‑ ing “literary.” “he relative scarcity of simple action verbs in early‑ nineteenth‑ century writing, for instance,” they note, “was a recent development” in literary language that reveals social transformations only through the me‑ diation of “competing ideals of literary reinement.” My thanks to Le‑ Khac for his e‑mail calling my attention to Underwood and Sellers’s article, which I had not seen. 20. Advocating for the humanities does not necessar‑ ily only mean defending older or disciplinary notions of the humanities. My “Humanities and Tomorrow’s Dis‑ coveries” attempts to reframe the humanities in common cause with other disciplines and with public needs. WORKS CITED Adams, Michael, ed. “Cunning Passages, Contrived Cor- ridors”: Unexpected Essays in the History of Lexicogra- phy. Electronic ed. Monza: Polimetrica, 2010. PDF ile. Alexander, Marc. Message to the author. 6 Oct. 2012. E‑mail. ———. “‘he Various Forms of Civilization Arranged in Chronological Strata’: Manipulating the HTOED.” Adams 309–23. “Ana ly tics: Super v ised vs. Unsuper v ised Learning.” MONK: Metadata Of fer New Knowledge. N.p., n.d. Web. 1 Dec. 2012. Bauer, Jean. “Who You Calling Untheoretical?” Journal of Digital Humanities 1.1 (2011): 68–74. Web. 23 Nov. 2012. Bearman, Peter S., and Katherine Stovel. “Becoming a Nazi: A Model for Narrative Networks.” Poetics 27 (2000): 69–90. Print. “Bibliography.” Ed. Marten Düring. Historical Network Research. N.p., 18 Sept. 2012. Web. 26 Nov. 2012. Bogen, James, and James Woodward. “Saving the Phenom‑ ena.” Philosophical Review 97.3 (1988): 303–52. Print. Burdick, Anne, Johanna Drucker, Peter Lunenfeld, Todd Presner, and Jefrey Schnapp. Digital_ Humanities. Cambridge: MIT P, 2012. Print. Cecire, Natalia. “Introduction: heory and the Virtues of Digital Humanities.” Journal of Digital Humanities 1.1 (2011): 44–53. Web. 23 Nov. 2012. ———. “When Digital Humanities Was in Vogue.” Jour- nal of Digital Humanities 1.1 (2011): 54–58. Web. 23 Nov. 2012. “The Digital Humanities Manifesto 2.0.” Humanities Blast. N.p., 2009. Web. 9 Sept. 2012. Feyerabend, Paul. Against Method: Outline of an Anarchis- tic heory of Knowledge. London: Verso, 1978. Print. 4Humanities: Advocating for the Humanities. Ed. Chris‑ tine Henseler, Alan Liu, Geoffrey Rockwell, Stéfan Sinclair, and Melissa Terras, with Lindsay homas. 4Humanities, n.d. Web. 16 Jan. 2013. Franzosi, Roberto. Quantitative Narrative Analysis. Los Angeles: Sage, 2010. Print. Gibbs, Frederick W., and Daniel J. Cohen. “A Conver‑ sation with Data: Prospecting Victorian Words and Ideas.” Victorian Studies 54.1 (2011): 69–77. Print. Gold, Matthew K., ed. Debates in the Digital Humanities. Minneapolis: U of Minnesota P, 2012. Print. Goldberg, David heo, and Patrik Svensson, ed. Humani- ties and the Digital. Cambridge: MIT P, forthcoming. Goldstone, Andrew, and Ted Underwood. “What Can Topic Models of PMLA Teach Us about the History of Literary Scholarship?” he Stone and the Shell: His- torical Questions Raised by a Quantitative Approach to Language. N.p., 14 Dec. 2012. Web. 9 Dec. 2012. Hayles, N. Katherine. How We hink: Digital Media and Contemporary Technogenesis. Chicago: U of Chi‑ cago P, 2012. Print. Heuser, Ryan. Message to the author. 13 Sept. 2012. E‑mail. Heuser, Ryan, and Long Le‑ K hac. “Learning to Read Data: Bringing Out the Humanistic in the Digital Hu‑ manities.” Victorian Novels 54.1 (2011): 79–86. Print. 1 2 8 . 2 ] Alan Liu 421 t h e c h a n g i n g p r o f e s s i o n ———. “Online Companion to ‘Learning to Read Data: Bringing Out the Humanistic in the Digital Humani- ties,’ Victorian Studies 54.1; and Stanford Literary Lab Pamphlet 4.” Stanford Literary Lab. Stanford Lit. Lab, May 2012. Web. 13 Sept. 2012. ———. A Q u a n t i t a t i v e L i t e ra r y Hi s t o r y o f 2 ,9 5 8 Nineteenth- Century British Novels: he Semantic Co- hort Method. Stanford Literary Lab. Stanford Lit. Lab, May 2012. Web. 1 Sept. 2012. Pamphlet 4. Historical hesaurus. Oxford En glish Dictionary. Oxford UP, 2012. Web. 25 Nov. 2012. Historical hesaurus of the Oxford En glish Dictionary, with Additional Material from “A hesaurus of Old En glish.” Ed. Christian Kay, Jane Roberts, Michael Samuels, and Irené Wotherspoon. Oxford: Oxford UP, 2009. Print. Historical hesaurus Photo Gallery. U of Glasgow, n. d. Web. 25 Sept. 2012. Jockers, Matthew L. “houghts on a Literary Lab.” Mat- thew L. Jockers. N.p., 4 Jan. 2013. Web. 10 Jan. 2013. Jockers, Matthew, and Glen Worthey. “Introduction: Welcome to the Big Tent.” Digital Humanities 2011. N.p., 2011. Web. 23 Nov. 2012. Kay, Christian. “Classiication: Principles and Practice.” Adams 255–70. ———. “What Is the Historical hesaurus of the OED.” Oxford En glish Dictionary. Oxford UP, 2012. Web. 25 Nov. 2012. Kay, Christian, and T[homas] J. P. Chase. “Constructing a hesaurus Database.” Literary and Linguistic Com- puting 2.3 (1987): 161–63. Print. Kirschenbaum, Matthew G. “Digital Humanities As/ Is a Tactical Term.” Gold 415–28. ———. Mechanisms: New Media and the Forensic Imagi- nation. Cambridge: MIT P, 2008. Print. ———. “What Is Digital Humanities and What’s It Doing in En glish Departments?” ADE Bulletin 150 (2010): 55–61. Print. Koh, Adeline. “More Hack, Less Yack? Modularity, he- ory and Habitus in the Digital Humanities.” Adeline Koh. N.p., 21 May 2012. Web. 23 Nov. 2012. Latour, Bruno. Reassembling the Social: An Introduction to Actor- Network- heory. Oxford: Oxford UP, 2005. Print. Le- Khac, Long. Message to the author. 4 Feb. 2013. E-mail. L emercier, Cla i re, a nd Pau l- A nd ré Rosent a l. “ T he Structure and Dynamics of Migration Patterns in Nineteenth- Century Northern France.” Vers. 3. HAL: Sciences de l’ homme et de la société. Centre pour la Communication Scientifique Directe, 5 May 2010. Web. 24 Nov. 2012. Liu, Alan. “From Reading to Social Computing.” Price and Siemens. ———. “he Humanities and Tomorrow’s Discoveries.” 4Humanities: Advocating for the Humanities. 4Hu- manities, 25 July 2012. Web. 14 Mar. 2013. ———. “Is Digital Humanities a Field? An Answer from the Point of View of Language.” Alan Liu. N.p., 6 Mar. 2013. Web. 6 Mar. 2013. ———. he Laws of Cool: Knowledge Work and the Culture of Information. Chicago: U of Chicago P, 2004. Print. ———. “The State of the Digital Humanities: A Report and a Critique.” Arts and Humanities in Higher Edu- cation 11.1–2 (2012): 8–41. Print. ———. “Where Is Cultural Criticism in the Digital Hu- manities?” Gold 490–509. McCarty, Willard. Humanities Computing. Basingstoke: Palgrave, 2005. Print. McGann, Jerome J. “Position Statement from McGann.” Panel: What Is Text? A Debate on the Philosophical and Epistemological Nature of Text in the Light of Hu- manities Computing Research. By Susan Hockey, Allen Renear, and McGann. N.p., 1999. Web. 25 Nov. 2012. ———. Radiant Textuality: Literature af ter the World Wide Web. New York: Palgrave, 2001. Print. Mohr, John W., and Vincent Duquenne. “he Duality of Cul- ture and Practice: Poverty Relief in New York City, 1888– 1917.” heory and Society 26.2–3 (1997): 305–56. Print. Moretti, Franco. Network heory, Plot Analysis. Stanford Literary Lab. Stanford Lit. Lab, 1 May 2011. Web. 6 July 2012. Pamphlet 2. “Names, Dates, People, and Places.” P5: Guidelines for Electronic Text Encoding and Interchange. Vers. 2.2.0. TEI, 25 Oct. 2012. Web. 14 Jan. 2013. National Endowment for the Humanities, Oice of Digital Humanities. Digital Humanities Start- Up Grants: Fre- quently Asked Questions. NEH, n.d. Web. 17 Nov. 2012. Owens, J. B. “Jack.” “What Historians Want from GIS.” ArcNews. Esri, 2007. Web. 15 Jan. 2013. Phillips, Amanda. “#transformDH—a Call to Action Fol- lowing ASA 2011.” HASTAC. N.p., 26 Oct. 2011. Web. 24 Nov. 2012. Pickering, Andrew. he Mangle of Practice: Time, Agency, and Science. Chicago: U of Chicago P, 1995. Print. Price, Kenneth M., and Ray Siemens, eds. Literary Studies in the Digital Age: An Evolving Anthology. MLA Com- mons. MLA, 2013. Web. 12 Jan. 2013. Ramsay, Stephen. “Algorithmic Criticism.” A Companion to Digital Literary Studies. Ed. Susan Schreibman and Ray Siemens. Oxford: Blackwell, 2008. Alliance of Dig- ital Humanities Organizations, n.d. Web. 25 Nov. 2012. ———. “On Building.” Stephen Ramsay. N.p., 11 Jan. 2011. Web. 20 June 2011. ———. “Who’s In and Who’s Out.” Stephen Ramsay. N.p., 8 Jan. 2011. Web. 23 Nov. 2012. Schreibman, Susan, Ray Siemens, and John Unsworth, eds. A Companion to Digital Humanities. Oxford: Blackwell, 2004. Alliance of Digital Humanities Orga- nizations, n.d. Web. 27 Nov. 2012. Sculley, D., and Bradley M. Pasanek. “Meaning and Min- ing: The Impact of Implicit Assumptions in Data 422 The Meaning of the Digital Humanities [ P M L A t h e c h a n g i n g p r o f e s s i o n Mining for the Humanities.” Literary and Linguistic Computing 23.4 (2008): 409–24. Print. Seefeldt, Douglas, and William G. homas III. “What Is Digital History? A Look at Some Exemplar Projects.” Faculty Publications. Dept. of History, U of Nebraska, Lincoln, 1 May 2009. Web. 12 Oct. 2012. Paper 98. Stallybrass, Peter. “Books and Scrolls: Navigating the Bible.” Books and Readers in Early Modern En gland: Material Studies. Ed. Jennifer Andersen and Elizabeth Sauer. Philadelphia: U of Pennsylvania P, 2002. 42–79. Print. “he Structure of the Historical hesaurus of the OED.” Ox- ford University Press. Oxford UP, 2009. Web. 25 Nov. 2012. Svensson, Patrik. “Envisioning the Digital Humanities.” Digital Humanities Quarterly 6.1 (2012): n. pag. Web. 23 Nov. 2012. ———. “Humanities Computing as Digital Humanities.” Digital Humanities Quarterly 3.3 (2009): n. pag. Web. 30 Nov. 2010. ———. “he Landscape of Digital Humanities.” Digital Humanities Quarterly 4.1 (2010): n. pag. Web. 23 Nov. 2012. Underwood, Ted. “Topic Modeling Made Just Simple Enough.” The Stone and the Shell: Historical Ques- tions Raised by a Quantitative Approach to Language. N.p., 7 Apr. 2012. Web. 14 Jan. 2013. Underwood, Ted, and Jordan Sellers. “he Emergence of Literary Diction.” Journal of Digital Humanities 1.2 (2012): n. pag. Web. 4 Feb. 2013. Vismann, Cornelia. Files: Law and Media Technology. Trans. Geofrey Winthrop- Young. Stanford: Stanford UP, 2008. Print. Wotherspoon, Irené. “Historical Thesaurus Database Using Ingres.” Literary and Linguistic Computing 7.4 (1992): 218–25. Print. ———. “he Making of he Historical hesaurus of the Oxford En glish Dictionary.” Adams 271–87. 1 2 8 . 2 ] Alan Liu 423 t h e c h a n g i n g p r o f e s s i o n