What Do We Want from the Standard Core Texts of the Digital Humanities Curriculum?

From Textbook to Texts

Let me begin with a useful objection: In fact, one might respond, haven’t most digital humanities texts been “pedagogical” in the sense that they must address absolute beginners to the field alongside more experienced audiences? Certainly, this was true of much work in text analysis and text encoding: writers were addressing their students and potential entrants into the field, and perhaps secondarily colleagues and peers. One thinks here of Matthew Jockers’s Text Analysis with R for Students of Literature, Nick Montfort’s Exploratory Programming for the Arts and Humanities, Taylor Arnold and Lauren Tilton’s Humanities Data in R, and the excellent lessons offered in the Programming Historian. The point of these texts is the processes they teach and the thinking that they model, not necessarily the results of their research. At the other end of the spectrum, Franco Moretti’s famous entry in the field offered little to no description of the process of assembling the graphs, maps, and trees he used to rethink literary history.1

Pedagogical volumes in the digital humanities demonstrate the challenges of writing for an interdisciplinary field in which many of the research questions and paradigms are (often covertly) disciplinary. Books like the very good Advancing Digital Humanities (by Katherine Bode and Paul Longley Arthur) mostly speak to specific research communities and questions, as they should, with some overview of the interdisciplinary issues and debates they parse. Each of the most recent volumes of Debates in the Digital Humanities has served in coursework as an introduction to the field, but its orientation toward internal debate does not always serve the interest of novices well. “Teaching the controversies,” on Gerald Graff’s model, assumes that we have a shared framework for understanding the basis of the controversies and something like a shared cultural and disciplinary background.2 But DH pedagogy cannot assume familiarity with a disciplinary “canon” or even a mode of investigating that canon.

A secondary problem arises for pedagogical texts written for the “digital humanities” student, whether entry-level or advanced: such students, in their purest form, do not exist. Rather than entrants into digital humanities minors and majors (who exist at relatively few institutions), we mostly teach our students within existing programs and institutions, in classes adapted to our particular disciplines, or in interstitial spaces like libraries and humanities centers.3 As we do so, we inevitably encounter resistance to digital methods, students’ resistance and our own.4 In our own writing, a gap looms between books written for our colleagues—often in specialist subfields, not the “digital humanities” writ large—and the books we want to teach with. This tension between student and peer audiences arises everywhere, of course, but in the digital humanities the problem is particularly difficult, first because research methods change so quickly, and, second and more importantly, because our pedagogy often remains embedded within disciplinary structures (even where our research is not). Unless and until the digital humanities “vanishes” into specialist subfields, this tension will remain present in pedagogical texts.

For students working in digital humanities majors and minors, there is a genuine need for pedagogical textbooks that not only discuss the history of the interdisciplinary conversation and its current debates but also model inquiry, critique, and creation in the field. Claire Battershill and Shawna Ross discuss one version of this model, and Digital Pedagogy in the Humanities (edited by Rebecca Frost Davis, Matthew K. Gold, Katherine D. Harris, and Jentery Sayers) offers an effective introduction to key terms and areas of inquiry, one that addresses a range of readers. Students in an Introduction to Digital Humanities course could benefit from standard texts that draw on these digital resources and others, employing them for critical and creative work at the undergraduate level. For those of us working in disciplinary contexts and departments, such introductions may be less useful, for reasons I will explore later, but even in those environments an attempt at a standard “introduction to digital humanities” would be welcome.

But a standard text for digital humanities minors will look quite different from the pedagogical texts that we want to teach in the contexts of digital history, digital classics, digital literary studies, and so on. Researchers working in these disciplinary contexts will raise a useful objection: Could there ever be a “standard text” of “the DH curriculum” given that the latter is irredeemably plural?5 Isn’t our real concern in developing a DH pedagogy the teaching of that portion of digital techniques and conversations that engages our core disciplinary objects and concerns, whatever those might be? This way of looking at DH pedagogy sees our curricula as necessarily multiple, loosely intersecting, disciplinarily bounded, and not amenable to standardization. I agree with this objection, in part, and think it could help refine what we actually want from our pedagogical texts.

Versions of an “introduction to digital humanities” do exist, and others are being written. Room exists for a standard introduction to the history and ongoing conversation of DH for undergraduate entrants into the field, one that elaborates on the “companions” edited by Susan Schreibman, Ray Siemens, and John Unsworth, and for more limited introductions to digital history or digital classics, among many other subfields. But in my view, such introductions must remain carefully descriptive rather than prescriptive, critical rather than exemplary, firmly grounded within their own disciplinary and hermeneutic limits rather than attempting to model inquiry in the digital humanities in general. Since the digital humanities are indeed plural, situated at the intersection of competing disciplines, methods, critical traditions, political identities, and objects of interest, such introductions need to clearly identify their particular situation, problem, and audience. While a history of the conversation and its institutions can be written, the idea of one “standard text” that would unite all methods and disciplines is implausible. The Norton Introduction to Poetry, to take one capacious example, is united by object (poetry), field (English studies), and method (close reading): the idea of a parallel Norton introduction to digital poetics seems implausible simply because no consensus on object, field, or method could be assumed.

What we should seek instead of one standard textbook, as we look to teach within the disciplinary and interdisciplinary spaces that tend to define our pedagogical work, are multiple standard texts that firmly situate themselves and advance our understanding of core objects, questions, and methods of interest. These texts, whether written for an undergraduate course or for graduate methods courses, should, in my view, locate themselves clearly at the conjunction of a disciplinary context and an interdisciplinary method or object. Introductions to literary geography, data critique, media archaeology, electronic literature, or critical interface design are all imaginable (or exist) at exactly such conjunctions.6

What Do We Want from Our Texts: Criteria

Let me summarize what I see as desirable for “standard texts” of the DH curriculum. Such texts would need, first, to acknowledge the plurality of approaches and disciplines in the community itself, either through embracing multiplicity (the approach of Advancing the Digital Humanities and many Debates volumes) or by carefully limiting the approach to a specific object, goal, and community, while framing it within the larger conversation (the approach of some Programming Historian lessons). Second, texts should be oriented toward a specific humanistic research problem and discursively located within the history of the research literature on that problem, rather than describing the use of technology as if in a disciplinary vacuum—not “using MALLET on historical texts,” as if either MALLET or “historical texts” exist in a floating epistemological and methodological void, but instead positioning both within a clear epistemological framework, discussion of method, and continuing research question.7 Third, and in logical development, standard texts should address the research questions that we actually confront in humanist fields, rather than introductory questions with no clear audience. A corollary is that they should thus be carefully limited in scope and goals. Finally, standard texts should carefully separate their digital dependencies (whether in an apparatus or otherwise) from the lasting intellectual contribution they envision. I elaborate on these requirements later.

Because I see the digital humanities as plural even when contingently united by scholarly method, I advocate for developing pedagogical texts that clearly address and identify themselves by disciplinary community and subfield: “literary geography in the nineteenth-century French realist novel,” to give one example from Melanie Conroy’s work, rather than grasping at large-scale readings of “space” in “the novel,” two categories that are both too large and too vague to sustain a productive research question.8 In this respect, to make a brief parenthesis, the example of Moretti has not been salutary for the field of digital literary studies in particular, as his work depends on the Marxist tradition of macrosocial reading, along with a hypertrophic homological method in which every moment of a text (or map or graph) stands in intimate relation to capitalist modernity itself. Maps, graphs, and texts must always be mediated, in both the Marxist and media studies sense: understood as media forms that connect disparate ontological modes and scales of intellection. Moretti’s work depends on a continual challenge and polemic at the grandest scale of literary history, “the atlas of the European novel,” for example, a polemical model that does not always generalize well. Part of the condition for the acceptance of digital literary studies and Moretti was precisely his affiliation with a Marxist tradition of large-scale social and literary analysis, but the condition for acceptance has led to work that writes checks that do not quite cash out.

A related contention: pedagogical volumes in DH are most valuable when most carefully limited. Careful limitation is one of the things that makes the tutorials in the Programming Historian valuable, despite violating some of the other criteria just listed, and it is the reason I still teach Stephen Ramsay’s reading of the gendered vocabulary of The Waves.9 Ramsay’s approach to sorting out the key vocabulary that shares and divides the voices of The Waves takes seriously the underlying literary questions posed by Virginia Woolf and feminist scholarship, places that conversation alongside an argument for a particular kind of digital paratext within digital literary scholarship, and produces (for me at least) a moment of genuine revelation both in relation to the text and for digital literary studies. But the key moment of Ramsay’s argument is not its general discussion of the mechanics of reading gendered vocabulary, but rather the algorithmic re-creation of a paratext that lets us see The Waves (and by extension the gendering of dialogue in any text) anew. Standard core texts can and should be addressed to pressing research questions and workers in the field itself, not just to early entrants in the conversation; to actual writers and researchers, even in a survey course.

Because “digital humanities” majors are rare, but students at every level throughout the disciplines can practice digital humanities techniques, pedagogical texts should orient themselves toward the research questions on which we actually work: the making of Samuel Beckett’s En attendant Godot, as reconstructed in a digital genetic edition, rather than “Python for humanists” or a beginner’s guide to digital textual editions. My experience with teaching digital editions of Virginia Woolf’s Monday or Tuesday (1921) has been that if undergraduates are asked to learn either basic editorial tools or early experimental fiction in isolation, they have little context or interest; but if they are working toward a concrete digital edition that has never been done before, with original facsimiles of rare versions and manuscripts, within the context of a course that engages the personal and political traumas of 1917–21 that produced the experimental disjunctures of Woolf’s text, they can engage deeply and fruitfully. But that model of engagement has implications for the kinds of pedagogical texts we want as well: these texts must traffic in the deep, rich, but focused humanistic work of the kind we want from our classrooms, rather than technical tools and questions addressed to a broad and vague humanist audience. We do not require an “introduction to visualization,” but rather an account of how data visualization engages intersectional feminism, to take the example of Catherine D’Ignazio and Lauren Klein’s Data Feminism, or the long history of visualization as an intellectual activity, as in the case of Johanna Drucker’s Graphesis.

Because our tools and code dependencies change quickly, even as the underlying questions and objects do not, we need pedagogical books that thoughtfully articulate the discursive intellectual contribution (whether in print form or otherwise) to a separate (and separately reviewed) digital apparatus.10 Many current pedagogical volumes do this, in some form, and Quinn Dombrowski’s radically honest postmortem of Drupal for Humanists gives us clear reasons for doing so.11 All platforms are as grass, and the glories they produce as the flowers of the field. All coding languages, APIs, and platforms change continually, some change quite radically, and we have to be prepared for changes and endings both quick and slow. Entire paradigms of work in spatial humanities depend (or depended) on the Google Maps API, for example, and much thoughtful work was done with deprecated versions of that API: the lasting contribution of that work must be articulated in ways that outlive the platform on which they depended, and in pedagogies that outlive their temporary homes.12 Work created in these paradigms must be translated, remediated, and gracefully memorialized as a condition of their lasting value. Julia Flanders and Fotis Jannidis’s The Shape of Data in the Digital Humanities provides one account of how to do this, through their inquiry in modeling as intellectual activity separate from any given project or domain of digital inquiry. Ideally, standard texts would be able to update their dependencies and digital resources in some form, whether through a digital apparatus or a larger platform open to revision, like the University of Minnesota Press’s Manifold platform or the online Cambridge Core model. Such apparatuses and code dependencies should be reviewed both for replicability (does the code or link function?) and for pedagogical clarity (e.g., is the code usefully annotated and presented, for example through Jupyter Notebooks?).

The ideal for “standard texts” in the digital humanities would thus be clearly defined in scope, research question, and field, explicitly located in a tradition of humanistic inquiry, peer reviewed for reproducibility and clarity, and revisable at the level of digital apparatus and dependency. Rather than address all of the digital humanities at once, these texts should position themselves within a delimited research field and set of questions. In address, such introductory texts should aspire to the condition of the best mid-twentieth-century poetry “introductions,” accessible to the interested undergraduate reader but formalizing and deepening the knowledge of the expert as well.

Research as Pedagogy in Digital Literary Studies

How do the foregoing considerations play out in the construction of actual volumes? Let me address this concretely through the case of our midlength “minigraph” series from Cambridge, Elements in Digital Literary Studies (edited by Katherine Bode, Adam Hammond, and myself). Throughout the planning and proposal process, we had to think through the questions and concerns I have articulated and differentiate the series from other possible models. We decided to address “digital literary studies” in particular because we saw that formulation as an area both specific enough to shape work within and broad enough to appeal to a large range of communities (from digital textual editors to theorists of electronic literature).

This series is not only or avowedly pedagogical: it “defines and expands the elements of digital literary studies through a series of short exemplary texts,” midrange in length between monograph and article.13 Yet since not all of the communities and subfields that we address employ digital tools or methods, each volume must instruct the reader in the techniques it employs to achieve its particular literary-critical or literary-historical conclusions. These volumes are thus necessarily “teaching” as well as “research” texts, and they work to query the distinction between the two. A pedagogical or methodological component is necessary simply for peer review, we claim, let alone peer replication of the kind at issue in Nan Z. Da’s critique of “computational literary studies.”14 Most important to us was that each Element “address a pressing research question of clear scholarly interest within a defined area of literary studies and articulate clear conclusions on the literary insights achieved,” rather than addressing the digital humanities intramural conversation exclusively.

Yet there is a tension between achieving insights within a particular field and demonstrating for the reader the methodological validity of those insights, particularly if the method chosen is unfamiliar. One initial structure we have suggested as a solution to this problem is a methodological introduction, on the lines of a “methods” section in the natural sciences, which provides the equivalent of a pedagogical guide to replication, followed by chapters written with readers in the literary subfield as the primary intended audience. Some volumes could be imagined largely as primers in established methods in digital literary studies, closely tied to literary conclusions, without any division between method and result. Other kinds of volumes might annex their materials and methods to a digital appendix, on the model of Andrew Piper’s work in Enumerations, and remain focused on literary questions at hand. In his Element on literary generalization, Piper employs a mixed-methods approach that integrates the digital corpora and derived illustrations alongside the literary-critical text, a model that works particularly well in digital-first publication.15

Our central question in the series is that of the title: What methods, questions, and approaches should count as “elemental” for digital literary studies? That question turns out to be more controversial than one might think: electronic literature scholars have an entirely different notion of core skills and activities from digital medievalists or book historians. In DH teaching contexts, this is sometimes handled by addressing technologies rather than the interpretive communities that use them—XSLT rather than eighteenth-century letters—but we think the field is now at the point that technologies should be clearly and confidently tied to their humanistic payoff, even in an introductory pedagogical text.

What core elements of digital literary studies do we then imagine? In the following, I give a few examples, and many more could be imagined. Some familiar areas call out for a pedagogical text, in supplement perhaps to the ways they have been taught at the various teaching institutes in the digital humanities: encodings, for example, need to be addressed as a fundamental step in digital work, both at the level of the XML/TEI architecture built up for literary uses and at the level of the basic text encoding required for fundamental transformative work with textual corpora. Encoding is addressed by almost every digital literary studies book, but few sources I know outside the TEI guidelines think through the subject in any depth. Citation analysis is a fundamental skill that every graduate program should teach, formally or informally, as are the more complex forms of “search” at the level of patterns of syntax and semantics.16

Other kinds of core texts arise from specific single-author communities of interest: the Beckett Digital Manuscripts project comes to mind. But why not a “digital approaches to Henry James,” for example, using his works to define a limited test set against which to prove digital-literary hypotheses? Such a volume might ask whether late James has a kind of signature “late style” that compares to the late work of other writers, for example—here I am drawing on work in progress by Jonathan Reeve—or whether his lateness is an idiolect of James himself, a radicalization of his existing style. Digital correspondences also provoke new questions: How do the semantic, editorial, or correspondence networks of Willa Cather figure in relation to her prose? In what kinds of contexts and relations do her works (or the work of any writer in correspondence) gain new significance and prominence?

Enormous areas of contemporary digital textual production lack a guide to studying the field in a way that clearly addresses the digital dimension of the materials themselves. Think of fan fiction studies, which often brilliantly expound on the meaning of the materials for their participant-observers, but do not always address the formal dimension of those materials within the original networks of their production. Where is the standard text on reading the e-book, a task for both forensic materialists and platform studies? How should we read the algorithmically regenerated print-on-demand copy of a Victorian masterpiece that we assigned in quite a different version? What difference does it make to take those copies seriously, as literary and material objects? What economies do they live within? What reading communities do they affect?

In all these particular cases, we need to address our primary intellectual communities first, and secondarily the broader digital humanities conversation. We should remain committed to the importance of what has been built by all our communities even after the digital projects themselves undergo graceful degradation.17 Thinking through what we want from the core texts of the digital humanities curriculum, then, turns out to be an activity that moves outward toward specific disciplinary and interdisciplinary intersections, rather than inward toward a fixed “core curriculum.” Standard texts move as well in two temporal directions at once: forward, to new communities and conversations, and backward toward the intellectual legacies of lost projects, toward the intellectual foundations of our work, toward the past formation of human minds and lives in other moments of crisis.

My argument in this essay has been for a pluralist notion of digital humanities pedagogy mirrored in pluralized “standard texts,” texts located clearly in relation to intersections of method, discipline, and object. In their discussion of method, these standard texts should be descriptive rather than prescriptive, particular rather than exemplary, grounded within their own disciplinary and hermeneutic limits rather than attempting to model inquiry in the digital humanities in general. Standard texts should acknowledge the plurality of approaches and disciplines in the larger community and carefully frame their own object, goal, and audience in relation to those disciplines. Standard texts should address a clear research question or questions and locate themselves in the research literature on that problem, rather than addressing tools or methods in isolation. Standard texts should articulate digital dependencies to published volumes in ways that anticipate renovation, remediation, and preservation. Standard texts should address the questions and objects that humanists actually care about, in ways that further humanistic understanding. The consequence of these “standard texts” would be a pedagogy that is anything but standardized: and that is exactly the point.

Notes

1. 1. See Moretti, Graphs, Maps, Trees.

2. 2. Graff, Clueless in Academe.

3. 3. Thanks to Brandon Walsh for this formulation. See Cordell, “How Not.”

4. 4. Battershill and Ross’s Using Digital Humanities is exceptional in dealing with this problem explicitly.

5. 5. See Fitzpatrick, “Humanities.”

6. 6. See, for example, Rettberg, Electronic Literature; and D’Ignazio and Klein, Data Feminism.

7. 7. This is not to impugn thoughtful tutorials using MALLET, which acknowledge the disciplinary and epistemological limitations of their work: see, for example, Weingart, “Topic Modeling for Humanists.”

8. 8. Conroy, Literary Geographies.

9. 9. Ramsay, Reading Machines.

10. 10. Not all publishers have resources for a full review of code, for example, but ideally this would be done alongside standard peer review of the full manuscript.

11. 11. Dombrowski, “Sorry.”

12. 12. Here, I am thinking of the exceptional work captured in Presner, Shepard, and Kawano, HyperCities.

13. 13. These descriptions of the series and others are taken from Hankins et al., “Cambridge Elements.”

14. 14. See the controversy recorded in Critical Inquiry, “Computational Literary Studies.”

15. 15. Piper, Can We Be Wrong?

16. 16. On the latter point, see Shore, Cyberformalism.

17. 17. See Nowviskie and Porter, “Graceful Degradation Survey.”

Bibliography

1. Arnold, Taylor, and Lauren Tilton. Humanities Data in R Exploring Networks, Geospatial Data, Images, and Text. Cham, Switzerland: Springer International Publishing, 2016.
2. Battershill, Claire, and Shawna Ross. Using Digital Humanities in the Classroom: A Practical Introduction for Teachers, Lecturers and Students. 2nd ed. London: Bloomsbury Academic, 2022.
3. Bode, Katherine, and Paul Longley Arthur. Advancing Digital Humanities: Research, Methods, Theories. Hampshire, U.K.: Palgrave Macmillan, 2014.
4. Brooks, Cleanth, and Robert Penn Warren. Understanding Poetry: An Anthology for College Students. New York: H. Holt, 1943.
5. Centre for Manuscript Genetics. “Samuel Beckett: Digital Manuscript Project.” https://www.beckettarchive.org/home.
6. Conroy, Melanie. Literary Geographies in Balzac and Proust. New York: Cambridge University Press, 2021.
7. Cordell, Ryan. “How Not to Teach Digital Humanities.” In Debates in the Digital Humanities 2016, edited by Matthew K. Gold and Lauren F. Klein, 459–74. Minneapolis: University of Minnesota Press, 2016. https://dhdebates.gc.cuny.edu.
8. Critical Inquiry. “Computational Literary Studies: A ‘Critical Inquiry’ Online Forum.” In the Moment (blog), March 31, 2019. https://critinq.wordpress.com/2019/03/31/computational-literary-studies-a-critical-inquiry-online-forum/.
9. Davis, Rebecca Frost, Matthew K. Gold, Katherine D. Harris, and Jentery Sayers, eds. Digital Pedagogy in the Humanities: Concepts, Models, and Experiments. New York: Modern Language Association, 2020. https://digitalpedagogy.hcommons.org.
10. D’Ignazio, Catherine, and Lauren F. Klein. Data Feminism. Cambridge, Mass.: MIT Press, 2020.
11. Dombrowski, Quinn. Drupal for Humanists. College Station: Texas A&M University Press, 2017.
12. Dombrowski, Quinn. “Sorry for All the Drupal: Reflections on the 3rd Anniversary of ‘Drupal for Humanists.’” Quinn Dombrowski (blog), November 8, 2019. https://quinndombrowski.com/blog/2019/11/08/sorry-all-drupal-reflections-3rd-anniversary-drupal-humanists/.
13. Drucker, Johanna. Graphesis: Visual Forms of Knowledge Production. Cambridge, Mass.: Harvard University Press, 2014.
14. Fitzpatrick, Kathleen. “The Humanities, Done Digitally.” In Debates in the Digital Humanities, edited by Matthew K. Gold, 12–15. Minneapolis: University of Minnesota Press, 2012. https://dhdebates.gc.cuny.edu.
15. Flanders, Julia, and Fotis Jannidis. The Shape of Data in the Digital Humanities: Modeling Texts and Text-Based Resources. New York: Routledge, 2019.
16. Gold, Matthew K., ed. Debates in the Digital Humanities. Minneapolis: University of Minnesota Press, 2012. 
17. Graff, Gerald. Clueless in Academe: How Schooling Obscures the Life of the Mind. New Haven, Conn.: Yale University Press, 2003.
18. Hankins, Gabriel, Katherine Bode, and Adam Hammond. “Cambridge Elements: Digital Literary Studies: About This Elements Series.” Cambridge Core. Cambridge University Press. https://www.cambridge.org/core/what-we-publish/elements/digital-literary-studies.
19. Jockers, Matthew Lee. Text Analysis with R for Students of Literature. Cham, Switzerland: Springer, 2014.
20. Montfort, Nick. Exploratory Programming for the Arts and Humanities. Cambridge, Mass.: MIT Press, 2016.
21. Moretti, Franco. Graphs, Maps, Trees: Abstract Models for a Literary History. London: Verso, 2005.
22. Nowviskie, Bethany, and Dot Porter. “The Graceful Degradation Survey: Managing Digital Humanities Projects through Times of Transition and Decline.” Paper presented at the Digital Humanities Conference, London, July 2010. https://dh2010.cch.kcl.ac.uk/index.html.
23. Piper, Andrew. Can We Be Wrong? The Problem of Textual Evidence in a Time of Data. New York: Cambridge University Press, 2020.
24. Piper, Andrew. Enumerations: Data and Literary Study. Chicago: University of Chicago Press, 2018.
25. Presner, Todd Samuel, David Shepard, and Yoh Kawano. HyperCities: Thick Mapping in the Digital Humanities. Cambridge, Mass.: Harvard University Press, 2014.
26. Price, Kenneth, and Ray Siemens, eds. Literary Studies in the Digital Age: An Evolving Anthology. New York: Modern Language Association, n.d. https://dlsanthology.mla.hcommons.org.
27. Programming Historian. Programming Historian. https://programminghistorian.org.
28. Ramsay, Stephen. Reading Machines: Toward an Algorithmic Criticism. Urbana: University of Illinois Press, 2011.
29. Rettberg, Scott. Electronic Literature. Cambridge, U.K.: Polity, 2019.
30. Schreibman, Susan, Ray Siemens, and John Unsworth. A New Companion to Digital Humanities. Malden, Mass.: Blackwell, 2016.
31. Shore, Daniel. Cyberformalism: Histories of Linguistic Forms in the Digital Archive. Baltimore: Johns Hopkins University Press, 2018.
32. Van Hulle, Dirk, and Pim Verhulst. The Making of Samuel Beckett’s “En attendant Godot / Waiting for Godot.” Antwerp: University Press Antwerp, 2017.
33. Warren, Robert Penn, and Cleanth Brooks, eds. Understanding Poetry; an Anthology for College Students. New York: H. Holt and Company, 1939.
34. Weingart, Scott. “Topic Modeling for Humanists.” scottbot irregular (blog), July 25, 2012. http://www.scottbot.net/HIAL/index.html@p=19113.html.
35. Wellek, René, and Austin Warren. Theory of Literature. New York: Jonathan Cape, 1949.