A Keyword Analysis of Climate Change in Contemporary Literary Studies, 2000-2022

The Marxist scholar and critic Raymond Williams was one of the first to suggest the importance of keywords for literary studies. Originally published in 1976, Williams’s book Keywords: A Vocabulary of Culture and Society challenges conceptions of everyday terms through historical analysis. Specifically, Williams’s keywords demonstrate how shared discourse changes over time. As Williams notes, “When we come to say ‘we just don’t speak the same language’ we mean something more general: that we have different immediate values or different kinds of valuation, or that we are aware, often intangibly, of different formations and distributions of energy and interest” ¹. Studying conceptual keywords in this view not only enables insights into semantic shifts but also reveals how historically contingent conditions embed themselves within such shifts. As such, composing keyword entries revealed to Williams that “words which seem to have been there for centuries, with continuous general meanings, have come in fact to express radically different or radically variable, yet sometimes hardly noticed, meanings and implications of meaning” ¹. These insights have since generated a rich tradition of keyword analysis projects in the humanities, including but not limited to select critical theory books, the “Living Lexicon” series from the journal Environmental Humanities , as well as the _Keywords _ series from New York University Press, which includes entries on disability, gender and sexuality, and environmental studies.

Since Williams, scholars have relied on keywords in reading, writing, and teaching. For many, a different kind of keyword engagement influences research methods. Keyword indexing in databases such as the Modern Language Association (MLA) International Bibliography facilitates access to published scholarship’s metadata. Selecting effective keywords on such databases enables the act of finding, citing, and responding to scholarship. Scholars interested in climate change, for instance, search databases for articles that effectively deal with the topic’s complexity, which, as Andrew Ross [2016] notes, connects to historical ramifications for scientific knowledge and planetary, geopolitical conflict ². However, it is important to note for not only contemporary literary studies but also the digital humanities that disjunctions in database indexing impede digital access to scholarship. In this article, I explore how select academic journals from contemporary literary studies exemplify this disjunction through the occurrence of climate change as a keyword. Applying the VOSviewer bibliometric analysis tool, I analyze thousands of unique citations in four academic journals representative of work in contemporary literary studies that scholars in the field consider trendsetters. Large datasets of bibliographic data allow us to survey scholarship through their indexed citations and authorial networks ^{3 4 5 6 7}. Digital bibliometric analyses therefore enable scaled-up analyses of how keywords impact precise framing of and engagement with indexed research.

The field of contemporary literary studies presents a valuable case study in how we categorize humanistic scholarship about climate change and related environmental issues because its scholarship reviews direct representations of climate change. Owing to established fields such as ecocriticism, the discipline enjoys many conferences, journal articles, and monographs dedicated to climate change ^{8 9 10 11}, and scholarship will only proliferate in the coming years. Digital humanists will thus benefit from looking at the use of keywords in contemporary literary studies. The field reflects ongoing some of the more significant developments in humanities scholarship on myriad environmental issues, including but not limited to climate change. Deriving insights from a bibliometric analysis of contemporary literary studies will, in turn, suggest new possibilities for indexing scholarship that encourage engagement and access. While I go into more detail about the results below, this project’s central purpose is to identify how climate change occurs as an object of scholarly inquiry in the humanities through its use as a keyword in contemporary literary studies.

Corpus and Methods

The following sections encompass a bibliometric analysis of four journals from contemporary literary studies: ISLE: Interdisciplinary Studies in Literature and Environment ( ISLE ); Contemporary Literature ( CL ); MFS Modern Fiction Studies ( MFS ); and PMLA: Publications of the Modern Language Association ( PMLA ). I analyze the _ISLE _ corpus as a reference case for the occurrence of climate change as a keyword in a journal explicitly interested in environmental issues. With this corpus as a basis, I compare the other three corpora as benchmarks of contemporary literary studies scholarship. In selecting journals for analysis, I gathered indexed citations from journals with approximately 500-750 results in the MLA International Bibliography. These results derive from filtering out non-English language results and citations that occur prior to 2000. The _ISLE _ corpus counts 665 unique citations; the _CL _ corpus counts 620; the _MFS _ corpus counts 763; and the _PMLA _ corpus counts 1,964. Though I considered journals such as _Environmental Humanities _ and Resilience: A Journal of Environmental Humanities as supplemental reference cases to ISLE , they were excluded because their indexed citation counts in the MLA International Bibliography were less than 500 at the time of data collection.

Although the periodization of contemporary literature is highly contested ¹², my scope includes literature published from 1985 to the present while limiting scholarship on said literature from 2000 to early 2022. I adopt this timeline because I am most interested in scholarship that addresses recent representations of climate change that benefit from engagement (or lack thereof) with the past three decades of climate science. At the same time, I acknowledge that these journals — besides ISLE — do not tally significant counts of climate change as a keyword because the topic was likely peripheral to their interests up until the last decade. However, I find them valuable test cases because they have since published on the topic and, thus, can be analyzed as a corpus with keyword connections that either do or do not reflect the topic’s many connections. With these variables in mind, I assembled each journal’s corpus, limiting their indexed citations in the MLA International Bibliography to English-language citations from 2000 to the present. ¹³ I then exported the citations as research information system files to process with VOSviewer, a bibliometrics tool with modularity algorithms for visualizing author networks and keyword occurrence clusters ¹⁴. I initially isolated keywords that occur at least ten times in the top fifty keywords across the run of journal articles and book reviews, removing keywords that convey period duration or generalist categories. ¹⁵ I began with 10 keyword occurrences because that number suggests a keyword’s relative frequency in a corpus of at least 500 citations. If a climate change cluster did not emerge at ten keywords, I decreased the occurrence count to five and, if necessary, one keyword to find the climate change keyword cluster. I simultaneously increased the top count in permutations of 50, 100, and 1,000, while decreasing the occurrence count (if a cluster does not emerge in the top fifty, it will emerge at subsequent scales). I followed these steps to determine the relative significance of climate change as a keyword in each reviewed corpus.

While this article derives inspiration for bibliometric analyses elsewhere in literary studies ¹⁶, it also follows bibliometric analyses in applied environmental sciences that propose new avenues for research ^{17 18 19}. Reconciling differences between these projects will inform framings of climate change in digital humanities research and points to the utility of digital interventions in the environmental humanities. Survey approaches of contemporary literary studies, ecocriticism, and the environment humanities precede this project, as we see in select articles from Ursula K. Heise [2006a] [2006b] [2008] [2013]. But because digital humanities tools can analyze thousands of citations in databases, scaled-up bibliometric analyses can determine common terms of engagement between otherwise disconnected disciplines. As such, I focus on climate change as a keyword because keywords such as climate malaise, climate fatigue, and climate perturbation are not commonly recognized. The project also must rely on the keywords provided by the indexers of the MLA International Bibliography: climate change is the keyword because it is what appears on the database. I follow this parameter because climate change is a discursive mainstay, a keyword in the truest sense of Williams’s work. For example, a brief search for publications with Environment in the Scopus database reveals the predominance of climate change as a keyword (both for authors and the database indexers) in pertinent research. Of 2,000 publications surveyed, selecting for at least 10 keyword occurrences, the keyword for climate change occurs in 188 articles with 723 links to other pertinent topics such as carbon, drought, and climate models (Fig. 1). ²⁰

While not the leading keyword in the Scopus corpus (these being article, controlled study, and nonhuman), climate change is in the top 15, indicating its relative significance. As a keyword, then, climate change presents significant inroads for a variety of ecological issues with which literary studies, as well as the digital humanities, will benefit from informed engagement.

When evaluating the corpora, I do not critique any author, journal, or publisher for how a given citation does or does not employ certain keywords. This is in part because the MLA International Bibliography’s keyword selection process is unclear. The Modern Language Association’s website for the database notes the number of indexers who cite publications and the thousands of terms they use for indexing ²¹, but does not mention if keywords derive from authors, journals, or other parties. I also do not know if the initial stage of keyword selection affects later indexing of pertinent metadata. I therefore survey this corpus of journal citations mindful of the many components that result in a keyword’s appearance in a scholarly database. The article’s conclusion considers how responding to these limitations may inspire more effective indexing practices for future digital environmental humanities projects, in addition to future contemporary literary studies and digital humanities scholarship (not to say that the two are distinct). These digital environmental humanities projects include not only analyses of emergent cultural forms ^{22 23 24}, but also ongoing conversations about how the digital humanities will respond to climate change through its methods, tools, and critiques of extant technologies ^{25 26 27}.

Reference Case: ISLE

Since its 1993 debut, _ISLE: Interdisciplinary Studies in Literature and Environment _ (hereafter ISLE ) has served as a hub for environmental scholarship in literary studies. As its editors wrote in its first issue, ISLE ’s founding “reflects the rapid growth of ecological criticism in the United States in recent years, which in turn reflects the steady increase in the production of environmental literature and nature writing over the past several decades and the attendant increased visibility of such writing in college literature courses” ²⁸. At the time of writing, 29 volumes of _ISLE _ have been published as part of the Association for the Study of Literature and environment. The journal’s mission statement pinpoints its influence as a pillar for literary studies: “Celebrating the rich confluence of environmental humanities, ecocriticism, and environmental justice, ISLE welcomes submissions from authors creatively engaging these fields from a broad range of disciplines, geographies, and perspectives. ISLE invites scholarly articles and creative writing that interpret the environment in complex, imaginative, and generative ways” ²⁹. Owing to the journal’s long publication history and dedicated interest in the ways cultural texts grapple with environmental issues, and how these issues are explored through such genres as climate fiction and ecopoetics, _ISLE _ presents a fruitful archive for bibliometric analysis. I therefore refer to _ISLE _ as a reference case that generates an array of keywords for literary studies scholarship concerned with the environment.

The _ISLE _ corpus presents multiple occurrences of climate change as a keyword. The VOSviewer tool presents clusters that cohere around significant keyword occurrences at various scales. Links between clusters suggest the relative significance of keywords to other relevant topics of interest. An initial reading of the _ISLE _ corpus suggests the significance of keywords (referred to as subject terms in the MLA International Bibliography) such as ecocritical approach, american literature, and nature (Fig. 2), many of which characterize the initial stages of ecocriticism in the US.

While these keyword clusters are the most frequent at a scale of 10 occurrences in the top 50, they are not the most recent. More recent clusters of significance include postapocalyptic novel, environmental crisis, and climate change. The environmental crisis keyword occurs 85 times, with 29 links to other clusters; the climate change keyword occurs 18 times with 13 links. As we can read environmental crisis in several ways, a discrepancy emerges between these two keywords in the ISLE corpus. Does environmental crisis refer to climate change, a mass extinction event, or something else entirely? Does indexing environmental crisis preclude the mention of “climate change,” making the keywords mutually exclusive?

Closer analysis of the _ISLE _ clusters hardly distinguishes the keywords. Scaling down from ten to five keyword occurrences produces environmental crisis and climate change nodes with more connections in the top 50: the former adds 23 links for a total of 52 connections while the latter gains three for a total of 16. The node for environmental crisis links with meat, petroleum industry, and social justice. The node for climate change does not link to any nodes not already connected to environmental crisis, suggesting that the conceptual keyword of environmental crisis includes but is not limited to climate change. Three ISLE articles in the corpus indexed with climate change are also indexed with environmental crisis ^{30 31 32}. Our empirical knowledge of climate change daily evolves with applied modeling and research,³³ and, whether in popular media coverage or in everyday conversation, we commonly do not associate the phrase environmental crisis with ecological catastrophe more than we do climate change. As a conceptual keyword, climate change is commonplace among environmental humanists, scientists, and policymakers alike. At the same time, the conceptual keyword environmental crisis may lend itself to complexities that climate change does not present because the former touches on issues such as pollution, biodiversity loss, and changing ecosystems. These more specific problems and their local occurrences are not always direct effects of climate change. Consequently, indexing an article with the keyword environmental crisis may misrepresent its account of environmental issues that require further contextualization.

It should not surprise us that climate change occurs significantly in the _ISLE _ corpus. Yet it is surprising that the environmental crisis cluster seemingly absorbs the climate change cluster in that the former attaches to topics discussed in concert with climate science. Though I argue that the distinction between environmental crisis and climate change remains significant, the _ISLE _ reference corpus presents a valuable example of wrestling with the two keywords as only two among many in its set. Now that we have seen a sampling of keywords from the _ISLE _ corpus, we can see how these keywords emerge, or do not emerge, in the other corpora that are not exclusively concerned with the environmental humanities.

Test Cases: Contemporary Literature, MFS, and PMLA

This section reviews the Contemporary Literature ( CL ), _ MFS Modern Fiction Studies_ ( MFS ), and _Publications of the Modern Language Association _ ( PMLA ) corpora. While these journals have not historically foregrounded climate change as an object of inquiry, I look to their corpora to consider how their recent keywords reflect contemporary literary studies’ emergent interests in climate change. With the _ISLE _ keywords as a reference for work elsewhere in literary studies, I am interested in analyzing how journals indexed in the same database do, and do not, engage similar and dissimilar keywords. I begin with CL , which publishes scholarly articles and book reviews. CL published the “first articles on Thomas Pynchon and Susan Howe and the first interviews with Margaret Drabble and Don DeLillo; it also helped to introduce Kazuo Ishiguro, Eavan Boland, and J. M. Coetzee to American readers” ³⁴. As of late, CL has also published ecocritical and environmental humanist scholarship. The most recent volume in the corpus includes articles that consider Black American conceptions of land use ³⁵, postapocalyptic environmental conditions in the novel Red Plenty ³⁶, and a review of scholarship on the Anthropocene and cultural production in the Global South ³⁷.

In the CL corpus, climate change as a keyword occurs infrequently. At 10 occurrences in the top 50, no climate change or environmental crisis clusters occur. Scaling the corpus down further by selecting for five keyword occurrences in the top one hundred produces ecocriticism and environmental crisis nodes quite distant from the network’s center (Fig. 3). Each cluster numbers six occurrences, indicating their relative insignificance in the corpus.

The occurrence of environmental crisis changes at subsequent scales. Selecting for one keyword in the top five hundred, environmental crisis occurs six times with eighteen links. Selecting for one keyword in the top seven hundred-and-fifty, the occurrence count of environmental crisis remains the same but adds new links. A cluster for climate change does not occur at any of these scales. We must scale down considerably to find a cluster: selecting for one keyword in the top one thousand registers a single occurrence of climate change with four cluster links.

At first glance, these results corroborate the MLA International Bibliography’s listing of CL ’s citations: only one article in the last 20 years has been indexed with climate change ³⁸.³⁹ This does not mean the corpus does not include citations geared toward prominent environmental issues, as the environmental crisis keyword surfaces when scaling down. The six citations indexed with environmental crisis are not indexed with climate change, though they are indexed with important topics such as global warming, pesticide, poisoning, and violence. In the corpus, these citations include both reviews ^{40 41 42 37} and articles ^{43 44}. Because the environmental crisis cluster in the _CL _ corpus is attached to a constellation of topics, the corpus presents a notable discursive gap between the use of environmental crisis and climate change. If the latter is used seldomly, the former is used in reference to important topics that may, or may not, be discussed separately from climate change.

I now turn my attention to the _MFS _ corpus. _MFS _ publishes articles and book reviews concerned with twentieth- and twenty-first century literature on a quarterly basis, with an emphasis on modernist, postmodernist, and contemporary fiction. A journal with a prestigious publishing record of nearly 60 years, MFS is a benchmark publication for innovative debates and topics in contemporary literary studies. Of note to this article is the fact that _MFS _ has published multiple special issues concerned with climate change and related environmental issues. These special issues include “Modern Fiction and the Ecological: The Futures of Ecocriticism” ⁴⁵, “Anthropocene Fictions” ⁴⁶, and “Literature and Extraction” ⁴⁷; the first special issue in that list published prominent environmental humanist Rob Nixon’s early writing on slow violence ⁴⁸.

Compared to the CL corpus, the occurrence of climate change as a keyword in the _MFS _ corpus differs greatly. The emergence of an environmental crisis node at the scale of 10 keyword occurrences in the top 50 suggests the cluster’s relative significance. However, no climate change cluster emerges. Scaling down to five keyword occurrences in the top fifty and top one hundred returns the same result. Scaling the corpus down to one keyword in the top five hundred produces a climate change cluster with three occurrences and fourteen links to other keywords (Fig. 4). Several clusters related to the environment occur at this scale, including ecology, ecocritical approach, environmental movement, natural gas industry, petroleum industry, and plant imagery.

Different keyword connections for climate change and environmental crisis characterize their occurrence in the _MFS _ corpus. While the former keyword primarily occurs around geopolitical clusters such as united states imperialism and afghanistan war (2000-2021), the latter connects to petroleum industry, nature, and ecocritical approach. While I do not mean to focus on these two keywords throughout the essay, these distinctions are worth noting, as many of the environmental crisis connections are relevant to any literary studies account of climate change. Three citations appear that explicitly address climate change ^{49 50 51}; searching elsewhere in the corpus, citations with environmental crisis represent a range of important but different environmental issues.

Databases such as the MLA International Bibliography operate with directed user engagement, and the difference between keyword selection may elide opportunities for accessible engagement. Regarding such engagement in the _MFS _ corpus, we can turn to Caroline Levine’s article “The Long Lure of Anti-Instrumentality: Politics, Aesthetics, and Sustainability” as an example of disjunctions in keyword indexing. Levine’s article asks readers to revise conceptions of literature’s anti-instrumentality in the context of climate change’s emergence ⁵². Levine situates this argument for literature’s instrumentality in the context of extant issues such as climate migration, resource allocation, ecological degradation, and climate denialism and pessimism ⁵², Levine’s article provokes new ways of thinking about climate change and contemporary literary studies in scholarship and pedagogy. In the MLA International Bibliography, Levine’s article is indexed with three keywords: themes and figures; aesthetics; and left-wing politics.⁵³ Climate change is absent from this list. Yet the term climate is used 18 times throughout the article, with 8 unique instances of climate change. In highlighting the indexing of Levine’s article, I do not aim to criticize any party for the absence of climate change as a keyword; we should still read, consider, and cite the article regardless of how it is indexed. While reading through journals may remain the best way to assess their scholarship, many contemporary scholars instead rely on Googling, library search portal browsing, and databases. Scholars’ targeted research practices help them get what they need when they need it (especially when increased teaching loads and immensely reduced funding opportunities undermine research capacities), and precise keyword selection makes the difference. Yet the absence of climate change as a keyword impedes our engagement with articles such as Levine’s that discuss it with significance.

My final test case, the _PMLA _ corpus — which is significantly larger than either the _CL _ or _MFS _ corpora — publishes scholarship on a diverse range of fields and periods in literary studies, including contemporary literature. The journal’s active cultivation of novel approaches speaks to its importance, not only as the cornerstone publication of the Modern Language Association but also as a publication for established disciplinary conversations. As its website states, “The ideal  PMLA essay exemplifies the best of its kind, whatever the kind; addresses a significant problem; draws out clearly the implications of its findings; and engages the attention of its audience through a concise, readable presentation” ⁵⁴. PMLA is also a competitive site for submission: disciplinary journals may therefore not be as receptive to relatively new work, or even submissions from graduate students and junior scholars ⁵⁵. However, PMLA has published work relevant to contemporary literary studies, the digital humanities ^{56 57}, the environmental humanities, and even an early iteration of digital environmental humanities referred to as the “ecological digital humanities” ⁵⁸, even as much of the journal’s writing offers surveys of scholarship in their respective topics.

Starting with 10 keyword occurrences in the top 50, no environmental crisis or climate change clusters emerge in the corpus. The notable absence of climate change is still visible when we scale for at five keywords in the top fifty and top one hundred. At each of these scales, zero clusters emerge for topics associated with environmental issues. A cluster for climate change appears if we select for one keyword occurrence in the top five hundred: the keyword occurs six times and links with twenty-six additional clusters (Fig. 5).

Additional clusters associated with environmental issues emerge at this scale, including nature, ecocriticism, landscape, and even anthropocene epoch. In two unique citations, climate change is indexed with nature; in two other citations, with weather (also seasons of the year in one of those two); in another unique citation, with environmental studies; and, in a final unique citation, pollution and industrialization ^{59 60 61 62 63 64}. In the PMLA _ corpus, then, climate change as a keyword enjoys distinct cluster connections that speak to different yet interconnected topics. For example, we have Philip Steer’s [2021] recent article on nineteenth-century Britain, Australia, and the Anthropocene with climate change as a keyword. Environmental crisis occurs as a keyword twice, and neither citation is indexed with climate change or the above terms ^{65 66}. Is there a tradition of selecting environmental crisis in lieu of or as complement to climate change in ISLE , CL , and MFS but not in PMLA ? This difference may account for the concerns of field- or period-based journals versus disciplinary journals. ISLE , _ CL , and MFS publish scholarship with foci relevant to scholars; _PMLA _ publishes scholarship with a broad focus for multiple fields. Regardless of their differences, each corpus presents different accounts of climate change as a keyword.

Conclusion: Keywords and Digital Environmental Humanities

How fields such as contemporary literary studies and the digital humanities respond to climate change requires shared conceptual keywords that extend beyond disciplines. In the future, we may wish to derive inspiration for these terms from various keyword projects ^{67 68}. In addition, we may write a publicly available keyword selection guide for precise database indexing and engagement: Lauren Klein and Catherine D’Ignazio’s [2020] discussion of datasheets as a contextual tool for big datasets may serve as first step toward realizing this approach, encouraging what Johanna Drucker may refer to as a more sustainable digital humanities praxis [2021]. In this case, I suggest a specific lexicon for environmental issues and how they connect to climate change. Although some may say that all humanities work is in some way imbricated in climate change ⁶⁹, I argue that only humanities work that explicitly addresses climate change should merit attention as climate humanities scholarship. For instance, an essay that addresses topics such as climate-induced biodiversity shifts, climate engineering, or climate fiction should be indexed with climate change only if the essay makes a significant connection between said topics and climate change. Such meaningful connections allow us to avoid the nebulousness of environmental crisis, which does not offer as many interdisciplinary inroads. This is not to overlook the fact that keyword indexing can be an imperfect practice; however, I hope my results show that it can be standardized for renowned publications interested in publishing on the topic.

A systematic adoption of keywords, as Williams writes, grants us “a vocabulary to use, to find our own ways in, to change as we find it necessary to change it, as we go on making our own language and history” ¹. Whether one works in contemporary literary studies or the digital humanities (or both), following similarly determined examples of keyword selection will ensure continued collective meaning making in scholarly discourse. Students, scholars, and teachers interested in digital environmental humanities will also benefit from deliberate indexing choices. As the digital environmental humanities take on new, forms, effective indexing practices that precisely frame discussions of climate change in teaching, scholarship, and outward facing projects will enrich the field’s interdisciplinary significance. At the same time, because climate change is a phrase with meaning beyond academic indexing, it can be a keyword that encompasses diverse yet globally interconnected environmental issues; subsuming it within the environmental crisis hierarchy only dilutes outward access to such scholarship.

Acknowledgements

My thanks to Matt Hannah for feedback on the project in its many stages. My gratitude, as well, to the DHQ reviewers for their suggestions.