Behind the scenes: The language of retrenchment—A discourse analysis of budget cutting in higher education.

While the 2016 and 2017 academic year was personally traumatic because of the presidential election, it was also a time of upheaval for many institutions of higher education. As a result, many institutions underwent budget cuts, which is a perpetual condition at many institutions and in the field of higher education in general.

The impetus to do this critical discourse analysis study on budget cutting emails came from the use of a single phrase: “frost.” Used in an email to describe an institution’s approach to halting spending costs, the term “frost” stood out to me as curious. It contrasted with a term I was more used to seeing in communications about cost cutting measures across the business world: “freeze.” Freeze is most often used to describe decisions made to halt hiring, such as in a “hiring freeze.” However, the use of the word “frost” to describe a general directive to halt spending across an institution including a hold on hiring was immediately interesting. Why was frost being used instead of freeze entail? Did the use of frost entail more impermanence than a freeze (as in, the frost will melt soon, and spending can resume)? Or was frost an indicator of more serious freezes to come, as the first frost of a season indicates that the worst, meaning the true cold, is yet to come. Was the use of frost an institutional way to say, “Winter is coming?”

This use of the word frost came from an institutional email communicating information about budget cuts at a urban, public higher education institution. As I explored that email further, I found more discursive and linguistics areas of interest. Overall, the email used language in a way that made it feel like real information was being communicated when the actual nature of the information presented was actually masked or hidden, such as in the use of “frost” to hide what exactly was being cut, who these cuts would affect, and for how long the hold would last. Instead, from that first email, it was clear that the emails were designed to generate support from those reading the emails without actually communicating much about what institutional leadership was asking them to “buy” into. And, thus, a discourse analysis project was born.


Like my prior critical discourse analysis projects, I followed the same steps in this CDA. After identifying the corpus, which consisted of emails from the president to faculty, staff, and stakeholders (publicly available as public communications) at two public institutions (one a large, urban, public institution; the other a small, public, regional institution), I returned to the literature to find out what prior linguistic analyses had been done on retrenchment communications to faculty, staff, and the public. Surprisingly (or not, depending on your perspective), not much existed in the literature about the communications or language used in these processes; much more had been written about the type of cutting done in times of financial stress or the decision-making processes used to make retrenchment decisions. What existed was limited to two studies, one by Slaughter (1993), and a CDA by Ayers (2014). Those studies found that language use was designed to communicate the nature of the crisis and generate support from stakeholders into the plan created to address that budget crisis.

Since limited research existed on the linguistic tools used to communicate these messages, I explored my favorite linguistic analysis of higher education language-in-use, Biber’s (2006) University language: A corpus-based study of spoken and written registers. I checked this book out so many times via ILL, that I finally realized I just needed to buy it. Through review of the text and my own reading of the emails, I identified pronoun use, stance, and nominalization as key linguistic tools through which I could explore how the emails communicated budget cutting behaviors. First, pronoun use was important to analyze because pronouns are often used to create an “us” versus “them” positioning to generate support for an idea or message. Second, pronouns can be used to mask who is actually making decisions (such as in the use of “we” when describing decisions made by one or a small group instead of the whole. Second, stance was important to explore because it provided insight into how the crisis and budget cuts were being communicated (e.g., were they being presented as urgent? Incontrovertible fact?). Finally, I explored nominalization to understand how verbs and adjectives were being turned into nouns which, in the process, obscured their meaning (This is what was happening with the use of “frost.” By calling a halt to spending a frost, all information about what was included in the frost was hidden from the reader.) These three tools would provide the linguistic foundation for the thematic “discourse” analysis that was the last stage of the data analysis process. In this study, we were exploring how discourses persisted across the corpus and how these discourses reinforced both the conclusions drawn in the linguistic analysis and how they perpetuated systems of power in higher education. As with other “Behind the Scenes” posts, we discuss the results in the full manuscript, which you can find here.

Path to Publication

Initially, we (my coauthors and I) were unsure about where this manuscript belonged. Among the three of us, I was the methodology “expert” and conducted the analysis and did most of the writing, while my co-authors were experts in financial decision-making in higher education. I, as quickly as I could, caught up on what I didn’t know about retrenchment decision-making in higher education, but the limited amount of research I could find meant that I also did not have a lot of foundational literature to build on and journals to reference that had published similar work before. As a result, we initially submitted to a linguistic journal that quickly rejected it for reasons I do not quite remember. It seemed clear, however, from the feedback, that we needed to find a higher education specific journal for this work. I set off on a google journey to find a journal where it might fit. Because of my limited knowledge of higher education organization and administration, and I had written the majority of the manuscript including implications, I was hesitant to submit to an org and admin-focused journal. Instead, we settled on the Journal for Post-secondary and Tertiary Research (JSPTE) out of the University of Idaho. JSPTE’s focus on higher education more broadly and in an international setting made it an excellent home for this work. Their review process is delightfully transparent, and we received an R&R in the normal timeframe with clear directions on how to improve and refine our manuscript. After turning around those edits, the manuscript was accepted and, then, published. It was a smooth, well-organized, and clear process.

Final Thoughts

As this post and all of my posts in my CDA projects have made clear, I have a singular approach to critical discourse analysis, one that stands apart from other discourse analysis projects I have read recently in the higher education literature. Many of those projects focus on “discourses” as evidence of Discourses of power, or they have focused on discourses as analysis of a reciprocal communication. In contrast, while I explore what I call “big D” discourses in the final stage of a CDA project, my analysis always begins by understanding the language used to communicate those discourses. For me, beginning with linguistic decisions often provides insight into the Discourses I will identify later and provides evidence of how those Discourses are being communicated. This is often easier to do in the text-based documents I analyze. However, I do not suggest that there is more or less validity in the way I conduct CDA project, just that there is a reason I like to begin and root my conclusions in linguistic choices.

Second, if you are reading this on the day it is published (November 13, 2019) or the following day (November 14) and are attending the ASHE conference in Portland, OR. You can join us for a discussion of the paper on November 14, 2019 at 2:15 PM in Hilton Portland Downtown, 3rd floor, Directors room. In that presentation, we will focus the bulk of our time talking about the results.


Ayers, D. F. (2014). When managerialism meets professional autonomy: The university ‘budget update’ as a genre of governance. Culture and Organization, 20(2), 98-120.

Biber, D. (2006). University language: A corpus-based study of spoken and written registers. Amsterdam: John Benjamins Publishing Company.

Slaughter, S. (1993). Retrenchment in the 1980s: The politics of prestige and gender. The Journal of Higher Education, 64(3), 250-282.

Behind the Scenes: Digital Media Responses to a Feminist Scholarly Article

In the first Behind the Scenes essay, I briefly discussed the alt-right response to my first publication, a feminist critical discourse analysis of STEM higher education syllabi. In this essay, I’ll discuss more about those responses, my critical analysis of them, and the resulting publication Digital Media Responses to a Feminist Scholarly Article: A Critical Discourse Analysis published in Feminist Media Studies. Doing the literature review, analysis, and writing of this manuscript was perhaps the most personal work I have ever done – but in saying that, I also hoped that by doing the work and publishing what I learned, my findings could help others also under attack from the alt-right world to understand the nature of the attacks against them, see how those responses were to a recontextualization of their work, not the original publication itself, and that the linguistic tools used in the responses were designed to silence or provoke not to begin an authentic discourse.

While findings are discussed at length in the full manuscript, in sum, in exploring 31 documents (13 twitter conversations, 3 discussion boards, and 15 articles), analysis suggested that the linguistic tools of stance and deixis were used across the corpus to create and reinforce an “us versus them” argument where I was the enemy. Second, analysis revealed how trolling strategies were used to provoke an emotional response and to silence me and those who might defend my article or do similar work. Finally, thematic analysis indicated that the linguistic tools and content reinforced the idea of knowledge as objective, the masculine nature of STEM education, and the appropriateness of a chilly STEM classroom climate.


This research refined what has become my approach to critical discourse analysis (CDA) research. I followed many of the same methodological steps as the Syllabi CDA research, and I approached the entire process systematically. Because I was analyzing digital media responses to my own work, it was especially important to seek external validity.  It was never my intent to theoretically distance myself from the process (or to pretend that I even could) – this was personal, and I was analyzing written attacks that were framed as personal attacks on me or my work. As such, I was not ready to begin that analysis immediately, so even though I recognized that I needed to save the emails I was receiving, I simply moved them upon receipt to an inbox folder I called “trolls.” For the most part, I stopped reading them completely.  I knew that I could return to them when I was ready to begin the analysis. It took almost a year and a half to return to the emails, articles, reddits, discussion boards, blog posts, and tweets I had been saving.  About a year after the original article had been published, I started searching the literature for what research was out there on similar attacks.

Seeking understanding of how others had analyzed and interpreted digital media responses to their work led me to Schmidt (2017) and Kamenetz (2018) and the theory of recontextualization.  As Attenborough (2014) explained, recontextualization, or when an event is described by another, leads to content that is responding to the recontextualization of the event even when it is being framed as a response to the original event. Understanding recontextualization was critical in understanding how so much vitriol and misunderstanding was being directed toward me when it was clear that most responses either hadn’t read or had just skimmed the original article. Those messages were in response to the recontextualizations of the article in the conservative media outlets, not the original article itself. At that time, the body of literature was still growing (although there is much being written about it now), so I expanded my search to include the research being done on trolls and trolling (trolls are the comments themselves, not the troller, a person, or the act of leaving trolls, referred to as trolling). This, in addition to the linguistic tools of stance and deixis, provided the additional linguistic tools I would code for in my analysis (See Hardaker, 2013 and Cole, 2015). Before I started analyzing data, I built a plan for analysis that built on prior research and utilized relevant linguistic tools, as discussed by Gee (2014). The CDA methodology process I follow for each study follows these steps:

  1. Develop the research questions
  2. Explore prior linguistic research on similar corpuses and genres
  3. Explore prior research on the research question and identify gaps or areas that need future research
  4. Create a plan for data analysis that builds on the body of research on this topic, both from a linguistic and thematic perspective.  Thematic analysis is always the final step in my CDA data analysis process.
  5. Identify the corpus (documents that will be analyzed) and seek ethical approval if required (most of my CDA work has been publicly available documents, but our IRB still requires a Request for Determination of Human Subjects submitted and a determination that it is not human subjects research before I can proceed with analysis and publication)
  6. Analyze data systematically, one analysis pass through the entire corpus for each linguistic tool, concluding with thematic analysis which will take as many analysis passes as necessary for saturation

In addition to building a strong theoretical and methodological framework for my analysis before I began the data analysis process, I also sought validity through the analysis and writing process, most importantly through critical friends. One particularly important friend provided critical feedback throughout the process, calling my attention to areas I had missed and also prompting me to consider which critiques might be valid. She read and re-read several drafts of the manuscript, helping me think critically about how my own bias and emotional responses to the comments might blind me to themes and implications resulting from the analysis.

Although the emails I received were often the most terrible – threatening harm and calling me terrible names, they were not public, and so I chose not to include them in the analysis. Ethically, I felt like that was the right choice, and the emails were simply more aggressive versions of the same tactics used by the more public digital media responses.  It is lucky, perhaps, that the original manuscript had contact information for me from an institution that was quick to shut down my email address after I graduated.  After 2016, for the most part, I stopped receiving the emails except for the ones from those especially upset individuals who hunted my University of Louisville email address.

Publication Process

One way to identify a journal for publication is to look at the journals where the work I am citing have been published.  One that recurred in my literature review was Feminist Media Studies, so while it is not a higher education journal, as I read about the mission of the journal, it seemed like the best option and had published similar work in the past. One time-consuming part of the publication process was transitioning to Chicago Style. While Chicago isn’t a difficult publication style on its own, it did take some time to transition the manuscript that I’d written in APA to a new format.  I also had to learn endnotes, and instead of doing them the right way (e.g., letting Word organize them for me), I painstakingly did them by hand. That meant that any time I needed to add a reference to, say, the methods section, I had to revise the entire reference list and each superscript throughout the entire manuscript. That was a poor choice. I was being lazy, and it ended up adding way more work to what should have been a simple process. Luckily, the manuscript was accepted without revisions (aside from the editorial edits and proofs), so I was saved from more reviewer-suggested additions to my literature review. Otherwise, working with Feminist Media Studies was delightful, and the publication process was fast.

After Publication

I felt like this work was important, so I submitted it to the 2019 NASPA Ruth Strang Research Award which, “recognizes individuals for outstanding contributions to the body of literature concerning women in higher education.” Despite feeling like the article was an important contribution, I didn’t expect it to win, so it was a wonderful surprise when I received the notification that I had won!

Finally, I have never responded to a single trolling email, troll, or article aside from this manuscript. I’m not suggesting that this should be your approach, but once I understood that despite any messaging to the contrary, the trolls, tweets, and emails were not an attempt to open up the discourse but to provoke a reaction, I felt like it was not worth my emotional energy and time to respond.  That being said, I also acknowledge that there is value is others seeing a response to trollers and trolls – indeed, it literally makes my day every time I see Roxane Gay clap back.  Responding to them individually is not the path I have chosen for now – and it is your choice whether or not you respond.  Either way, if you choose to respond, know what you are responding to with the knowledge that enlightenment or even an apology are unlikely to occur.


Attenborough, Frederick. 2014. “Jokes, Pranks, Blondes and Banter: Recontextualising Sexism in the British Print Press.” Journal of Gender Studies 23 (2): 137154. doi:10.1080/09589236.2013.774269.

Cole, Kirsti K. 2015. “‘It’s like She’s Eager to Be Verbally Abused’: Twitter, Trolls, and (En)Gendering Disciplinary Rhetoric.” Feminist Media Studies 15 (2): 356358. doi:10.1080/14680777.2015.1008750.

Gee, John P. 2014. An Introduction to Discourse Analysis: Theory and Method. New York: Routledge.

Hardaker, Claire. 2013. ““Uh…Not to Be Nitpicky, but…the past Tense of Drag is Dragged, Not Drug’.: an Overview of Trolling Strategies.” Journal of Language Aggression and Conflict 1 (1): 5886.10.1075/jlac

Kamenetz, Anya. 2018. “Professors Are Targets in Online Culture Wars; Some Right Back. nprED, April.

Behind the scenes: Are STEM syllabi gendered? A feminist critical discourse analysis

In the “Behind the Scenes” essay series, I dive into the background of my own published scholarly works to provide more insight into the background of the piece, the publication process, and, most importantly, an in-depth discussion about the methods of data collection and analysis that led to the findings described in the publication. Through this series, my goal is to provide insight into the unseen work that goes into the publication of higher education research. 

Are STEM Syllabi Gendered? A Feminist Critical Discourse Analysis was my first peer-reviewed publication. As a research project, it was formative in the development of my approach to Critical Discourse Analysis (CDA) as a research method.  As an open-access publication with a title that had words like “gendered” and “feminist,” it inspired vitriol from the alt-right interwebs that was unexpected, shocking, and largely unrelated to the actual content of the manuscript itself. Yet, because of those responses, the manuscript is still one of TQR’s most popular papers in terms of number of downloads, the paper won the 2015 Sandra Donaldson Award in Women and Gender Studies, and a subsequent CDA of responding articles, discussion boards, and tweets led to another publication. In retrospect, almost four years later, the negative attention had an important, formative impact on my development as a scholar, and reinforced my motivation to do critical work. But at the time, it was isolating, shocking, and it almost succeeded in silencing me.

The research reported on in the manuscript began in my discourse analysis class, taught by Marcus Weaver-Hightower, PhD in the final semester of my doctoral coursework at the University of North Dakota. When I was applying to doctoral programs, I attended an information session about the doctoral program at the University of Utah; in that information session, a graduate student panel recommended that doctoral students try to leave every methods course with research and a manuscript that could lead to publication, and I took that advice to heart (not recognizing, at the time, the pressure that would put on me, but also, more importantly, the pressure and added work this puts on instructors of introductory research methods courses).  I approached every methods course I took in my doctoral program with that intention, and left all but one with research that (much much later in some cases) was published.

In this study, I explored publicly available STEM syllabi for language use that perpetuated a chilly climate, specifically seeking to understand if and how STEM syllabi were gendered. This feminist CDA was my first experience with discourse analysis, I completed the study before I started working on my dissertation proposal.  However, this was not my first research project; my master’s degree required a thesis, and I’d worked as a research graduate assistant for two other research projects.  Altogether, I was still a young scholar and new researcher with so much to learn – I was just starting to problemmatize the connections between discourse, knowledge, and power. Through this work, I observed how institutional language replicated and reinforced power and oppression without using language that was overtly or obviously gendered. These findings informed my dissertation research and introduced me to Acker’s (1990; 2000; 2012) theory of gendered organizations and Biber’s (2006) book exploring the linguistics of Higher Education language, University Language.


Perhaps because this began as a project for an introduction to discourse analysis course, I started my critical discourse analysis as a traditional discourse analysis. I began with a focus on language-in-use and conducted a linguistic analysis of the language used in the syllabi. I started with a coding process similar to the sentence diagramming I did in elementary school, focusing on identifying parts of speech like pronouns and modal verbs. Informed by Gee’s (2014) approach to discourse analysis and Biber’s (2006) exploration of institutional language, I identified pronouns and modal verbs as important indicators of how power was negotiated and communicated in the STEM classroom. To make this methodological decision, I explored the literature to identify what parts of speech and linguistic tools might be evidence of power in written academic documents; that search suggested that pronouns, modal verbs, and interdiscursivity would be key linguistic tools to inform understanding of how gender and power were reified in language. While I didn’t expect to find this, I also planned to look for and highlight any obviously gendered language, such as referring to scientists as men or using he/she/ze.

The second component of my analysis sought to identify the discourses that were evidence of gendered power at the STEM institutional level. Discourses can be hard to identify, especially because discourses seem, by their very nature, natural and “logical.”  Strauss & Feiz (2014) defined discourse as, “the social and cognitive process that reflects, creates, shapes, re-creates, and reifies meaning in the lifeworld” (p. 1). My favorite definition of discourse, particularly relevant in CDA work is from Johnstone (2008; as quoted in Strauss & Feiz, 2014): “‘Discourses are ideas as well as ways of talking that influence and are influenced by the ideas'” (p. 2). Dominant discourses are, in my experience teaching discourse analysis, especially hard to identify by members of privileged groups because they are thought of as natural, normal, and “logical.” Dominant discourses convey a natural order and define what is right/wrong, good/bad, and true/untrue for everyone, selectively choosing what knowledge is acknowledged as valid, how knowledge is evaluated, and who is viewed as a holder of knowledge (or authority). It is, perhaps, easier to identify those discourses from a position of oppression because one is able to identify the processes, procedures, systems, structures, and discourses that create challenges for them – this is why I frequently use standpoint theory as a theoretical framework for my work. In this discourse, since I was exploring gendered discourses, I sought to identify what ideas about teaching (e.g., good students learn on their own), ideas about science/math (e.g., every question will have one right answer), and what knowledge (e.g., the egg and the sperm are, respectively, passive and active) was presented as correct, obvious, and logical that was, in fact, gendered or based on a gendered ideal scientist or student (read more about the findings in the full manuscript).

This approach to critical discourse analysis has informed all subsequent CDA projects. After developing research questions, I revisit linguistic tools to identify which tools might inform understanding of my research questions. Second, I explore existing CDA and DAs within the genre (or in genres that are similar) to see what linguistic tools have been used to explore this genre. Finally, I explore what discourses have been identified in this genre or field of study to inform my thematic analysis for discourses.  Data analysis, then, proceeds in iterative passes where I only look for one linguistic tool at a time – trying to look for all of them at once is overwhelming, and I’ve found that miss things if I try to look for everything at once.  If each linguistic tool I look for is a separate complete pass through the data, then by the time I start coding for discourses, I’ve already been through the entire corpus at least three times.  I have found this familiarity with the corpus (and the memos I take on the initial passes where discourses jump out) means that I’m better able to identify discourses having been subconsciously thinking about them throughout the analysis process. Further, analysis of linguistic choices is also an important part of understanding how power is reified in language.  Simply choosing to say “we” when “I” is meant is a choice that communicates shared responsibility or decision-making when the reality is quite different, and often those choices are often made in ways that are evidence of a discourse.

Publication Process

Having heard about the perils and challenges of the publication process, I was shocked when this manuscript was accepted without revisions. This has only happened a few times since, and it felt like it couldn’t be real when it happened for this manuscript.  I had expected rounds and rounds of revisions. I knew this was unique, but because I am plagued by imposter syndrome, I felt sure that the journal had made a mistake. Perhaps they’d confused my manuscript with another? Maybe I misunderstand the acceptance email and revisions were still coming?  Nevertheless, it was all real, and from my flat in Budapest, I celebrated my first peer-reviewed publication.

Note: Although this was my first peer-reviewed publication, this was not my first time submitting a manuscript for publication.  Another manuscript that didn’t end up being accepted for another year had already been through three rounds of revisions for a journal that later rejected it at this time this manuscript was accepted.

After Publication

Soon after publication, however, I would wake up to an email inbox full of hateful messages from strangers. Some were simple – they called me gendered and racist names and threatened violence and future harassment. Other emails were long, and critiqued the manuscript, saying my findings were wrong, saying I was clearly not a scientist, not a research, and stating that I wasn’t smart enough to get a PhD.  The first set of emails were scary but easy to ignore, especially living in Budapest, far from the English-speaking alt-right world that was criticizing me.  The second set, however, were harder to ignore.  I found myself wondering if the research was bad science, a poorly conceived manuscript, deeply flawed. Not knowing who to talk to and feeling isolated because I was so far away, at first, I didn’t tell anyone about it.  I think I was afraid that someone would read these emails (and tweets, reddits, discussion boards, and articles) and tell me that the writer was correct.  Eventually, however, I did tell my father, and I remember his response clearly – “well, if you are going to do this kind of work, this is the response you are going to get from this part of the world. It means you hit a nerve and are on the right track.”

I realized, in that discussion, that I had a choice to make – if this was going to be a response I received to my work, then I either 1) accepted this and continued with the knowledge that this would likely happen again – or 2) I could choose a different path and research something else, something safer, perhaps, with less vocal opponents. From there, it was an easy choice.  I wanted to do critical work, research that informed understanding of how marginalization was institutionalized into our discourses, structures, practices, and procedures – I wanted to, hopefully, help to make a difference.  I also realized, around this time, that I needed to treat these written responses as data, shifting my lens from viewing them as personal attacks to viewing them as data.  The latter, especially, helped me to shift my focus, and each new email or article published meant I had more data to analyze. Viewing it as a researcher also began to help me to really read the content of the attacks instead of internalizing the inherent messages in them. Much later, one of my students told me that I needed to be the “GI Jane of feminist Higher Education Research.”  He was, I think, being quite generous, as there are many more qualified individuals who better embody that title, but thinking of myself in that way has become a part of my academic identity.

Questions? What else would you like to know about the research and publication process of this manuscript? What is the story of your first publication?


Acker, J. (1990). Hierarchies, jobs, bodies: A theory of gendered organizations. Gender & Society, 4(2), 139-158. doi:10.1177/089124390004002002

Acker, J. (2000). Gendered contradictions in organizational equity projects. Organization, 7(4), 625-632.

Acker, J. (2012). Gendered organizations and intersectionality: Problems and possibilities. Equality, Diversity and Inclusion: An International Journal, 31(3), 214-224. doi:10.1108/02610151211209072

Biber, D. (2006). University Language: A corpus-based study of spoken and written registers. Philadelphia, PA: John Benjamin Publishing Company.

Gee, J. P. (2014). An introduction to discourse analysis: Theory and method. New York, NY: Routledge.

Strauss, S., & Feiz, P. (2014). Discourse analysis: Putting our worlds into words. New York,

NY: Routledge.