Why is academic writing so bad?


June 1, 2018

According to the stereotype, academic writing is at turns dry, jargony, esoteric, discursive, self-conscious, inward-looking, and—worst of all—just plain incomprehensible. The purpose of writing is to communicate ideas clearly and concisely, but academic writing achieves the opposite. In short, academic writing is bad.

Every researcher knows there is some truth to this stereotype. Yet there are also plenty of exceptions. While you have doubtless suffered through bland, confusing papers, you can probably also think of colleagues whose writing you consistently admire. Why, then, is academic writing often so bad, and what distinguishes the good writing?

When I speak of “bad writing” I do not primarily mean bad prose. Although sloppy prose is tedious to read, and the most broken or convoluted prose can genuinely impede comprehension, poor prose is not enough to explain why researchers and laypeople alike find academic writing to be so unpleasant. Plenty of academics write decent prose but unreadable articles. Moreover, non-native English speakers are a large and growing constituency in academia and especially in STEM fields. If English is to be the lingua franca of the global research community, it simply isn’t fair to expect that all researchers write beautiful prose. Thankfully, while beauty in prose is commendable, it is not necessary for comprehension.

Neither is the problem poor organization. In my experience,1 research papers tend to follow standard organizational conventions, with clearly labeled sections and subsections. Most researchers understand that scholarly writing is not free writing or poetry and that readers should not need to conduct a detailed exegesis to determine the topic or thesis of a paper.

Prose (low-level structure) and organization (high-level structure) are the basic technical proficiencies of any writer. If these aren’t what’s wrong with academic writing, what is? I suggest that the problem is not technical at all. A writer can be highly technically skilled, yet still fail to achieve the prime directives of clarity and cohesion. In my view, the problem with academic writing lies in the attitude that researchers take to their writing and to their readers.

Professional narcissism

I recently had the pleasure to read Steven Pinker’s excellent guide to writing style, The Sense of Style (Pinker 2014). I recommend the whole book to any nonfiction writer, but here I wish to call attention to just one idea, which Pinker calls professional narcissism.2 Writers engage in professional narcissism when they privilege the idiosyncratic demands and traditions of their professional specialty over the needs of their broader audience. Professional narcissism is not restricted to the academy, even if it is particularly acute there. It manifests itself in a variety of ways.

The curse of knowledge

Pinker devotes a whole chapter of his book—the best one, in my opinion—to a phenomenon called the curse of knowledge: “a difficulty in imagining what it is like for someone else not to know something that you know” (Pinker 2014, 59). He argues that the curse of knowledge is the main cause of incomprehensible writing.

A typical researcher spend years, decades, even a whole lifetime, cultivating knowledge about abstract, possibly counterintuitive concepts. After a time, these concepts become so deeply embedded in the researcher’s thinking that they forget what it was once like to not know them. When they write, they rely heavily on jargon as a convenient shorthand for complex ideas. They leave implicit the familiar logical connections between ideas. There is inevitably a gap between the writer’s and reader’s implicit conceptual models. If the writer cannot fill it with enough explicit explanation, the reader gets lost. The curse of the knowledge is the tendency to underestimate the size of the knowledge gap.

The curse of knowledge is a largely unconscious form of professional narcissism. To overcome it, we must make a conscious effort to see through the eyes of people outside our immediate research communities. Cutting down unnecessary abstraction and jargon is a good start. Pinker offers many practical tips on how to do so without compromising rigor.

Fixation on academic history

While we occasionally hear about academic misconduct related to insufficient citation or outright plagiarism, I think the more common failing is not too little, but too much, citation. The importance of acknowledging your influences is a universal scholarly value. However, explaining previous work should never interfere with, or take the place of, explaining your own.

Not long ago, I tried to read an influential book about scientific realism. After much preamble the authors announce their main thesis:

Our metaphysic is a synthesis of constructive empiricism and scientific realism based on our broadly Peircean verificationism. We aim to show that a view defended over several years by one of us (Ladyman), ontic structural realism, just is such a synthesis. We will do this by incrementally adjusting constructive empiricism and standard scientific realism in the face of each other’s objections—at least, those objections we can endorse on the basis of the PNC [principal of naturalistic closure]—until their residues form a consistent mixture that is a form of ontic structural realism. (Ladyman, Ross, and Spurrett 2007, 67)

The authors then deliver the promised “dialectical narrative,” indirectly developing their own position by contrasting it with several opposing ones. Don’t do this. The statement of your thesis should be as free-standing and self-contained as possible. Only after you have clearly introduced your own ideas should you compare them with competing approaches. In this case, because I didn’t know the competing positions any better than the authors’, I couldn’t even figure out what ontic structural realism is! That’s too bad: I ought to have been a receptive reader because I share the authors’ skepticism about metaphysics divorced from science.

Fixation on academic history is a form of professional narcissism because no one outside of your closest colleagues cares about your subspecialty’s history, which is probably very boring. Your readers want to know what you think and why you think it. That is what you should tell them.


In some fields, academic writing tends to be defensive, self-conscious, even apologetic. No statement, however straightforward, escapes decoration with a list of conditions, provisos, caveats, and pedantic clarifications. The writer is always on guard against their intellectual enemies. Even caveats that should be obvious to a competent, charitable reader are explicitly noted, lest a cagey opponent score a “gotcha!” criticism.

Self-conscious writing is professionally narcissistic because it prioritizes a small group of academic disputants over a broader readership that is competent and motivated to learn, but is not initiated into (or interested in) the author’s inner circle. When the main argument is constantly interrupted by caveats and clarifications, the uninitiated reader must work harder to keep the major points straight and may even lose the thread entirely.


At the other extreme, academic writing can suffer from excessive hype.3 The problem is most acute in fashionable scientific fields receiving mainstream media attention and heavy investment from funding agencies or corporations. As researchers pile into the field in search of funding and citations, they must work ever harder to attract attention over their growing mass of competitors. They are motivated to overstate the novelty and importance of their contributions, argue glibly to avoid calling attention to their work’s limitations, and write clickbaity titles more befitting a blog post than a serious work of scholarship. Such behavior is self-reinforcing and gives rise to the vicious cycle of hype.

Although it opposes self-consciousness, hype is professionally narcissistic in much the same way: the writer’s interest in advancing their own career is privileged over the reader’s interest in information presented clearly, correctly, and fairly. Hype is among the most pernicious forms of professional narcissism. Nothing does greater violence to the bond between writer and reader than the reader’s suspicion of being swindled. Once the author’s intellectual integrity is at question, a cloud of doubt is cast upon the whole work that can never be completely dispelled. Hype is so corrosive because it undermines trust in scientists and, by extension, in science itself.

What can we do about it?

To summarize, I think that the major problem with academic writing is professional narcissism in all its guises. Now that I know what to look for, I cannot help but see it everywhere. And I’ll be the first to admit that my own writing has sometimes been burdened by professional narcissism. So what can we, the research community, do about it?

I see reasons for both optimism and pessimism. Sometimes there are fundamental mismatches between the incentives faced by writers and readers. Writers respond rationally to their incentives—and readers pay for it. The hype cycle seems to fit this mold. On the other hand, I know I’m hardly the only researcher who routinely finds himself annoyed by research papers and books and wants to do better in his own writing. It is here that I find grounds for optimism.

Perhaps you’d like to improve your writing but worry that deviating from disciplinary norms will hurt your publishing prospects, invite censure from your stodgier colleagues, or otherwise harm your career. Helen Sword’s Stylish Academic Writing is a powerful antidote to such thinking (Sword 2011). The book is both a conventional style guide and a systematic literary study of academic writing. Sword (and her intrepid research assistant) assembled a dataset of one thousand academic articles: twenty articles from each of five journals in ten academic disciplines.4 They also studied one hundred recently published style guides, making the book a kind of style guide meta-analysis. As a statistician, I’m delighted to see a little data introduced into an area usually reserved for the (often contradictory) opinions of writing authorities.

An important finding of the book is that most—but crucially not all—published academic writing conforms to disciplinary norms, regardless of whether those norms constitute good writing style. Remarkably, although norms differ greatly across disciplines, Sword consistently finds a small but nonnegligible minority of writers willing break from the mold. In her sample, “roughly 10 percent of the articles across the board diverged from any given disciplinary trend” (Sword 2011, 18). Sword interprets this observation as a cause for hope. So do I. It proves that academics can, and sometimes actually do, deviate from stifling conventions while still publishing in major journals. Writing badly is a matter of choice, not an imposition of nature. So let us all, as responsible researchers, make the choice to write just a little bit better.


Ladyman, James, Don Ross, and David Spurrett. 2007. Every Thing Must Go: Metaphysics Naturalized. Oxford University Press. DOI:10.1093/acprof:oso/9780199276196.001.0001.
Pinker, Steven. 2014. The Sense of Style: The Thinking Person’s Guide to Writing in the 21st Century. Penguin Books.
Sword, Helen. 2011. Stylish Academic Writing. Harvard University Press. DOI:10.4159/harvard.9780674065093.


  1. My comments are biased towards those fields where I have some acquaintance with the professional literature, namely mathematics, statistics, computer science, and philosophy.↩︎

  2. To some extent, I regret the term “professional narcissism” because I fear it trivializes the more severe forms of narcissism, which are such a scourge upon politics and civil society. Then again, the term is memorable and evokes the right idea.↩︎

  3. Does my complaint about hype not contradict my previous complaint about self-consciousness? Here there seems to be variation across fields. With the possible exception of lawyers, no group of people share a greater love for arguing with each other than philosophers. Consequently philosophers tend write defensively and self-consciously. Meanwhile, certain branches of computer science, particularly artificial intelligence and machine learning, perennially suffer from terrible extremes of hype, in the worst cases becoming indistinguishable from self-parody.↩︎

  4. The ten disciplines are medicine, evolutionary biology, computer science, higher education, psychology, anthropology, law, philosophy, history, and literary studies. See the appendix of Sword’s book for more information about selection criteria.↩︎