The Problem with "I Argue that..."

Catherine Cocks's picture

Guest post by Jenny Tan, Associate Editor, University of Pennsylvania Press

In the vast ocean of rules and opinions about academic writing, few issues seem to be as divisive as that of signposting. Ask someone whether it’s okay to say “I argue that…” in your writing, and you’ll get a range of responses, from those who insist that you must use this phrase in any article or book you write, to those who will tell you to kill it with fire wherever it appears.

Here’s my take on this debate: Setting stylistic preferences aside, signposting is effective when it allows the reader to do less work to follow the writer’s thinking and understand their argument. Signposting is not effective when the writer has used it to do less work, to take a shortcut through necessary developmental steps.

Detractors of the “I argue” constructions, for example, often say that if you have clearly stated your argument, prefacing it with "I argue" is unnecessary. This is probably true, but all the same, the phrase can be a helpful shortcut for the reader, like a second-hand book where someone has highlighted the most important bits. It says, "Pay attention! Here’s the main argument! Keep this in mind as you read!" It ensures that you're understood and helps the reader come away with the one big idea you want them to come away with more easily.


Doing the writerly work of thinking through how all your ideas relate to one another is necessary so that you can successfully recognize and formulate your one main argument.


If, however, "I argue" and other argument-signposting phrases proliferate throughout a piece of writing, the construction loses its utility, like a book where someone has highlighted an entire page—likewise, if what follows “I argue” isn’t actually the main argument or even an argument at all, like a book where someone has highlighted at random. This kind of signposting is ineffective because it serves as a shortcut for the writer, who is often trying to lead the reader along a path that they haven’t carved out yet, and thus ends up creating more work for the reader. Doing the writerly work of thinking through how all your ideas relate to one another is necessary so that you can successfully recognize and formulate your one main argument. (For more on how to do this work, see Laura Portwood-Stacer on “Getting into Arguments.”) You can put "I argue" in front of that one!

It can be difficult in practice for a writer immersed in writing and revising to maintain a clear perspective on what their signposting is doing. But there are some tells of ineffective signposting. A multitude of “arguments” is one. Another is the overuse of metalanguage to describe what your book or article is doing. Words like explore, examine, investigate, and analyze are helpful when used sparingly to narrow down your focus, but it’s important to keep in mind that they do not themselves convey an argument. 

“Shedding light on” something is not an argument, either, and is rarely an effective use of metalanguage: just tell us what we learn as a result of the light shed. Same with offering “insight” into something—tell us what the insight is! Verbs like problematize, nuance, and complicate especially suffer from misuse, because they promise a payoff that is often not realized or else is hard to discern. These words make a claim about your argument—that it is more sophisticated than other arguments or perspectives—but stop short of actually stating an argument.

The tendency to overuse this metalanguage strikes me as a response to the pressure that scholarly authors feel to not just make an argument, but an original, important one. And so, writers fall back on familiar techniques to signal that they’ve fulfilled this obligation. But the signaling can unintentionally disguise the fact that a clearly stated argument is missing, since claims about “the argument” (and what it contributes, what it builds upon, etc.) give the impression it has already been established in text—thus shifting the onus off the writer and onto the reader to discern where and what it is.

Similarly, signposting can inadvertently disguise a not-yet-fully developed structure: the logical sequence of analytical steps through which the argument unfolds for the reader. Where this often reveals itself most clearly is in roadmapping. The analogy of “roadmapping” (and “signposting,” too, for that matter) points to its own limitations as a writing technique; no amount of mapping will get the reader from point A to point B if there isn’t already a reasonably clear path there.


...an attempt at roadmapping often becomes a blow-by-blow account of the text [...] But this is an outline of tasks for the writer to execute, not of the conceptual moves used to develop an argument.


In this case, an attempt at roadmapping often becomes a blow-by-blow account of the text, e.g., “First I explain the theory I draw on, then I provide contextual information for my object of study, then I analyze that object, then I make some practical takeaways.” But this is an outline of tasks for the writer to execute, not of the conceptual moves used to develop an argument. The information it provides is not useful, especially when the roadmap states things that should be self-evident, such as, “I will analyze the poem that is the topic of this chapter” or “I will explain my argument.”

Effective roadmapping provides information that is useful even taken on its own, should the reader choose to quickly skim the rest of the text or stop reading at that point altogether. It exposes the connective tissue of your argument, showing not just the points on the map, but how those points are linked. The same logic applies to other forms of signposting, such as transitions and topic sentences (which are, in fact, the most important and indispensable elements of effective signposting). Rather than saying, “Having discussed X, I will now turn to Y,” a good transition explains why and how Y follows from X.

Those who hate signposting sometimes object to the way it makes the writer an explicit presence in the text, while those who favor it embrace that presence. But, while the technique of signposting is to reveal your authorial intentions and craft by talking about them, that is not the purpose. Good signposting, counterintuitively, requires thinking from the point of view not of the writer, but of the reader.

Jenny Tan is Associate Editor at the University of Pennsylvania Press, acquiring books in the social sciences, in particular anthropology and politics. She received her Ph.D. in Comparative Literature and Medieval Studies from UC Berkeley in 2019.


Have something to say on this topic? Reply to this post or email the Elephant about writing for us. We welcome submissions from stakeholders on all sides of scholarly publishing.