It’s important for undergraduates not to sound like AI bots when they’re writing philosophy. While there are no foolproof ‘tells’, here are four things you should think about to improve your writing. It will make you sound more like somebody who thinks for themselves, and who thinks about what they write.
First, ask yourself, “could somebody reasonably disagree with this statement, or does it describe something so subjective that nobody could really disagree with it?”
Some phrases that sound meaningful, but on second thought nobody could disagree with them because they are not literally true and refer only to the author’s perspective.
‘X’s view aligns with Y’s view’ (views don’t literally line up, and whether two views align isn’t a thing you can reasonably argue about)
‘Fact X highlights the importance of Y’ (facts dont literally highlight things, the ‘highlighting’ is in the author’s mind and the reader can’t disagree)
‘Evidence X suggests that perhaps Y’ (evidence doesn’t suggest, people suggest, and who can argue with a suggestion?’)
‘Claim X plays a crucial role in theory Y’ (what makes the role crucial? what exactly is the role?)
A little bit of this is okay when transitioning from paragraph to paragraph or section to section — there is ‘transitional language’ in writing. But writing that heavily uses subjectivistic, vague language like this in the most important parts where it should be precise ends up saying little of substance and it sounds like ‘robotic’ filler.
Second, ask yourself, “am I personally present in this text, or could anybody have written this?”.
Philosophy allows you to use the first person (“I” “me”) in ways other academic disciplines discourage. Descartes starts us off this way: cogito ergo sum. In fact we feel offended when somebody says something that should be in the first person (“I argue that X is wrong”) but is phrased in the third person (“X is wrong”) unless we’re embedded within an argument where the author is working from premises. While philosophy is not primarily about personal experience, the philosopher (or philosophy student) is a living person and I expect them to have their own opinions, reflections, impressions, connections they notice, and so on. “Robotic” writing has no self. It is non-commital, or stays overly objective, or sounds almost scripted when it tries to talk about its own opinions. The best thing is simply to aim for authenticity in your thoughts and allow yourself some idiosyncratic ways of writing (let yourself have quirks).
Some especially annoying phrasing
‘Critics argue that X is false’ (which critics? cite them)
‘We clearly see that Y’ (do we?)
‘It is intuitive that X’ (to whom? note – most professional philosophers are guilty of this one at some point!)
Third, ask yourself, “am I overstating the case I’ve made, or overly hedging and minimizing my case, when I should be moderately confident?”
Robotic writing tends either to use forceful word choices and vocabulary that overstate what’s actually been proven in an argument, or else to be overly cautious and non-commital when something has been successfully shown or is widely accepted in the context. Aim instead to write straightforwardly so the meaning is plain. It is better to err on the side of using overly informal language than to use language that sounds presumptuous, like the language and vocabulary are covering up emptiness. Suppose for example that you give a well written argument for the conclusion that Aristotle’s ethics is correct. Compare:
“This compellingly demonstrates that Aristotle’s ethics is correct, underscoring the profound and enduring significance of his virtue-based framework ” (overstated, bad)
“While ethical theories are inherently complex and nuanced, Aristotle’s ethics presents a compelling perspective” (overly hedging, bad)
“It follows that anyone who accepts the premises I have defended in this paper must also accept that Aristotle’s ethics is correct” (appropriate, good)
Fourth, ask yourself, “am I using a term that carries with it presuppositions that are not part of this discourse and treating them as though they are shared?”
‘Framework’ is the best example. Within analytic philosophy, the term “Framework” is a very specific reference to Rudolf Carnap’s notion of Sprachrahmen, a linguistic system which gives meaning to questions of being and truth; for Carnap and other positivists, there are no questions of being or truth outside a Framework. For whatever reason, Gemini introduced the term to mean something like “an individual’s structured system of beliefs”, which wasn’t really a use before AI came around. But now if somebody within philosophy throws around the word “Framework”, they don’t realize they’re carrying with this term a whole philosophy (logical positivism) that they may not actually buy into (or, at least, they need to justify it if they do). Basically, use the technical vocabulary that is already used in a particular discourse (course readings, lectures, etc.) and don’t add in technical terms from somewhere else. Other common examples:
“substrate” in place of “substance” / “ontological ground”. “Substance” or “ontological ground” is the fundamentally existing thing in virtue of which other things (like people, cups, etc.) come to exist. “Substrate” is a term from science and technology, it isn’t used in philosophy normally.
“prove” meaning “give empirical evidence for” (wrong) rather than “construct a deductively valid argument” (correct). This predates AI, but AI has made it worse, because consumer models of AI can’t do basic logic very well and so they throw around logical vocabulary which has a formal meaning when really what they mean is something less.
“metaphysical” to mean “supernatural” (wrong) as opposed to “regarding the fundamental structure of reality as a whole” (correct). This also predates AI, but has gotten worse because of it.
“holistic” to mean “considering more than one point of view” (wrong) as opposed to a reference to Quine’s coherentist epistemology or theory of language (correct).
“paradigm” to mean “a way of viewing the world” (wrong) as opposed to a reference to Kuhn’s theory of changes in science over time (correct).
You shouldn’t really worry too much because undergraduates make this mistake all the time and have for centuries, it’s a matter of learning a specialized vocabulary and everybody is learning it. It is not a big deal here or there when people make mistakes. It becomes “robotic language” when somebody is using a lot of technical terminology that doesn’t actually fit the context and the reader feels like maybe they don’t know what they are saying at all!