False
Score | 33
Prince Nigeria Student @ Babcock University
In Africa • 6 min read •
The Accent of Intelligence

The Accent of Intelligence


There is something interesting about being considered too intelligent to be believed.


Imagine spending hours constructing an argument, carefully selecting your words and rearranging sentences until they communicate exactly what you intend, only for someone to read your work and conclude that you could not possibly have written it. Not because your argument is weak, but because it is too good. Your grammar is too polished, your vocabulary too sophisticated and your sentences too organised.


Suddenly, your intelligence becomes evidence against you.


I encountered a version of this experience during a short engineering quiz at university. My lecturer questioned my answer, not because the final result was wrong, but because the method I used to arrive at it differed from the one he had taught us. He assumed I had used AI. What made the situation particularly ironic was that the method he found suspicious was actually presented in the very textbook he had recommended for the course. I showed him the textbook and the section I had referenced, and he eventually returned my marks.


It was a relatively small incident, but it left me thinking about how easily a difference in approach can be mistaken for dishonesty. I had not introduced an unfamiliar method from nowhere; I had consulted a recommended academic resource and applied what I had learnt. Yet, for a moment, my answer had to be defended simply because it did not resemble what was expected. If a student can be suspected of using AI for solving an engineering problem differently, what happens when an African writer expresses an idea differently?


For African writers, this suspicion is not entirely new. Artificial intelligence has simply given an old prejudice a new vocabulary.


Long before ChatGPT, African writers were already being told what their stories should look like and what their characters should sound like. In her TED Talk, The Danger of a Single Story, Chimamanda Ngozi Adichie recounts how a professor questioned the authenticity of her novel because its characters were educated, middle-class Nigerians who drove cars. They did not resemble his expectations of African characters.


What makes this particularly troubling is that the professor was not necessarily questioning the quality of her writing. He was questioning whether her reality qualified as African enough to be represented.


The problem, therefore, was not the story. It was the audience’s definition of authenticity.


Today, that same expectation has found another expression. A writer uses an unfamiliar proverb, an unusually sophisticated word or a sentence that flows differently from what a reader expects, and the accusation of AI follows. It is almost as though African intelligence must first demonstrate that it is human before it can be appreciated.


This concern is not merely hypothetical. In a 2023 study published in Patterns, researchers affiliated with Stanford University examined seven AI detection tools and found that they incorrectly classified an average of 61.3% of 91 TOEFL essays written by non-native English speakers as AI-generated. These were essays by Chinese writers, not African writers, and that distinction matters. However, African writers who work in English as a second or additional language have reason to be concerned about the same kind of linguistic bias. The study raises an important question: when detection tools mistake human writing for machine-generated content, how much confidence should we place in their verdicts?


The Aké Arts and Book Festival 2026 prompt also draws attention to Lola Shoneyin’s observation that African writers are sometimes wrongly accused of using AI because of differences in language and literary style. This reinforces the need to recognise that unfamiliar expressions are not necessarily artificial ones.


But who established the standard in the first place?


We have become so accustomed to associating certain forms of expression with intelligence that anything outside those boundaries becomes suspicious. We recognise intelligence when it arrives in an accent we understand, a vocabulary we are familiar with and a structure we have been taught to respect. When it arrives differently, we begin searching for an explanation.


This is where the idea of native intelligence becomes important to me.


Intelligence did not begin with formal education, universities or, certainly, artificial intelligence. Long before algorithms could generate paragraphs, African societies had systems of knowledge, reasoning and communication. Our elders preserved history through oral traditions. Proverbs communicated complex lessons in remarkably few words. Communities developed ways of understanding agriculture, medicine, astronomy, governance and human relationships through observation and generations of experience.


Chinua Achebe, in his 1994 interview with The Paris Review, recalled a proverb: “There is that great proverb that until the lions have their own historians, the history of the hunt will always glorify the hunter.”


Its power is not simply in its imagery. It contains an argument about representation, ownership and the consequences of allowing others to tell our stories for us.


That, to me, is native intelligence: knowledge shaped by experience, culture, language and the ability to interpret the world from our own perspectives.


However, I would be careful about treating the African voice as though it were one identifiable sound.


Africa is not a language, an accent or a writing style. A Nigerian writing in Pidgin, a Kenyan writing in Swahili, a Ghanaian writing in English and a Senegalese writing in French are not necessarily expressing the same cultural experience. Even within Nigeria, the voice of a writer from Enugu may carry different linguistic influences from that of a writer from Lagos or Kano.


Our languages influence how we think, construct arguments, use humour and communicate emotion. Pidgin, for instance, is often dismissed as informal or grammatically inferior, yet consider the expression, “You dey whine me?”


Its meaning goes beyond asking someone whether they are joking. Depending on the context, it can communicate disbelief, irritation and an almost immediate rejection of what someone has said. Translating it into formal English may preserve the basic meaning, but some of its personality disappears.


A proverb translated directly from an indigenous language may also appear grammatically unusual in English. That does not make it poorly constructed. It may simply carry a thought that was never originally designed to exist within English grammar.


Authenticity is not about sounding African. It is about having the freedom to sound like yourself, even when your expression does not fit somebody else’s definition of African.


Nevertheless, defending African expression should not mean dismissing the problem of AI-generated writing.


AI misuse is real. Some people submit generated work as their own, and this creates legitimate concerns about originality, intellectual effort and recognition. Writers who spend time researching, developing ideas and refining their work deserve to have that effort respected.


But we must distinguish between evidence and assumption.


A sophisticated sentence is not evidence of AI use. Neither is flawless grammar, an unfamiliar vocabulary or the effective use of proverbs. Polish proves nothing about the author. If we are going to question originality, we need more meaningful ways of doing so.


Writing communities should encourage transparency, examine the development of ideas and create opportunities for writers to explain and defend their work. We should be interested in how an argument was developed, not simply whether its final version sounds too polished to be human.


More importantly, we must avoid creating an environment where writers feel pressured to deliberately weaken their writing to appear authentic.


There is something perverse about a writer having to make their work less intelligent to convince others that it is genuinely theirs.


Ultimately, what makes a voice yours is not the complexity of your vocabulary or the absence of grammatical errors. It is the collection of experiences, observations, influences and ideas that shape what you choose to say and how you choose to say it.


AI can reproduce patterns of language. It can imitate styles, construct arguments and arrange words into convincing paragraphs. But the ability to produce convincing writing is not proof of who produced a particular piece of writing.


We should also be honest enough to acknowledge that human writing has never existed in isolation. We learn from books, teachers, conversations, cultures and other writers. Originality does not mean producing something that has never been influenced by anything else. It means contributing something of your own to the influences that have shaped you.


Perhaps the real question is not whether an African writer sounds like AI.


Perhaps it is why we are so quick to believe that intelligence must have come from somewhere else.


The African writer should not have to choose between being understood and being authentic, between being brilliant and being believed.


If a sentence is intelligent, let us examine its intelligence. If an argument is original, let us appreciate its originality. And if there is evidence of dishonesty, let us address that evidence rather than prosecute an entire style of expression.


The court should be interested in the truth, not the accent of the accused.


After all, if we spend so much time teaching machines to sound human, we should be careful not to start treating intelligent humans like machines.


References


1. Adichie, Chimamanda Ngozi. The Danger of a Single Story. TEDGlobal, 2009.

2. Achebe, Chinua. Interview by Jerome Brooks. The Art of Fiction No. 139. The Paris Review, Winter 1994.

3. Liang, Weixin, et al. “GPT Detectors Are Biased Against Non-Native English Writers.” Patterns, vol. 4, no. 7, 2023.

4. Aké Arts and Book Festival. Native Intelligence: African Voice on Trial – Who Gets to Decide What Sounds Real? 


2
views 3
comments icon on TwoCents 1 share

Other insights from Prince

Contributor Points

Points-to-Coupons

Referral Earning


Insights for you.
What is TwoCents? ×