Nicola Grandi is one of Italy’s most renowned linguists, with an academic career spanning morphology and linguistic typology. A full professor at the University of Bologna, he has made significant contributions to the study of linguistic processes and their variations. With his book Grammar and Error: Natural Languages Between Rules, Violations and Exceptions, he invites us to look at linguistic errors in a whole new light.
REMIDA meets NICOLA GRANDI
F: In one of our recent features, we talked about how language evolves alongside the time and culture in which it’s spoken. We could almost picture language as a house in constant renovation. Among the many building blocks shaping this ongoing transformation, we find errors — the very topic of your latest book. What are linguistic errors, and what role do they play in natural languages?
N: Error is perhaps the most complex notion to address in technical linguistic terms, precisely because it’s where we find the biggest gap between public perception and academic understanding.
From a technical, linguistic point of view, errors don’t actually exist. Any linguistic production by a speaker is the result of what their competence suggests, shaped by the input they’ve received throughout life, and so every utterance is as legitimate as any other.
F: So errors don’t exist?
N: Not quite. The concept of error exists in the listener’s perception: when someone notices a gap between a linguistic production and their own internal reference standard, that’s when they get the sense that a mistake has been made. Saying that errors simply don’t exist would be reductive. It might be true in theory, but in everyday linguistic exchanges, people perceive errors as something very real.
F: So how can we determine which linguistic productions are “errors” and which aren’t?
N: We do have an official point of reference: standard Italian, the version you find in grammar books taught at school. But context matters even more when it comes to evaluating errors: a linguistic deviation from the standard might be perfectly acceptable in one context and completely unacceptable in another. Take, for instance, Jovanotti’s song Un ragazzo fortunato, where he sings: “perché non c’è niente che ho bisogno.” According to standard Italian grammar, that’s an error. But it certainly didn’t stop the song from being a success, and I doubt anyone waves their arms in protest when Jovanotti sings it live. That’s why it’s so hard to define how many errors exist, or which ones matter: some linguistic forms make our eyes widen in certain settings… and go completely unnoticed in others.
F: From a communication standpoint we’re living through an unprecedented moment in human history.
There’s never been a time when we’ve written and spoken as much as we do now. Every day, we produce enough written text to fill a million Libraries of Alexandria.
If context is so key to judging errors, how is that changing over time?
N: As with anything, we tend to get better at what we practise the most. Today, the most widely used writing form is messaging, unlike thirty years ago. The most visible symptom of this modern shift, a true product of our time, is the difficulty many students now have with punctuation, as shown in recent analysis of student writing. There's a clear and growing gap between what we’re theoretically taught, the standard, and what we actually do.
F: Language — and you know this well — is not only made up of rules and conventions. Language comes above all from those who speak it. At Remida, we’ve developed a framework called Linguistic Direction®, which is based on this very principle: putting people at the centre of communication. For us, this means understanding the audience we’re speaking to, knowing their background, their cultural references. And even in this case, the study of linguistic errors can reveal a lot. How can linguistic errors reflect social, regional or register-based variations?
N: We use people’s linguistic productions to judge them, to create a kind of social profile. In short, based on how someone speaks, we build an identity that we then use to evaluate them. This happens because there’s a systematic correlation, typical of complex societies like ours, between social and linguistic parameters. Rather than judging the production itself, we judge the stereotype attached to it.For example, a deviation from the standard in a particular geographical area may point to dialectal influence, whereas in another context it might reflect social class or educational background. Additionally, errors may reflect the use of an informal register in settings that would normally require formal language, signalling familiarity or intimacy between speakers. The study of linguistic errors not only helps us understand how people use language, but also gives us a window into the social and cultural dynamics that shape communication.
F: So linguistic errors have an impact on how identity is perceived and constructed, both individually and collectively?
N: Language is fundamental to identity and ideally everyone should feel fully comfortable speaking their own language. We need to lighten the ideological weight currently attached to many uses of our language: a weight that, in some cases, makes it almost impossible to approach the topic scientifically.
F: Speaking of science, let’s shift for a moment to a more technological subject. Large Language Models are increasingly used by people online, and they show how linguistics and artificial intelligence can come together to create powerful tools for language processing. But can humans really be removed from that marvellous system of conventions and errors we’ve shaped over millennia — language itself?
N: All AI systems that generate language do so by being trained on texts produced by us. As a result, the texts they generate reflect the specific traits of the input they’ve been exposed to, both in terms of linguistic structures and embedded stereotypes. Even when trained on vast corpora, machines remain excluded from the social dimension that defines human beings, the same dimension that shapes language through its rules and its errors. The reins of language creation are still very much safey in human hands.
F: That said, language is the result of ongoing transformations over time, and that includes errors.
We also know that machines don’t “make mistakes”, while humans most certainly do. So… could it be that error is the key to keeping the evolution of language firmly in human hands?
N: Absolutely, if we see linguistic error as a source of creativity and innovation. Error lies at the heart of human diversity, and for that reason, it’s one of our greatest strengths. In my book, I quote Frank Zappa: “Without deviation from the norm, there is no progress.” Linguistic errors aren’t flaws, they’re opportunities for language to evolve and grow. Recognising the value of error allows us to embrace the full complexity and beauty of human communication.
And you, what kind of mistakes do you make with your language? Tell us all about it at supernova@remidastudio.com.
