Asking the Right Questions in the Age of AI: Malta and the Boissevain Legacy
Back in 1968, Jeremy Boissevain, a Dutch anthropologist living in Malta, perspicaciously asked why the Maltese, generally, and already starting from a young age at school, “ask so few questions” (Boissevain, 1968/1990). To attempt to answer this question, he identified four structural culprits. First, a colonial heritage of passive dependency, which rested on a power dynamic whereby it was deemed imprudent to speak out, and any resistance was best expressed harmlessly and covertly through informal gossip at best. Second, a hierarchy of infallibility that punished dissent: bred out of a ‘command and control’ culture and economy whereby those in power are always right; and silence is a deliberate strategy of outward consent that avoids reprisal. Third, the social fishbowl of a small island society, where the low threshold of intimacy and the density of social networks make anonymity impossible and guarantees that anything said or done would be quickly traced and attributed to its messenger and exponent, for better or for worse. Fourth, a language barrier that silenced those least confident in the English language. He warned that “the key to planning is, after all, merely the ability to ask intelligent questions about the future.” Decades on, that warning has become existential.
In the age of AI, the stakes of not questioning — or of asking the wrong questions — have multiplied dramatically. AI systems produce fluent, confident-sounding answers at scale; and these answers keep getting better. For a culture like Malta’s, still shaped by habits of deference and conformity, this carries a profound risk: the ‘hierarchy of infallibility’ now has a technological face. If Maltese students once hesitated to challenge a teacher, a priest or a politician – and they still do – they may now accept a chatbot’s output with a similar blind faith and a lack of scrutiny. Moreover, that chatbot’s output is a function of prompts: how then to encourage the Maltese to develop a critical (meaning questioning) attitude to AI?
Knowing how to prompt an AI platform — what exactly to ask, how to probe its limits, how to verify its outputs, how to follow up on an answer — is rapidly becoming a core literacy requirement. It is these and similar skills that should now be included in any educational activity. Becoming active and participative citizens in a democracy today also means not being gullible to technology. Asking “what is the answer?” is not enough. One must also ask: “is this answer valid and reliable, and whose interests does it serve?” AI-generated ‘replies’ – like all replies – are also incomplete by definition: and so, another question to ask is: “what has been left out, and why”? What is the politics and economics of absence?
One could argue that Malta’s colonial inheritance may have bequeathed a tendency to treat foreign expertise as trustworthy and therefore unquestionable. AI, overwhelmingly built outside Malta, in English, reflecting non-Maltese priorities, is the latest iteration of that deferential dynamic. To be sure, our long exposure to bipartisan politics – embedded and amplified through radio, television, newspapers and social media – might suggest that the Maltese are not as naïve as one might think. They have been handling and living with competing truths for decades. Behind that ‘silence’ is a strategic game of survival and the pursuit of self-interest; and perhaps a sharp sense of judgment. However, being quiet and ‘shutting up’ alone is not savviness: in the generative AI universe, seeking knowledge is impossible without asking questions.
Boissevain’s challenge to educators to cultivate and groom a questioning frame of mind is therefore not merely still relevant. In the age of AI, it is urgent, and the quality of the question matters more than ever.
Educators at any level really need to step up to the challenge. Much of what happens in the classroom remains a throwback to the industrial revolution, where obedience and adherence to authority was paramount. Times have changed. There is no point in educators behaving as if AI does not exist. There is no point in educators churning out assignments and assessments that are likely not to be the original work of their students; and then they grumble, despair and play police in order to ‘call out’ the students for their misdemeanour. There is no learning in that. Calling out students, by the way, is becoming increasingly challenging as the tech
improves. Instead, how are educators to generate a skill set that encourages their students to: identify reliable sources of knowledge; engage in suitable trajectories of interrogation; and fact-check the results of their AI prompts? Badly needed are safe spaces and opportunities where students are encouraged to ask questions, and not just answer them, often in implicitly predetermined ways. This is what builds character and critical attitude. It also future-proofs students for tomorrow’s labour market and gives them the tools and skills to avoid the massive disruptions envisaged by the mainstreaming of AI.
Further Reading
Boissevain, J. (1990). Why do the Maltese ask so few questions? Education (Malta), 3(4), 17–19. (Original work published in Ferment, 1968). https://www.um.edu.mt/library/oar/bitstream/123456789/52787/1/Education3(4)A3.pd
