How we meet the future

Generalising is a natural mental shortcut where we use past experiences to make sense of new situations. For example, a child might assume all dogs are friendly based on their pet golden retriever. Overgeneralising, however, is making broad assumptions without sufficient evidence, like believing all XL Bully dogs (a breed recently banned in the UK) are dangerous after one incident. While overgeneralising can lead to harmful mistakes, generalising itself is not a bias but a useful cognitive tool. It helps us structure our experiences and focus on new information by ignoring irrelevant details. The article highlights that the problem lies not in generalising but in making unwarranted leaps beyond the evidence we have.

In 1930, philosopher Richard Braithwaite discovered notes by his late friend Frank Plumpton Ramsey, a genius in philosophy, mathematics, and economics who died at age 26. Ramsey proposed that believing a generalisation, like 'all humans are mortal,' is not about listing every fact but about having a disposition or habit to act accordingly. For instance, if you believe all humans are mortal, you avoid actions that could harm them. Ramsey rejected the idea that generalisations are infinite lists of facts, as earlier philosopher Ludwig Wittgenstein argued. Instead, he saw them as tools to 'meet the future' by guiding our actions in new situations without needing perfect accuracy. This view helps explain why people can disagree on generalisations even when agreeing on all known facts.

Ramsey distinguished between two types of generalisations: universal and generic. Universal generalisations admit no exceptions, like 'all cats are furry,' and require high precision. If even one hairless Sphynx cat is encountered, the belief must be revised. Generic generalisations, however, are more flexible. They describe what is generally true without specifying proportions, like 'cats are furry.' These beliefs tolerate exceptions and don’t need to be revised after encountering counterexamples. For example, knowing that only 1% of mosquitoes carry the West Nile virus doesn’t stop people from believing 'mosquitoes carry the West Nile virus.' Generic generalisations simplify the world by focusing on broad patterns rather than exact details, making them practical for everyday life.

Stereotypes are a form of generic generalisations applied to social groups, such as nationalities, genders, or ethnicities. Like other generalisations, stereotypes simplify experiences by associating certain traits with a group. For instance, assuming 'all Brits are rude' after one encounter is a stereotype. Stereotypes are useful because they help structure our understanding of the world, but they can also lead to harmful errors. They often rely on flawed or limited evidence, such as stereotypical portrayals in media, and may cause us to misinterpret ambiguous situations. While stereotypes can be morally wrong and perpetuate discrimination, the article argues that their irrationality stems from the evidence used, not the act of generalising itself.

The article explores whether stereotyping is inherently irrational. It notes that humans have limited cognitive resources, which is why mental shortcuts like generalising are essential. Psychologist Daniel Kahneman, a Nobel Prize winner in 2002, demonstrated this in his work on bounded rationality and heuristics. Heuristics are rules of thumb that help us make decisions under uncertainty but can lead to errors. For example, in a 1983 study, participants incorrectly judged a statement about Linda being a 'bank teller and feminist' as more probable than her being just a 'bank teller,' despite the former logically implying the latter. This shows how generalising can sometimes override precise reasoning, but it also highlights why such shortcuts are necessary for navigating a complex world.