We often talk about probabilities casually – “probably,” “likely,” “almost certain” – but how accurately do these terms translate to real-world outcomes? The gap between everyday language and precise mathematical probability is surprisingly vast, and understanding this mismatch isn’t just an academic exercise. It affects how we interpret everything from dinner plans to existential threats like climate change.
From Ancient Rhetoric to Modern Misunderstanding
The ambiguity of probability isn’t new. Ancient Greeks already recognized the difference between what seemed likely (eikos ) and what was persuasive (pithanon ), noting that the latter doesn’t necessarily align with actual odds. This linguistic instability carried over into Roman rhetoric, where both concepts were lumped under probabile, the root of our modern word.
It wasn’t until the 17th century, with the rise of gambling and the Enlightenment, that mathematicians began developing a quantifiable approach to probability. Philosophers followed, attempting to map degrees of belief onto a spectrum. John Locke, in 1690, proposed ranking certainty based on consensus, personal experience, and secondhand testimony – a framework relevant even to legal principles today.
The Legal and Economic Pursuit of Clarity
The need for precise probability extended into law and economics. Jeremy Bentham, in the 19th century, lamented the “deplorably defective” language used to quantify evidence in court. He even proposed a 0-to-10 scale for ranking belief strength, but deemed it impractical due to subjective variation. A century later, John Maynard Keynes favored relational comparisons – focusing on whether one event was more or less probable than another, rather than assigning absolute numbers.
The CIA’s Solution: A Probability Dictionary
The breakthrough came unexpectedly from an unlikely source: the CIA. In 1964, Sherman Kent, an intelligence analyst, drafted a classified memo, “Words of Estimative Probability,” to standardize language in National Intelligence Estimates. Kent recognized the tension between “poets” (those relying on qualitative language) and “mathematicians” (those demanding hard numbers). His solution? Assigning specific probabilities to vague terms: “almost certain” became 93%, though he strangely left gaps in the 0-to-100 scale.
From Intelligence to Science: A Patchy Adoption
Kent’s framework influenced scientific disciplines, though imperfectly. Surveys show some overlap between his scheme and how healthcare professionals interpret terms like “likely,” but inconsistencies remain. The IPCC, for example, defines “very likely” as a 90-100% chance – a sobering reminder that we’ve likely already exceeded the 1.5°C warming threshold.
The Psychology of Framing: Why Negatives Fail
Despite these efforts, probability perception remains skewed. Research shows that framing matters: people perceive “unlikely” forecasts as less credible than equivalent “likely” statements. This cognitive bias mirrors how we react to life-or-death scenarios: most prefer a treatment that saves 200 lives over one that lets 400 die, even though the outcomes are identical.
In conclusion, communicating uncertainty requires precision. When hard numbers aren’t feasible, shared linguistic understanding is crucial. And when possible, framing probabilities positively can increase acceptance – even if the underlying truth remains unchanged.
