Via Negativa

Definition

Via negativa is the principle that knowledge and improvement are more reliable when subtractive rather than additive: it is easier and more robust to know what is wrong, fragile, or false and remove it than to know what is right and add it. Applied to forecasting, Taleb argues the only valid form of long-range prediction is negative — identifying what is currently fragile and therefore will not survive — rather than imagining specific new things that will appear: "the way to do it rigorously... is to take away from the future, reduce from it, simply, things that do not belong to the coming times."

In the Book

Chapter 20, "Time and Fragility," builds via negativa around the idea that "time is the same as disorder" — whatever survives exposure to time's disorder must be robust, so identifying what is fragile lets you predict what will disappear even without knowing what will replace it. Taleb illustrates with his own evening routine — walking in shoes similar to a 5,300-year-old design, eating with Mesopotamian-invented silverware, drinking wine using millennia-old technology — arguing that predictions of the future by figures like Jules Verne or H. G. Wells failed because they added imagined new technologies rather than asking what of the present would survive. He notes that longer-term predictions of this negative kind are paradoxically more reliable than short-term ones, since given enough time, anything Black-Swan-prone is certain to eventually be swallowed, whereas ordinary (additive, specific) predictions degrade the further out they reach. Via negativa is the master principle of Book VI, which also uses it to argue for subtraction in medicine (removing harmful treatments) and in architecture, diet, and technology adoption.

Why It Matters

Via negativa offers a route to reliable action under uncertainty when positive prediction is unreliable: you often cannot know what will work, but you can identify what is broken, fragile, or unnecessary and remove it — a lower-error, more robust form of improvement. This applies wherever addition is the default instinct (more rules, more medication, more features, more predictions) and reframes progress as subtraction: strip out the fragile and harmful, and let what's left, tested by time, carry the system forward.