What Nassim Taleb’s recommendations reveal is a single, ferocious
preoccupation: the tyranny of rare, high-impact events over the smooth models
people prefer. The keystone is his mentor Benoit Mandelbrot’s The
Misbehavior of Markets, which Taleb calls "the deepest and most realistic
finance book ever published" — the fractal, wild-randomness view of markets
that everything else in his work extends. From there the list is a rigorous
apparatus for taking tails seriously: Per Bak’s How Nature Works
explains, through self-organised criticality, why avalanches and crashes are
intrinsic rather than anomalous; Modelling Extremal Events is the
technical "bible" for measuring them; and The Kelly Capital Growth
Investment Criterion is how you bet and survive when they are real.
The second thread is skeptical empiricism — a distrust of tidy
theory that isn’t grounded in practice. It runs through Gary Taubes’s Good
Calories, Bad Calories, which Taleb praises as the work of "a true
empiricist" against nutritional dogma, and John Gray’s Straw Dogs,
whose assault on human hubris about knowledge and progress Taleb folded into
Antifragile. Peter Bevelin’s Seeking Wisdom and Emanuel
Derman’s Models Behaving Badly extend it into decision-making and the
honest limits of models — books, tellingly, by practitioners rather than
academics.
The surprise, and the key to him, is the literature. The man who weaponised
probability names a 1940 novel — Buzzati’s The Tartar Steppe, about an
officer who spends his whole life at a frontier fort waiting for an attack that
may never come — as a personal emblem, and praises Graham Robb’s The
Discovery of France for how it upends a tidy national story. Taleb reads
fiction and history not as decoration but as the same subject by other means:
how humans live inside uncertainty, anchor to false expectations, and are
remade by the events they failed to imagine. His shelf is the argument of his
life — that the smooth, the average and the respectable are where the danger
hides.