The Advent of the Algorithm by David Berlinski, Harcourt, $28,
ISBN 0151003386
“ONCE I saw the Crown Prince Rudolf hurrying away from court in his fiacre,
his coachman Bratfisch whistling, `Wo bleibt die alte Zeit?’ while
somewhere to the east of the city, in a country castle, his mistress waited,
sloe-eyed, peach-lipped, a single rose pinned to her auburn hair.”
Ugh! Multiply this passage a hundredfold and you’ll have an idea of how it
feels to wade through David Berlinski’s The Advent of the Algorithm.
What have narcissistic digressions and self-indulgent prose to do with
algorithms? I became so exasperated that I threw the book across the room.
Advertisement
Take another random dip. Suppose you’re interested (as I am) in understanding
the clever numbering procedure underlying Kurt Gödel’s incompleteness
theorem. Do you really want the explanation broken up by a six-page diversion
detailing the author’s meaningless encounter with the Roman Catholic
cardinal of Vienna?
The writer of a work of popular science treads a narrow path. Readers come to
such a book primarily for information, not style—at least not in the
literary sense of beautifully crafted turns of phrase and deep insights into the
human drama. So the literary flourishes that distinguish a belle-lettres novel
from an airport thriller are exactly the features that distract a reader’s
attention from the business of the book. But the wooden prose that characterises
academic tomes is also forbidden because it ensures the kind of sales figures
that make publishers wince.
What is needed to gladden the heart of publisher and reader alike is a
judicious blend of hard facts and theories, together with short, pithy
anecdotes, spicy metaphors and clever examples to illuminate the technical
points.
But Berlinski seems never to have learned the distinction between an
integrated, gripping approach and a flabby and unfocused one. His “popular”
books, ranging from a treatment of modern systems analysis to an account of
calculus, are unhappy blends of literary self-indulgence and watered-down
science.
The Advent of the Algorithm, while ostensibly addressing one of the
greatest intellectual developments of the 20th century, is, unfortunately, a
prime example of Berlinski’s style.
It has to be said that Berlinski does manage to convey what an algorithm is
—simply a set of rules for carrying out a computation, an idea probably as
old as humankind itself. And he rightly emphasises that it is only in the past
fifty years or so that we have really begun to understand what an algorithm is
in a formal, mathematical sense.
Moreover, the book covers all the “right suspects”: the logicians Gottlob
Frege, Bertrand Russell and George Boole, together with the mathematicians
Gödel, Alan Turing and Alonzo Church. The work of these men and their
contemporaries led to today’s computers, as well as to a deep understanding of
just how far one can go in characterising events as the working out of a set of
rules.
Berlinski struggles valiantly to present the ideas underlying
algorithms—set theory, formal logical systems, the lambda calculus,
Turing machines—in a way that an uninformed reader might be able to
understand and digest. But I can only wonder who the reader for a book like this
could possibly be.
