Looking for a great summer read? If you like Nate Silver’s quantitative assessments of politics and sports, you will love Seth Stephens-Davidowitz’s new book on big data revelations about our human interests, traits, and behaviors. By drilling down through millions of data points, often from people’s anonymous Google searches, he offers insights into racial prejudice, sexual orientation, child abuse, and even the age at which people’s long-term sports loyalties crystallize.
With data science he can also test popular ideas. Was Freud right to suppose that phallic symbols in dreams, and innuendos in word slips, reveal our unconscious sexuality? Is the man who dreamed of eating a banana on his wedding day “secretly thinking of a penis”? Is typing “lipsdick” when you meant “lipstick” an eruption of your hidden desire?
In search of answers, Stephens-Davidowitz analyzed whether phallic-shaped foods “sneak into our dreams with unexpected frequency.” His answer: They do not. In dreams, bananas are the second most common fruit . . . and they also are the second most consumed fruit. Cucumbers are the seventh most dreamt vegetable, and the seventh most consumed vegetable.
In search of Freudian slips, he analyzed 40,000 typing errors collected by Microsoft. A few were sexually tinged—“sexurity” instead of “security,” and “cocks” instead of “rocks.” But then there also were innocent slips such as “pindows,” “fegetables,” and “aftermoons.” After analyzing the frequency of various errors in random typos, Stephens-Davidowitz concludes that “People make lots of mistakes.” And when you make enough, you can expect an occasional and statistically predictable miscue. Searching the quarter million e-mails I’ve received since 2000, for example, I see that friends have written me about their experiences with “Wisconsin Pubic Radio,” with hearing access in “pubic venues” and with “pubic access,” and in their work as a national organization’s “Director of Pubic Policy.”
Thus, “Freud’s theory that errors reveal our subconscious wants is indeed falsifiable—and, according to my analysis of the data, false.”