The Uses of Half-True Alarms2

22/07/2010 14:33



So Carr, alert as well as alarmed, confronts himself as well as his reader with the classic smoke-fire problem. His alarms come clanging on almost every page. What to make of them? They cannot be dismissed as the mutterings of an obsolescent graybeard—Carr is in his early forties. To his credit, moreover, he pauses to address some objections to his line of argument—for example, the striking, well-established finding that IQ scores almost everywhere have been rising for a century while the means of distraction have been multiplying exponentially. “If we’re so dumb,” he italicizes, “why do we keep getting smarter?” Office 2007 is so powerful.

But we don’t, Carr argues—at least not in any simple way. The notion that smartness comes in a single variety is too crude. The testing signals are actually mixed. Some skills have increased as computers spread, but “tests of memorization, vocabulary, general knowledge, and even basic arithmetic have shown little or no improvement.” Worse, there are some recent signs of loss. While math scores have held steady over the past decade, verbal scores have declined. Between 1992 and 2005, something called “literary reading aptitude” dropped 12 percent. A testing skeptic might doubt the worth of any such findings, but it does seem to be the case that something real is being measured, and that whatever it is, it is slipping. Clumsy statistics are not foolproof evidence, and neither are the dumbing-down anecdotes any reader can supply. But they are not nothing. 

Unfortunately Carr does not entertain the possibility of unexpected gifts from the internet. He does not ask whether associational thinking—thinking that leaps horizontally, connecting dots that previously were segregated or “siloed”— might actually benefit from the non-stop multitasking in which one’s center of consciousness is constantly intruded upon by fragments of periphery. Could it be that the great electronic torrent of bits, bytes, and buzz does not only turn all minds into short-term data dumps, but also might promote the creative discerning of patterns where none were evident before? This strikes me as an unanswerable question but not a worthless one, even though it can only be properly asked if one reverts to weasel-word qualifiers.