too much information?

Roger Bohn and James Short estimate that the average American consumes 100,500 words of written text per day and 34 gigabytes of data (most of which comes in the form of moving images). Their report, entitled How Much Information?, describes their data sources and methods very briefly, but apparently they drew for the most part on commercial marketing data. Assuming they are even roughly correct, the implications seem fairly profound to me.

First, there is the question of quality, which the authors themselves raise in noting that Lincoln just took a few words and bytes for his Gettysburg Address. In 2004, the National Endowment for the Arts found “literary reading … in dramatic decline with fewer than half of American adults now reading literature.” But according to Bohn and Short, people are absorbing just about one novel’s worth of text every 24 hours–day in and day out. I think the net change is bad–you’re much better of with one book about Lincoln or one novel by Virginia Woolf than 10 times as much text in the form of error messages, repetitive ads, and weather forecasts. I would not assert than any novel is better than any short electronic message–after all, this is a blog, which I take seriously and am happy to write. But the ratio of trivial to profound information must be changing for the worse, and one reason for falling quality is surely the overwhelming increase in quantity, which is extremely difficult to resist.

Bohn and Short estimate that the time spent consuming information grew from 7.4 hours per person per day in 2000 to 11.8 hours in 2008 (double-counting any hours spent consuming two or more information sources simultaneously). They consider that a fairly modest rate of increase, as is the increase in the number of bytes consumed: 5.4 percent per year (compounded). This rate seems low to them because the number of bytes that our electronic devices can handle doubles every other year, according to Moore’s Law. Watching a movie on NetFlix involves enormously more bytes than playing PacMan on your computer in 1980.

Evidently, lots of data are delivered to us but not used. We are filtering, ignoring, and searching more, and absorbing less–both for better and worse. At the same time, I would argue (against Bohn and Short) that we are actually increasing our data consumption at an extraordinary rate. We are not doubling our internal processing capacity every other year, so we are organisms under considerably strain.

Finally, the shift from TV broadcasts and books to computer searches and games means an increase of “interactivity.” The authors write: “we estimate that a full third of our [words consumed] is now received interactively, and 55 percent of our [bytes consumed]. This is an overwhelming transformation, and it is not surprising if it causes some cognitive changes.” Working interactively with data might seem liberating and empowering, compared to receiving someone else’s prepared stream of text or images. But I am not so sure about that. To struggle with a novel, a historical narrative, or a book-length argument means having to suppress to a degree your own assumptions, preferences, and habits in favor of the author’s. This can be profoundly liberating. On the other hand, to read what your own search queries generate, or to communicate in short bursts of text with your own friends, may do nothing but reinforce the views you started with. Just because it is intractable, an old-fashioned book can be quite a bit more “interactive” than the latest technologies.