The sentence turns up on conference slides and in book epigraphs, usually offered as prophecy: a wealth of information creates a poverty of attention. Herbert Simon put it in print in 1971, two decades before the web. Read whole, and in the room where he said it, it turns out to be doing something narrower and considerably more useful than predicting the future.
The room is worth a moment. The talk was given in a lecture series and published in 1971 by the Johns Hopkins Press in Martin Greenberger’s volume Computers, Communications, and the Public Interest. Simon spoke as Richard King Mellon Professor of Computer Science and Psychology at Carnegie-Mellon. His discussants were Karl Deutsch of Harvard and Martin Shubik of Yale. The moderator was Emilio Daddario, chairman of the House Subcommittee on Science, Research, and Development. A congressman was in the chair, and Simon gave an engineering answer.
He builds to the famous line through rabbits. A rabbit-rich world, he observes, is a lettuce-poor world. The obverse of a population problem is a scarcity problem, and therefore a resource-allocation problem. Only then:
In an information-rich world, the wealth of information means a dearth of something else: a scarcity of whatever it is that information consumes. What information consumes is rather obvious: it consumes the attention of its recipients. Hence a wealth of information creates a poverty of attention and a need to allocate that attention efficiently among the overabundance of information sources that might consume it.
The usual excerpt stops a clause early, at “poverty of attention,” which turns an allocation argument into an epigram. Simon’s operative words are the two that routinely get cut: allocate, and efficiently. He is opening a resource problem in order to do arithmetic on it, which is a different genre from lament.
The arithmetic comes next, and it is rarely quoted. A scarce resource needs a unit. Simon takes up Shannon’s bit and discards it: the bit capacity of any device or person for receiving information depends entirely on how that information is encoded, so it is not invariant, and is therefore an unsuitable measure of the scarcity of attention. His replacement is blunter. Measure how much time the recipient spends. Human beings, he writes, are like contemporary computers in being essentially serial devices, able to attend to only one thing at a time.
I should say plainly what that does to this column. The FTSeffect puts time in the denominator, and for weeks I have argued that the denominator is where saturation is actually decided. The decision to denominate attention in time is Simon’s, in print by 1971, and his reason was unrelated to politics: people are serial.
From there he turns to design, which is what the talk is really about. A computer divides its own attention among four activities: listening, storing, thinking, and speaking. Whether it relieves an overload problem or compounds it depends on how it splits them. The principle:
An information-processing subsystem (a computer or new organization unit) will reduce the net demand on the rest of the organization’s attention only if it absorbs more information previously received by others than it produces—that is, if it listens and thinks more than it speaks.
And then the test, which is the quietly devastating sentence in the essay:
It is conventional to begin designing an IPS by considering the information it will supply. In an information-rich world, however, this is doing things backwards. The crucial question is how much information it will allow to be withheld from the attention of other parts of the system.
Simon scores an information system on subtraction: how much of the world it agrees to keep from you. Every platform in this book is scored on the opposite quantity, how much it delivered and how long you stayed, and is optimized against that number. Judged by that number those systems are working exactly as specified. By Simon’s test the specification was written backwards.
He priced the problem too, in a line dropped almost in passing: free or underpriced resources are always in desperately short supply. Attention has been free at the point of use for the entire period this book covers. Every term in the numerator of the equation is a method for spending somebody else’s unpriced resource.
Simon wrote a design brief. Read as prophecy it arrives fifty-five years early. Read as a specification it remains largely unbuilt. His name for the system he wanted was an information condenser, and I do not think the obstacle to building one has been technical.
Next week: the cases where it didn’t work.
— J.W.B.
Note 17 of 24