This weekend, as a kind of follow-up to Csikszentmihalyi's Flow, I read John Kay's Obliquity: Why Our Goals Are Best Achieved Indirectly. I found a copy of Kay's book last weekend ago at a wonderful bookstore in Ely, and its argument resonates with Csikszentmihalyi's claim that optimal experiences are not always the happiest or most pleasurable ones, and with the more technical claims of van Nimwegen et al that contextual help can degrade users' long-term acquisition of problem-solving skills. The book grew out of a 2004 Financial Times essay, which is still online and well worth reading. But so is the book.

Obliquity, Kay writes,

describes the process of achieving complex objectives indirectly. In general oblique approaches recognize that complex objectives tend to be imprecisely defined and contain many elements that are not necessarily or obviously compatible with each other, and that we learn about the nature of the objective and means of achieving them during a process of exploration and discovery. (3-4)

This is not to say that there aren't problems that can be solved using direct, quantitative, or computational means; but the big goals that people pursue for years, or that companies pursue for decades, are by their nature oblique.

Why does obliquity work?

The answer is a bit like the famous line by Winston Churchill (another practitioner of obliquity, Kay argues) about democracy: it's the worst system, except for all the alternatives. Rigid ideologies or seductively elegant quantitative models can't account for the richness and vagaries of human experience. Neither the world nor success one-dimensional.

Further, objective measures are less objective than we think. First, they don't eliminate human judgment (which itself relies on craft and experience): Kay notes that profit can be measured in several ways, and CEOs' talk of "maximizing shareholder value" or ROI sounds precise but can be massaged to make lots of different outcomes look good. Further, in what Kay calls Franklin's Gambit, models are easily used to justify whatever outcome you happen to prefer: as he recalls from his days as a consultant, "I have never seen an analysis that didn't define the businesses the executives liked as stars and the ones they disliked as dogs." (100)

Direct approaches assume "a capacity for prediction that we can never possess." (104) Obliquity, in contrast, recognizes that complex institutions consist of many players, many of whom are constantly reacting to each other's actions and assessing each other's motives; this dynamic is pretty unpredictable.

Finally, direct approaches create perverse, self-defeating incentives. In the book's most compelling chapter, on "the profit-seeking paradox," Kay argues that the singleminded pursuit of profit is much more likely to destroy a company than a more oblique approach. Innovation and product development are inherently messy, complex activities that wither in a profit-maximizing environment. Further, as Kay memorably says, "a corporate culture that extols greed is, in the end, unable to protect itself from its own employees:" (37) as Churchill might have put it, CEOs who choose profit over honor end up with neither. Oblique approaches that value pragmatism, constant reevaluation, and are inherently modest may seem fuzzier, but they are also more honest, and tend to encourage more loyal behavior.

For all its being a book meant for corporate executives– it's one of those books you expect to see in the business class sections of airplanes, along with Tom Friedman's latest book– Obliquity reminds me a lot of writers like Oakenshott, Berlin and Hayek– people whose distrust of technocracy, statist projects, and ideology make them heroes of today's small-c conservatives. (Indeed, Kay himself opens with a broad criticism of the ultra-rationalist, tear-it-up-and-start-over Year Zero modernism represented by people like Le Corbusier and Pol Pot, and to a lesser degree by whiz kids like McNamara and Rostow.)

So how does this connect to contemplative computing? It's yet another reminder of the need to take a broader, more modest and human view of human action. This matters for HCI because as the range of places where computers are being used expands, and as computers are used in a wider range of problems and activities, we need to recognize that the formerly narrow world of computer use is no more: the old world of computers living in workplaces, and being used to support tasks that could be made more efficient in reasonably straightforward ways (keeping better control of inventory, printing checks faster and with fewer mistakes), is being muddied. Learning how people succeed under circumstances of fuzziness, complexity, and obliquity– how we pursue goals indirectly, find happiness through apparently unhappy means, and constantly seek meaningful experiences and flow– is an important first step to describing the world in which contemplative computing has to function.

Now on to the contemplative practices and self-experimentation literature.