Malcolm Gladwells Blink (2005) is a very smart book which presents a serious problem for democrats who believe in accountability and for those who, more simply, believe that George W Bush makes bad decisions.
The Tipping Point (2000) made the authors name with its one central idea: that there is a magic moment when ideas, trends and social behaviours cross a threshold, tip and spread like wildfire. Blink too has a single focus: what normal people call intuition our ability, without thinking, to reach the right conclusion by using our instinct, or gut feeling.
A great strength of the book is that it draws on the literature of experimental psychology and related disciplines to make intuition respectable. We all know that intuition exists and works very well. But we also know (or think we know) that it is not respectable as compared with the gold standard of proper deciding the use of rational analysis.
We are meant, surely, to arrive at the best answer by conscious review of the facts and the alternatives, in the light of our objectives and values. That process will typically take time, be conscious, explicit, transparent, accountable and auditable you can see how the path from argument to conclusion works.
But in the paradigm case, intuition has none of those characteristics you just know what the right answer is. It does not emerge from any auditable process often, you cant really say why you decided that way, or if you do, it is just after-the-fact rationalisation, the stuff that people who like analysis need to hear, not an account of how you came to decide this rather than that.
Picking up and popularising the work of neurologists and psychologists, such as Antonio Damasio, Gary Klein and Gerd Gigerenzer, Gladwell does an excellent job to show that intuition is not just what you have to use when you havent got the time or the facts or the intelligence to use the real thing, but a powerful and distinct faculty in its own right which should not be assessed by comparing it with analysis any more than a saw should be assessed by comparing it with a hammer.
As you would expect from a staff writer on the New Yorker, Gladwells popularisation is compelling. Appropriately, the power of the book lies not in its analysis of deciding, but on the relentless accumulation of anecdote and example typically from respectable experimental sources to buttress his championing of intuition.
Intuition and decision
But beyond that, Blink is a disappointment. By putting all the phenomena illustrated by his anecdotes under the heading blink (or by calling them as I do here intuition) he illicitly conflates a quite heterogeneous group of non-analytic decision methods from short cuts to internalised skills to Gigerenzers fast and frugal heuristics to Damasios somatic markers. But even if for the moment this portmanteau term is good enough, there then remain two serious loose ends.
First, Gladwell does not deal at all well with the hard part of the intuition story the so what? question. If you accept that intuition matters, what do you do next? One answer to this might be to provide a self-help manual which tells you how and when to use intuition. After all, the management literature is full of handbooks telling you how to decide analytically; Gladwell has a lot of fun with the dogged attempts of the United States department of defense to provide a systematic, step-by-step, manual for deciding what to do if, for example, the US were faced with a rogue dictator in the Persian Gulf.
But the (maybe self-contradictory) question remains what are the rules for using intuition well? It is a real question because intuitive answers do not carry on their faces a guarantee of quality far from it: for every good snap judgment there will be another when we just get it wrong.
The most obvious example is prejudice. A Metropolitan Police (London) poster which showed a black kid in trainers running away from a startled white man in collar and tie traded on the correct guess that most people viewing it would see the kid as the mugger and the man as the victim. In fact, the young man was a plain-clothes policeman running after the unseen thief, and the point of the poster was to make us realise that it is hard to see black kids as policemen. And intuition is all about drawing instinctively on a library of previously experienced patterns. So we want an answer to the question how do we use it when? Gladwell has no good answer.
Second, Gladwells book misses even more awkward issues which centre on our appetite for transparency and accountability in public and political life.
One of the reasons that we like analysis is that it is explicit. If we ask an expert a lawyer or a scientist to analyse a problem for us, we can see clearly why he arrived at that conclusion and hence why, if we followed his advice, we decided as we did. If we or he are challenged later, we can give an account of ourselves a transparent account which allows the shareholders, or the electorate, to see what the reasons were, and see where it went wrong, if it did: there werent any weapons of mass destruction (WMD), despite the intelligence: the movie didnt succeed, despite the market research.
Another reason that we like this model is its intimate link with democracy. When the king decided, you couldnt ask him why he had decided that way: there was nothing to make him answer. Getting him to answer was part of the anti-authoritarian revolution, and people in authority had then to be good not only at deciding, but also at explaining their decisions. And the two dont necessarily go together.
Intuition and accountability
Bringing together our need for transparency and the reality of the role of intuition in deciding leaves us with three hard problems.
The first is this. Supposing we accept that often intuition is indeed a very good way of deciding. There is then a serious problem about accountability. If the defining characteristics of intuition are that it operates unconsciously and you typically cannot say how you decided as you did, the only truthful answer I can give you when you ask me why I decided that way is the one I cant say: trust me.
In our ordinary life we all know people like that the friend whose judgment we respect who just says do that, and we do because she is just good at that sort of thing, even if she is not very good at giving reasons and we use such people all the time because we know that they have good instincts about situations. Unfortunately, the trust me answer is what the reasonably well-mannered but authoritarian king would have given.
Second, and even worse, the accountability model depends heavily on the idea that if I am given an account of why the decision was taken that way, I can assess the account, I can audit it. But often I cant. Take the discouraging example of personal finance. Anyone who takes out a mortgage or decides on a pension plan knows what it feels like when the advisor advises. You dont know how to assess the advice because you arent any good at that sort of problem. So when she gives you reasons for the advice, you dont know what to do. You often just have to trust her advice, which is not the same as checking it and finding it sound. And that trust in turn cannot be based on an analytic assessment of the advisors competence. To do that you have to be able to understand what a good advisor is like which requires you to be skilled in the field which you are not.
So even when the problem is capable of rigorous analysis, as pension problems generally are, and could in principle be decided on that basis, decisions or advice cannot be picked apart through the accountability process, unless the public to which the account is to be given is expert too. But the reason for using an expert is precisely that you are not expert. So you cant audit her advice. So you have to trust her or not.
Third, and most seriously, there are many cases when an account cannot be given in analytic terms because no satisfactory analysis can be given. Take the BSE, (later vCJD) crisis in Britain in the 1980s. When BSE (Bovine spongiform encephalopath, mad cow disease) broke out in 1986, the key question was whether it was safe to eat beef. This was put to a commission of experts, headed by Richard Southwood, who delivered the answer that for adults at any rate it was indeed safe.
This answer was unambiguous. If you look at the basis of that conclusion, it relied on a paucity of evidence and guesswork (Southwoods words). But an answer had to be found. In the end he would have had to say that although his answer was unambiguous, it was not water tight, it was no more than the best he could do. And if you ask why, in such uncertain circumstances, he concluded as he did, he would not have said that it followed analytically from the evidence. He would have said that he used his judgment. He could not say anything more about why it came out that. And the point of asking him despite all this was that he is an expert and we are not. He remains an expert and a better bet that you or I even though he cannot give us a proper account of why he decided as he did.
Intuition and democracy
This is equally true in relation to military strategy and foreign affairs. Richard Godwin asks in relation to the Vietnam war:
Why should we try to contain China? it is precisely because there is no sure and resistless logic that discussion often dissolves into empty generalities and false scholarship almost all any leader can do is to call upon wisdom and judgement and national principle, a sense of history and knowledge of present reality, and act on their speculative and intuitive guess that results.If thats how it is, the giving of reasons in a transparent and accountable framework comes to look very thin.
What sticks in the gullet of the democrat is the anti-intellectual, irrational, trust-the-expert (or the powerful), tone of all this. But we cannot dismiss the evident truth that we live in a world in which judgment and intuition are necessary and efficient tools for deciding: that we all know this from our everyday life and we have no way of handling this unpleasant truth when we come to holding our political leaders to account.
For those openDemocracy readers whose views on George W Bush range from unease to distrust via disapproval, the problem of judgment and intuition presents itself as follows. His is, in the words of Joe Klein, the Blink presidency (Time, 20 February 2005). Bush gives strong and decisive leadership via instantaneous and subconscious decision-making. He sort of explains what he is doing: but the words are much more about what he is like than about what process of ratiocination led to the conclusion.
People call Bush stupid. If that means that he does not conform with the ideal of rational analytic processes, fine. But what if he is just smart and is one of those smart people who cant explain himself very well, because he doesnt operate like that? People say he is authoritarian. True, if you cant give an account of yourself and your decisions in transparent terms, you arent playing your part in an important element of the democratic process. But again, thats what people who are like that are like. We all know and admire people like that. Why pick on Bush?
Take the decision of President Kennedy to outface Khrushchev over Cuba. If you read Aleksandr Fursenko and Timothy Naftalis One Hell of a Gamble, it is unsurprisingly clear that he conformed with Godwins model of speculative and intuitive guesses.
Winston Churchills decision not to surrender in 1940 wasnt the result of a careful analysis of alternative scenarios to 1947. In both these cases most people think of Kennedy and Churchill as good guys, approving their exercise of intuitive judgment, and accepting that those decisions had to be made like that. Would they approve of Bush and the neocons if after Iraq democracy spreads across the middle east? Certainly intuition can go badly wrong. So is the only accountability test how it turns out and forget what the leader is like, his principles and his character?
The unfortunate truth is that we have no way of fitting our democratic need for transparency and accountability with any realistic view of how decisions are and should be made.