by Guy Claxton
Who can wait quietly while the mud settles?
Who can remain still until the moment of action?
Observers of the Tao do not seek fulfilment.
Not seeking fulfilment they are not swayed by the desire for change.
Empty yourself of everything.
Let the mind rest at peace.
The ten thousand things rise and fall
While the Self watches their return.
They grow and flourish and then return to the source.
Returning to the source is stillness, which is the way of nature.
CHAPTER 6
Thinking Too Much? Reason and Intuition as Antagonists and Allies
The men of experiment are like the ant; they only collect and use. The reasoners resemble spiders, who make cob-webs out of their own substance. But the bee takes a middle course; it gathers its material from the flowers of the garden and of the field, but transforms and digests it by a power of its own.
Francis Bacon
Is it possible to think too much? Though people who cannot get to sleep for churning over their problems might answer in the affirmative, conventional wisdom sometimes seems to suggest not. In the classroom, the consulting room or the boardroom, we may operate as if the more analytical we were, the better. Or even if we do not, we may assume that we ought to; that a detailed listing and weighing up of considerations, for example, represents some kind of ideal cognitive strategy to which our actual behaviour approximates. As Benjamin Franklin wrote to the British scientist Joseph Priestley:
My way [of tackling a difficult problem] is to divide half a sheet of paper by a line into two columns, writing over the one Pro, and over the other Con. Then during three or four days’ consideration, I put down under the different headings short hints of the different motives, that at different times occur to me, for or against each measure . . . I find at length where the balance lies; and if, after a day or two of further consideration, nothing new that is of importance occurs on either side, I come to a determination accordingly . . . When each [reason] is thus considered, separately and comparatively, and the whole lies before me, I think I can judge better, and am less likely to make a rash step.1
The rationale for such a modus operandi is, presumably, that by making our thoughts and motives explicit and orderly, we can evaluate and integrate them better, and thus make better decisions. Or, as an influential textbook on decision-making puts it: ‘The spirit of decision analysis is divide and conquer: decompose a complex problem into simpler problems, get your thinking straight in these simpler problems, paste these analyses together with a logical glue, and come out with a program for action for the complex problem.’2 Be as explicit, as articulate and as systematic as you can be, and you will be thinking in the way that generates the best decisions and solutions. Given the evidence that we have looked at so far, however, we might have cause to question this ubiquitous, commonsensical assumption. D-mode and the slower ways of knowing work together, but they can get out of balance, and lose coordination.
Jonathan Schooler from the Learning Research and Development Center at the University of Pittsburgh has conducted a number of studies over the last few years that demonstrate graphically how thinking can get in the way of a whole variety of mental functions including everyday memory and decision-making, as well as intuition and insight. These studies go to the heart of the relationship between d-mode and intuition. One concerns how people choose between several possibilities, as when we are deciding which of a range of different foods we prefer. Schooler gave his subjects five different brands of strawberry jam to taste, and asked them to rate them and say which they liked best. The jams had recently been the subject of a consumer report, and those used had been ranked 1st, 11th, 24th, 32nd and 44th by the experts. Some of the subjects were told that they would be asked to explain the reasons for their choices, and to think carefully about their reactions and preferences. The results showed that those subjects who were left to their own devices ranked the jams in a way that corresponded closely to the judgement of the experts, while those who had been instructed to analyse their reactions disagreed with the experts.
Obviously this need not be a problem: maybe thinking about your choices makes you more independent. You decide what is right for you, rather than following the herd. If thinking carefully means that your decisions are based more closely on your true values and preferences, you would expect the thoughtful subjects to be more satisfied with their choices, and for this satisfaction to be more long-lasting. Unfortunately the reverse is the case. In a parallel study, subjects were shown five art posters and invited to choose one to take home. Those who had deliberated most carefully turned out to be significantly less satisfied with their choice, a few weeks later, than those who had chosen ‘intuitively’. The deliberating jam tasters were more ‘individual’ than their intuitive colleagues, but they were worse, not better, at making choices that reflected what they really liked.
Neither of these choices, you could say, has a terribly important impact on people’s lives. But deciding which courses to take in college certainly does. Schooler investigated university students as they selected their second-year psychology courses. They were given full information about all the options, including comments and ratings from those students who had taken each course the previous year, and asked to say which courses they thought they would take. Again, some students were asked to reflect in detail on the information provided, and the criteria they were using to make their choices. As in the other studies, those who thought most carefully were less likely to opt for the courses recommended by their peers, and more likely to change their minds subsequently. Later, when it came to signing up, the choices of the ‘deliberating’ students tended to revert to those that conformed more with received opinion, and correspond with the choices of those who had chosen intuitively.
On the basis of these studies, the researchers argue that there are a number of potentially negative effects of encouraging people to be more reflective and explicit about their decisions. In choosing a picture, or a jam, or a course, there are many interwoven considerations to be taken into account, not all of which are (equally) verbalisable. When the decision is made in an intuitive way, these considerations are treated in a more integrated fashion, and those that are hard to articulate are given due weight – which actually may be considerable. However, when people are forced (or encouraged) to be analytical, the problem is deconstructed into those considerations that are more amenable to being put into words. Thus the way the predicament is represented to consciousness may be, to a greater or lesser extent, a distortion of the way it is represented tacitly, and decisions based on this skewed impression are therefore less satisfactory.
In particular, d-mode may exclude or downgrade those nonverbal considerations that are primarily sensory or affective. Analytic thinking therefore tends to overestimate cognitive factors, which may be more easily expressed, resulting in decisions which seem ‘sensible’, but which fail to take into account non-cognitive factors. Additionally, the more carefully one analyses the different alternatives, the more one finds that there are good and bad aspects to each, and the greater the consequent tendency for judgements to become more moderate, more similar and therefore less decisive. Hence the tendency to come to decisions that differ from those which the acknowledged experts would have advised, and to feel obscurely dissatisfied with the choice one has made. Anyone who has ever agonised over a choice while shopping, and then regretted the decision immediately they have got the item home, will be familiar with this phenomenon. It is the dislocation between conscious and unconscious decision-making that people are referring to when they say that they should have listened to their ‘heart’, or their ‘gut feeling’, or their intuition.
As usual the issue is not black and white. We might now suspect that where a problem can be adequately represented verbally, and where the solution lies at the end of a logical chain of reasoning, a predominantly d-mode app
roach will be effective and efficient. While where the problem is more complex, contains aspects that are hard to articulate, or demands an insightful leap, d-mode will be less successful than a more receptive, patient approach. In another series of experiments, Schooler and his colleagues explored these two types of problems, looking particularly to see where active thinking helped and where it hindered.
‘Insight problems’ are those where people are in possession of all the information and ability necessary to solve them, but where there is a tendency to feel blocked or ‘stumped’, before suddenly having a kind of ‘Aha!’ experience in which the solution becomes immediately or rapidly obvious. In such problems, the difficulty is often caused by the tendency to make some unconscious assumptions that get in the way, or to fail to retrieve knowledge that would actually be helpful. We have met examples of such problems before – the ‘mutilated chessboard’, for example. Two different puzzles of this kind are shown in Figure 6.
The figure on the left represents a triangle made up of coins. The problem is to make the triangle point downwards by moving only three of the coins. The figure on the right represents a pen containing nine pigs. The problem is to build two more square enclosures that would put each pig in a pen by itself.
Contrast these with two so-called ‘analytical’ problems. In the first, imagine that there are three playing cards face down on the table in front of you. You are given the following pieces of information:
Figure 6. Insight problems. (a) Move three coins to invert the triangle. (b) Draw two squares to give each pig its own enclosure. (The answers are in note 6.3 on page 236.)
To the left of a queen there is a jack.
To the left of a spade there is a diamond.
To the right of a heart there is a king.
To the right of a king there is a spade.
Your job is to say what the three cards are.
In the second problem, the police are convinced that one of Alan, Bob, Chris and Dave has committed a crime. Each of the suspects in turn has made a statement, but only one of the four is true. Alan said ‘I didn’t do it’. Bob said ‘Alan is lying’. Chris said ‘Bob is lying’. Dave said ‘Bob did it’. Who is telling the truth; and who did the crime?3
In the two analytical problems, no additional knowledge has to be supplied by the problem-solver, and it is unlikely that any assumptions will be made unwittingly that would make the problems harder to solve than they already are. All that is required is a meticulous fitting together of the pieces of information – a non-trivial but in principle straightforward task – and the answer will emerge. We might imagine that, if people were asked to think out loud while they were attempting an analytical problem, their words might track their thoughts quite easily and accurately, and would be positively related to the actual solution.
But with the insight problems, we might argue that a different kind of ‘thinking’ is required, one which is more of the intuitive, behind-the-scenes kind; and in this case if people are required to think aloud, this might actually interfere with the intuitive process. As Schooler says in his paper: ‘Verbalisation may cause such a ruckus in the “front” of one’s mind that one is unable to attend to the new approaches that may be emerging in the “back” of one’s mind.’ Schooler’s study did in fact analyse what subjects said as they were working on the two types of problem, and found exactly what had been predicted. Subjects who are solving analytical problems are neither helped nor hindered by the demand to think about and to verbalise what they are doing. But when they are attempting insight problems, they are very considerably hampered when they are simultaneously required to attend to and articulate what is going on in their minds.
In one variation of the experiment, subjects were informed before they started that they would be working on two kinds of problems, and that one kind, the ‘insight’ problems, would typically lead them into an approach that did not work. They were given an example of an insight problem, and told that if they got stuck, it would probably help to try to find a different approach or a new perspective. As before, some of the subjects were told to think aloud, and others were not. There were two interesting results. The first was that this heavy hint was of no benefit at all in solving the insight problems, and did nothing to offset the decrease in performance produced by thinking aloud. It looks strongly as if the way of knowing that leads to success in the insight problems is not only outside of conscious awareness, but outside of conscious control as well. If the processes of intuition are beyond voluntary control, then there is no way in which the subjects can make use of the ‘helpful hints’ which they have been given.
The researchers also found that the information about insight problems did have a marked effect on the solution of the analytical problems. When subjects were thinking aloud, the hint severely damaged their ability to solve the logical puzzles. Just as the attempt to solve insight puzzles through exclusive reliance on d-mode is misguided, so performance in analytical problems can be impeded if you sow doubt in people’s minds about their straightforwardness. The suspicion that something might be trickier than it actually is causes the confident use of d-mode to falter, as people endeavour to seek out – intuitively – complexities that do not exist. This result reinforces the point that the selection of the right cognitive mode is a matter of appropriateness, and not of the absolute superiority of one way of knowing over another.
Listening to the tapes of people’s verbalisations, it became clear to Schooler and his colleagues that the contents of the problem-solvers’ thoughts, as they tackled the two types of problem, were different. People wrestling with the analytical problems talked fluently, and most of their comments referred to the problem itself. However, when they were attempting the insight problems, subjects paused more frequently, and the pauses were longer: there were many more occasions on which there was, seemingly, nothing going on in the problem-solvers’ minds. And when people doing the insight problems did verbalise, they were four times more likely to make the kinds of comments that referred not to the logic of the problem but to their own mental state. They would say things like ‘There is nothing that’s going through my mind that’s really in any kind of . . . that’s in a verbal fashion’; or ‘I know I’m supposed to keep talking but I don’t know what I am thinking’. And this experience of ‘nothing going on’ was actually correlated with success on the insight problems. Those subjects who paused more solved more problems. Keeping up a running mental commentary really does interfere with the slower, less conscious processes going on at the back of the mind, and causes a drop in intelligence and creativity. We must presume that people for whom such chatter is habitual are thereby hampered when it comes to dealing with problems of greater subtlety or indeterminacy.
Jonathan Schooler’s general point is of enormous significance. Some of what we know is readily rendered into words and propositions; and some of it is not. Some of our mental operations are available to consciousness; and some of them are not. When we think, consciously and articulately, therefore, we are not capturing accurately all that is going on in the mind. Rather we are selecting only that part of what we know which is capable of being verbalised; only those aspects of our cognition to which conscious awareness has access. We think what is thinkable; not what is ‘true’. And the disposition to treat all problems as if they were d-mode problems thus skews our thoughts and our mental operations towards those that can be made explicit.
Other areas of our psychological life show similar effects. Take memory. There are many experiences that defy articulation, and our memories of them must therefore rely on non-verbal records. Our ability to recognise a huge collection of human faces, for example, with a remarkable degree of accuracy and effortlessness, attests to the power (and, I would say, ‘intelligence’) of these unspoken processes. What we can say about a face or an expression is a small fraction of what we can know. Thus it should come as no surprise to discover that the effort to describe a face so narrows our attention, a
nd biases it towards the little that can be said, that memory is reduced quite severely. In another of his studies, Schooler gave subjects photographs of unfamiliar faces to study, and asked them to attempt to describe some of the faces but not others. These pictures were then mixed up with some new ones of rather similar-looking people, and the subjects were asked to pick out the ones they had seen before. The faces that had been described were recognised about half as well as those that had not, and this impairment was unrelated to how detailed or accurate a particular description had actually been. The same result is obtained if the stimuli are simple patches of colour.
The problem with description is twofold. First, the effort to describe the face forces one to break it up into its articulable features, and focuses attention on what can be said at the expense of what is genuinely (but non-verbally) distinctive. And secondly, at the time of recognition one may be trying to retrieve the ‘written records’ from memory, and match these to the pictures that one is being shown, rather than relying on the non-verbal, sensory records that have been registered. If this retrieval effect is a significant part of the problem, performance might be improved by preventing people from using the verbal ‘code’ while they are doing the recognition test. This could be achieved by forcing people to make their memory judgement very fast, perhaps. If you deprive them of the time it takes to think, they might have to fall back on the visual information which has been overshadowed by the verbal description, and thus overcome the interference. This is exactly what Schooler showed. When people had to make their recognition decisions quickly, the deleterious effect of verbalisation was removed.
