Thinking, Language, and Decisions
How people represent concepts, solve problems, and check rules, and why their judgements under uncertainty go wrong in predictable ways. The unit works through Bayes' theorem in natural frequencies, prospect theory with its loss aversion and framing effects, the stages of language acquisition and the critical period, and the laws of attention that decide what reaches thinking at all.
A wheel of fortune has nothing to do with the share of African countries among the members of the United Nations, and yet the number it stopped at moved people’s estimates of that share by twenty percentage points. Results of this kind made the study of judgement and decision making one of the liveliest fields of psychology in the second half of the twentieth century, and they earned a psychologist the Nobel Memorial Prize in Economic Sciences. This unit follows thinking from the concepts we think with, through problem solving and reasoning, to judgement under uncertainty and the choices we make between risky options. It then turns to the language we think in, and finally to attention, which decides what reaches thinking at all. On the way it keeps an eye on which of the famous findings held up when other laboratories repeated them.
1 Concepts and problem solving
1.1 How concepts are represented
Thinking works on mental representations, which are internal stand-ins for things, events, and ideas that are not present, and the most important of these representations are concepts. A concept is a mental grouping of things that are treated as the same kind, such as bird or tool, and the set of things it covers is its category; with concepts, a creature seen for the first time can be recognized as a bird and expected to fly, lay eggs, and have feathers.
The oldest account, the classical view, says that a concept is a definition, a list of features that every member has and that no non-member has. Mathematics works this way, because a triangle is any closed figure with three straight sides, but everyday concepts resist such lists. The philosopher Ludwig Wittgenstein (1889-1951) showed this in 1953 for the concept of a game, since board games, ball games, card games, and a child throwing a ball against a wall share no single feature that all of them have and nothing else has, yet everyone knows what a game is.
Eleanor Rosch turned this observation into experiments in the 1970s. When people rated each member of a category, on a scale from 1 to 7, for how good an example of the category it was, they agreed closely with each other: a robin was a very good bird and a penguin a poor one (Rosch 1975). The ratings predicted behaviour: people verified “A robin is a bird” faster than “A penguin is a bird”, and children learned the typical members first. The prototype view explains this by saying that a concept is stored as a summary of its most typical member, the prototype, and that a new object is judged by its similarity to that summary. Rosch and Mervis (1975) showed that the typical members are those with the highest family resemblance, which is the number of features they share with other members of the category. A rival account, the exemplar view (Medin and Schaffer 1978), says that no summary is stored and that a new object is compared with the remembered individual examples, which explains why people also know the unusual cases, such as that ostriches cannot fly. Concepts also come in levels, and Rosch and her colleagues (1976) found that one of them, the basic level, is special: “dog” rather than “animal” or “beagle” is the most general level at which the members still look alike and are used in the same way, so it is the level children learn first and adults name fastest.
1.2 Solving problems
A problem exists whenever there is a goal and no obvious route to it. The first laboratory study of problem solving was made by Wolfgang Köhler (1887-1967), who spent the years 1913 to 1920 at a research station on Tenerife with a colony of chimpanzees. His most famous animal, Sultan, was given two hollow bamboo sticks and fruit placed outside the cage beyond the reach of either; after an hour of failed attempts Sultan played with the sticks, pushed the thin one into the thick one, and at once used the long tool to pull in the fruit. Köhler called this insight, a sudden reorganization of the problem rather than a slow strengthening of a correct response.