Tuesday, October 11, 2011


Inside our minds, believe psychologists, there are things called concepts. They are the internal representation of our external behaviour towards objects. This external behaviour is often described as categorization - categorization being they way we behave differently towards different things. For example, if I see my mother, I will behave differently towards her than if I saw a stranger. My different behaviour shows that I have categorized her differently to a stranger; therefore, the argument goes, I must have a special concept of my mother that is different from a stranger.

Since the time of Aristotle, and probably before, man has been aware of definitions. A definition is a way of describing a concept, which can be used to decide which objects belong to its related category, and which do not. For example, I could create a 'necessary and sufficient' definition of a mother as 'a female relative one generation above me, of whom I am a direct descendant'. This excludes friends, aunts and grandmothers, and goes a long way to creating a boundaried concept we can all agree on.

However, is this how concepts really work - do we always create 'in or out' definitions? Not really. A psychologist called Rosch found that categorization is often a matter of degree, not absolute certainty. He used the term 'typicality' to describe this. Never mind definitions, he said, we seem to rate some dogs as 'very doggy', and some as 'not very doggy' - meaning that categories may be a matter of evaluating how typical of a concept something is. This is especialy evident at the boundaries. If you saw a fruit which was half way between an apple and a pear, you would look for features typical of either apples or pears - perhaps shape, or taste - and try to make a categorization judgement based on these different aspects. A psychologist called Smith (1998) suggested that we have in our minds a kind of 'prototype' of each concept, a cluster of features which we think typical - and these features may have different weightings depending on how important they are to the concept.

Returning to the 'mother' idea, a person may be heard to say, to someone who fits the above definition of a mother, 'you are not my mother!' What they may be saying, is: 'You are not typical of what I expect a mother to be, therefore I exclude you from the category'. Perhaps the mother has failed to show unconditional love, which the person considers a heavily-weighted characteristic of a typical mother, and has therefore failed to be rated sufficiently 'typical' of the prototype. They fit the classical definition, but this is not how the person is thinking.

However, do people even think prototypically? One big problem is context - whether I include an object in a particular category will depend on the circumstances. The person above, who denies their biological parent is their mother for nurturing purposes, may accept them as their mother for administrative purposes - for instance, when filling in a form. Even if we use prototypes to judge category inclusion, these sets of features seem to be constantly changing and context-dependent. In an excellent experiment, Rips (1984) asked people to categorize 5-inch objects as coins or pizzas. He deliberately chose a size which was between that of a coin and that of a pizza. He found people's evaluations were quite inconsistent: for instance, they might say the item was more similar to a coin, but more likely to be a pizza. This shows that people don't, for instance, only use similarity-to-prototype to evaluate category inclusion. They may be inventing special rules. Even prototypes, we could even argue, need underlying rules to decide which features are more important.

So, an alternative idea is that we don't use classical definitions, and we don't use prototypes. Perhaps, say 'theory' theorists, we have our own special theories about things, and use these pet theories to decide what categories things belong to. Certainly, as suggested by Kiel (1989), children, as they grow to adults, evaluate less and less by superficial characteristics, and more and more by using their own theories. Perhaps we all, then, have internal concepts based on our own internal theories.

Again, in the 'mother' example above, this developmental idea makes sense. We may, when young, think of a 'mother' as being something quite stereotypical, a sort of combination of our own mother, and easy stereotypes from films and books. In later life, we may develop our own deeper concepts of what a mother may be, and evaluate the category 'mother' according to much more complex criteria.

However, some psychologists believe that we don't really think this hard about things. 'Essentialists' believe that, even before we have thought about things in depth, we naturally have 'beliefs' in our minds about what things are. They assert that we believe a mother is a mother simply because she contains an essence called 'mother-ness'. In a sense, this is true - if you believe that a person is your mother, then even if you find out this is genetically untrue, or that she has not behaved as a typical mother, you might still describe her as your mother. Perhaps her 'mother-ness' is not dependent on attributes, but on a basic 'essence'.

Essentialism is difficult to justify logically, but ths is exactly the point. If someone believes in a concept, then who are we to disagree? In this sense, essentialism respects a fundamental subjectivity in thought, however illogical that thought might be.

We seem to have at least four ways to think about things. Perhaps each way of thinking has its place:

1. We can use classical definitions where we need to be clear about category inclusion (e.g. you can only call someone a thief if they have stolen something!)
2. We can use prototypes where we need to make approximate practical judgements (e.g. you can guess someone may be a potential thief if their characteristics and behaviour fit a stereotype)
3. We can use our own theories to develop our mature ideas about what a thief might be, and in what context it would be right to call them a thief
4. We can believe, without evidence, that someone is essentially a thief, based on our subjective view

Our mind, then, thinks in flexible ways, and concepts are handled differently at different times. Given this, it is not surprising that people disagree so often. They are often using different methods of thinking. A final example: let us imagine we are deciding whether someone is a 'top model'….

1. Classical definition: are they registered with a top modelling agency? Yes or no?
2. Prototype: are they pretty? do they work as a model? do they behave like a model?
3. Theory: I have decided someone is a top model these days if they work as a model internationally and earn more than £100,000 a year
4. Essentialism: they just ARE a top model - it's in their bones, through and through!

As you can see, there is always scope for disagreement. But at least we can develop an understanding of WHY we disagree about concepts! Enjoy your thinking!.......