Tuesday, October 11, 2011


In 1854, Boole described a set of laws governing human reasoning, as though we all, at our best, think completely logically. And Piaget, a key developmental psychologist, considered the ability to reason logically one of the pinnacles of human development. We do seem to make logical inferences all the time - combining known information to deduce an unknown. However, do we really think like this consistently? Some argue that it would be a waste of brain resource to think absolutely EVERYTHING through logically. Some of our thinking habits may be short cuts - for instance, we might make a calculated guess as to another person's intentions. This makes sense in evolutionary terms. Faced with an attacking lion, I am not going to ask it to stop while I think through my options with perfect depth and logicality. I will simply make a best guess as to a suitable action!

Basic logic goes like this: it starts with a first statement (an antecedent), and then suggests that if this is true, then something else (a consequent) inevitably follows. For example, the following is an if…then statement typical of the foundations of logical thinking:

Imagine that the following is true:

(A) If the bag is heavier than 20 kg, then (B) Andrew will drop it

There are two logical assumptions based on this - and each one has a Latin name:

MP (modus ponens): If A then B - i.e. If the bag is heavier than 20 kg, then Andrew will drop it
(This is logical because it simply states the inevitable consequence. 97% of people think like this.)

MT (modus tollens): If 'not B', then 'not A' - i.e. If Andrew does not drop the bag, then it is not heavier than 20 kg
(This is logical because if the inevitable consequence hasn't happened, then the cause can't have happened. Only 72% of people think like this.)

Also, there are two illogical assumptions:

AC (affirming the consequent): If B then A - i.e. If Andrew drops the bag, then it is heavier than 20 kg
(This is illogical because Andrew dropping the bag does not CAUSE it to be heavy - it's the wrong way round! But 63% of people think like this!)

DA (denying the antecedent): If not A then not B - i.e. If the bag is not heavier than 20 kg, Andrew will not drop it
(This is illogical because the bag being light does not CAUSE Andrew to hold on to it. But 55% of people think like this!)

Some psychologists believe that people should be expected to follow these rules. Their approach is called a 'mental logic' approach. They judge humans by the very abstract and complicated rules above, and where people are illogical, they assume it is because people have failed to appreciate the logic above. One problem with this approach, is that it expects a lot from human beings. If you got a headache when you read the bit above, it is a clue that we don't naturally think like that! So a mental logic approach may be unrealistically strict.

Some psychologists believe that people aren't born knowing all this, but they try to work it out in their heads. And because our working memories are limited (we can't think of everything at once!), we tend to have a preferred mental model including some bits of the logic above, but not all of it - that would just be too complicated.

Some psychologists say 'Forget all this business about logic, people just work out which outcome is most probable!' In the above example, a probabilistic approach would suggest people think about Andrew and the bag, and guess probabilities from known factors instead of working it all out strictly and logically.

Experiments have found that, when we're given extra information, it influences us in our logical thinking, even when it's not strictly relevant. So it seems that we adapt our assumptions, not just based on logical working out, but on the amount of information we have at our disposal. This makes sense in practical tasks - we have a big memory of past events, so why not use all that information to help us to guess what will happen next?

In particular, people find it very difficult to work through any logic involving 'not', i.e. negatives, as in 'modus tollens' above. For example, if I asked you to list things in your garden that could NOT be burned easily in a fire, it would take you a while to think about it. But if I asked you to list things that WILL burn easily, that's much easier.

An evolutionary psychologist called Cosmides suggests that all the logic stuff above is a bit unnecessary. He suggests that we think in terms of practical benefits, not in terms of abstract logic. He has worked through experimental examples that seem to show that, where the chain of benefits is clear, people find it easier to think in a logical way. This makes sense: as animals, we are quite goal-orientated, and therefore, if given a clear goal, we will find it easier to work out how the constraints of a situation affect us.

It seems that mental logic is unrealistic for most people, and that most of the time we don't even create our own logical mental models. I would suggest that the balance of research favours the idea that humans find it easier to think in terms of, not reason, but probabilities in the context of a goal or objective. A few people may be good at logical abstract reasoning, but it does not come naturally.

How can this understanding help us?

Well, what it means is, don't expect your friends always to appreciate the logic of what you are saying, even if it makes sense to you! Don't expect people to be fair, abstract and neutral, even if you are. Most people are not naturally built to have such understanding. It is much better, instead, to appeal to their natural goal-seeking ability. Ask them to imagine your situation, and imagine what THEY would do in your position, if they wanted to maximise benefit. THEN, and only then, may they see your point of view.

Those teaching children may wish to apply this in, for instance, science classes. To encourage them to think things through, it may be better for a science teacher to give children a project based on rewards. They will then be driven to improve their thinking in order to overcome the situation's constraints and seek the reward!