2. The use of models, and attacks on scientific thought

Many things in nature and in human nature are incredibly complicated. Humans have made great progress by simplifying reality into a “model”, and then using the model to test ideas and make predictions. Then these predictions are tested in the real world. Models are very useful. They are a simplification, though, so they can make big mistakes if you take the wrong conclusions.

Teachers and scientists use models all the time when teaching, and so do these notes. Because they simplify, they are easier to understand. They work very well, but they are far from perfect.

A model works by developing a formula or algorithm that produces an outcome based on some inputs. For example, a model of the temperature inside a glass jar left in the sun would predict how hot the air is inside the jar. Many things will affect this: how hot the air is outside, how windy it is, what type of gas is in the jar and what pressure it is, how thick the glass is, what the lid is made from and so on.

A model starts by asking which of these variables is interesting. Some variables may not be interesting because they don’t change very often, or because we think they don’t make much difference. A model can only be useful if it is accurate, and if you agree with the variables it is studying, and if the assumptions about what doesn’t matter much are true.

Models which study more than a few variables at the same time are extremely complicated. For example, it is easy to model one billiard ball bouncing around an empty table. It becomes much harder fi the table is full of balls, although not very hard. Modelling the weather is very, very hard.

Models are of course simplifications. Someone who wishes to criticise a model can easily introduce complications which the model has not addressed. This is a common tactic to attack a rational conclusion from a model. For example, evolution can be modelled easily: the basic concepts are easy to understand. Someone who does not want to accept evolution can say something like: here is the incredible human eye. How does your little model explain something so advanced?

There are articles below which talk about how science works, and these articles come back to this. For the moment, the most important question to ask someone who makes such a criticism is: what is your alternate explanation? And does their alternate explanation have any predictive power (can it be used to make predictions about things we don’t yet know). Models are often over simplifications, but they are predictive and verifiable, which is extremely important when it comes to proving that something is true.