We can distinguish between a priori and empirical knowledge. The former is knowledge that you have gained without having to investigate the world, while the latter is gained, at least in part, via a worldly investigation. An important kind of a priori knowledge is gained by introspection, which is where we ‘look inwardly’ and examine our own psychological states rather than ‘look outwardly’ and investigate the world. We can distinguish between inferences that are deductive and inferences that are inductive. The former kind of inference is where one moves from premise(s) to conclusion, where the premise(s) entail the conclusion (i.e. given that the premise(s) are true, the conclusion must also be true). Inductive arguments, in contrast, are inferences from premise(s) which provide support for the conclusion without actually entailing it (i.e. the premise(s) could be true without the conclusion being true). Good inductive arguments appeal to a representative sample in providing support for the conclusion. Many non-deductive inferences do not seem to have the same form as normal inductive inferences, even though they involve premise(s) which do not entail the conclusion. Instead, these inferences involve making an inference regarding what is the best explanation of a certain phenomenon – what is known as an abductive inference.