Hume, like the philosophers who preceded him, wondered about the nature of the mind’s perceptions and considered it correct to categorise them into two different types. The first are ‘impressions’, which coincide to sensations and the reflections that reason makes on them. Clearly, their origin is experiential, since sensations come from the outside world and are independent from the thinking mind’s will. The latter, on the other hand, are ‘ideas’, which correspond to the forms of thinking (knowing, doubting, believing, esteeming etc.), which instead depend entirely on the internal processes of the thinking mind. It is precisely because of their very different origins, impressions are identifiable as the original model on which ideas are based, which are nothing more than mental processes that reshape and structure the impressions acquired from sensory experience; for this reason ideas are weaker and fainter perceptions than impressions, and moreover the creative power of the mind is limited by the presence, quantity and quality of the latter.

     Given this fundamental structure, the first decisive step taken by Hume concerns the analysis of what actually happens when the mind sets out to reason about external reality i.e. the perceptible world. In shortening the course of reasoning, it should be noted that the fundamental principle applied by the mind in its striving towards external reality is that of causality. In other words, when the mind tries to order and give a rational structure to the world, it does so by applying links of cause and effect that allow it not only to give reasons for what happens to the impressions it receives. It also, therefore, predicts what may or may not happen in the future. For example, whenever fire is experienced, there have also been impressions of ‘light’ and ‘heat’; and since this has always happened, the general law determines that fire causes effects such as light and heat. What Hume particularly reflects on is the need to have prior experience in order to establish the causal link. For it is from the multiplicity of prior experiential cases that the implication of a cause with its effect is derived through induction. Imagine a fire burning, but set aside all concrete memories of previous experiences with fire; would it be possible to state its relation to heat as necessary? Hume says no, for there is no quality implicit in ‘fire’ that determines its necessary relation to ‘heat’. Since ’cause’ and ‘effect’ are different and distinct events, it is not possible to have an a priori knowledge of effects, i.e. irrespective of the application of the principle of causality to previous experience. The question that arises philosophically at this point is the following: Is there a rational foundation that determines the necessary truth and reality of the conclusions drawn from the application of the principle of causality to perceptions derived from experience? The answer is ‘no’, because even if fire has always been perceived as heat, this does not necessarily mean that every time a fire is encountered heat will also be present in the future; in this sense Hume completes his critique of the principle of causality and determines that it is not possible to demonstrate the regularity of nature – and thus of cause-effect relations between natural events.

     In order to complete this line of reasoning and move on to the next one, a few points need to be made explicit. First of all, it is necessary to say that the principle of causality is nothing more than a conjecture, a presupposition of the thinking mind; in this sense it cannot have – and does not have – any logical value, but its value is purely psychological. The path that led to the establishment of causality as a fundamental principle is of this type: habit – belief – faith; one has faith that the world is ordered according to chains of causes and effects. Secondly, once this principle has disappeared, the possibility of inducing from perceptible reality the existence of a supersensible reality also disappears, i.e. inducing from the physical world the existence of a metaphysical world. In this sense, to the extent that it is based on the principle of causality (and for the èlenchos sake, there would be much to add) every ontological argument or proof of God’s existence turns out to be founded on a principle that is, at best, only probabilistic. Thirdly, while all judgments on factual truths (those based on experience) can only have a probable value, there are judgments about truths of reason that have universal and necessary value. This obviously refers to the fields of mathematics and logic, i.e. the knowledge of relations between ideas that is obtained through deduction. For example, numbers and mathematical operations belong to the category of ‘ideas’, and for this reason the sum 5+5=10 is a judgement of reason, since in 5+5 the result 10 is already included a priori even before we experience it in concrete terms by solving the addition. This, however, applies again only to ideas, and it is not possible to extend this nature to impressions, hence to knowledge of the external world – such an extension would be unjustified. 

     The second decisive step taken by Hume concerns substantiality. As mentioned above, Locke and Berkeley had already strongly criticised the concept of ‘substance’; for the former, substance is inscrutable, whereas Berkeley even denies the existence of any kind of material substance. Hume continues this trail of thought and shows that in truth there is no impression (i.e. let us remember, perception derived from empirical experience) that corresponds to the idea of ‘substance’. Therefore, the idea of substantiality must have a different origin, and in fact Hume makes two dimensions of it explicit. The first is related to the principle of causality: since the mind holds impressions that are accidental and transient, i.e. not necessary and permanent – e.g. the apple being initially ‘green’ and then ‘red’ –, he considers that such accidents must necessarily be sustained by a substance that has a certain degree of necessity and permanence. If something cannot exist by itself and yet exists, then it must exist thanks to something else, and this something else necessarily exists. This inference based on the principle of causality, of course, loses all value once it is recognised that the principle itself is only a conjecture. The second, however – but these are, as I said, two complementary dimensions – is related to the way in which impressions present themselves to the mind. If one takes the idea of ‘home’ (perhaps one’s own home), one can see that it is basically nothing more than a group of similar perceptions that occur together at intermittent intervals. They are present when we leave home, they disappear while we are away, and reappear when we return again. Since groups of very similar impressions occur on different occasions at different chronological intervals, the rational mind identifies them and assumes their permanent existence even when they are not actually present. Again, this is a conjecture, a fideistic belief with no real rational basis. 

     Hume’s work can be briefly summarised as follows: he showed that experience is nothing more than a collection of facts, which are given a rational regularity and structure only by virtue of belief. He then radically reduced the sphere of influence of reason and, consequently, of the reality on which an episteme – a science – can be constituted. Next, he defined as devoid of rational value any metaphysical belief and, equally, any scientific knowledge. There is obviously a difference between metaphysics and science: the latter has a practical utility, at least. In other words, although Hume states that principles such as causality are only surreptitiously applicable to reality, he simultaneously recognises that it would be impossible for human beings to live otherwise (imagine convincing yourself that every time you leave your house, your house ceases to exist!). All this work will be the indispensable propaedeutic to a real philosophical revolution by Hume’s most famous ‘pupil’: Immanuel Kant.