David Hume wrote *A Treatise of Human Nature* in his mid-twenties, in retreat in the town of La Flèche near a Jesuit college in France. The first two books appeared in London in 1739, the third in 1740. By Hume's own famous formulation the work 'fell dead-born from the press'. Only later would it become clear that this was the moment modern philosophy changed direction.

Hume takes Locke's empiricism and pushes it to the end. All ideas derive from impressions; so let us ask, for each ordinary concept, what its impression-base actually is. The answer is, again and again: none. Consider *causation*. When one billiard ball strikes another, all we ever observe is constant conjunction of two events; we never perceive a 'necessary connection' between them. Causation is not a feature of the world but a habit of mind.

This generates the *problem of induction*. What rational ground do we have for believing the sun will rise tomorrow? Not logic — that the future will differ from the past involves no contradiction. Not past experience — because that 'the past resembles the future' is itself an inductive assumption. Hume's resolution is anti-resolution: habit and nature compel us to believe, and that is our constitution rather than our justification.

The same scalpel cuts the self. When Hume looks inward, he never finds a 'self', only impressions of heat, cold, light, shade, love, hate. The self is a *bundle* of such impressions, not an underlying substance. The bundle theory is still alive in contemporary philosophy of mind.

The *Treatise*'s third book opens the 'is-ought' gap: no quantity of descriptive premises can entail a prescriptive conclusion. Morality is grounded in sentiment, not reason — 'reason is, and ought only to be, the slave of the passions'. Kant later writes that Hume woke him from his 'dogmatic slumber'. Contemporary analytic philosophy, philosophy of science, behavioural science, and decision theory all still walk the edges of the pits Hume dug.