Imagine dropping a stone a thousand times, and a thousand times it falls. You are utterly certain it will fall on the thousand-and-first. But pause and ask: what, exactly, justifies that certainty? You have never seen the “necessity” of the stone’s falling — you have only seen stones falling. The leap from what has happened to what will happen seems obvious, yet when we examine it, no logical proof holds it up. This small abyss, opened by David Hume in the eighteenth century, has never quite been closed. It swallows causation, induction, the self, and much of metaphysics — and it is the best place to start understanding why Hume is perhaps the most unsettling of modern philosophers.
1. Hume and the Scottish Enlightenment
1.1 Life and context
David Hume was born in Edinburgh in 1711 and died in the same city in 1776. His maturity coincided with the height of the Scottish Enlightenment, the extraordinary intellectual flowering that gathered in eighteenth-century Edinburgh and Glasgow such figures as Adam Smith (a close friend of Hume), Francis Hutcheson, Thomas Reid, and Adam Ferguson. It was a milieu that prized empirical inquiry, a morality grounded in human nature, and a growing distrust of metaphysical speculation and religious dogma.
Hume studied at the University of Edinburgh, tried his hand without success at law and commerce, and turned early to philosophy. He was also a celebrated historian — his History of England brought him, in his lifetime, far more fame than his philosophy — a shrewd economist, and an elegant essayist. His reputation as an “infidel” cost him the university chairs he sought in both Edinburgh and Glasgow.
1.2 The decisive works
A Treatise of Human Nature (1739–1740) is Hume’s great systematic work, published when he was barely in his twenties. Its reception was glacial: in his own words, the book “fell dead-born from the press.” Convinced that the failure owed more to form than to content, he recast his central theses in more accessible works. The foremost is An Enquiry Concerning Human Understanding (1748), a leaner and more mature version of the Treatise’s epistemology, complemented by An Enquiry Concerning the Principles of Morals (1751). Even so, it is in the Treatise that the boldest formulations appear — and it is there that we return for the most radical arguments.
2. The empiricist program
2.1 Impressions and ideas
Hume inherits from Locke and Berkeley the central empiricist thesis: all the mind’s content comes from experience. But he systematizes it with a precise distinction. All “mental contents” (what he calls perceptions) divide into two kinds:
- Impressions: the vivid and immediate perceptions — sensations, passions, emotions. Seeing red, feeling pain, experiencing anger.
- Ideas: faint copies of impressions in memory and imagination. Recalling the red, imagining the pain.
The difference is one of force and vivacity: the impression is the burning original; the idea, its pale reverberation.
2.2 The copy principle
From this Hume draws a powerful critical tool, the copy principle: every simple idea derives from a corresponding simple impression. If there is a genuine idea in the mind, it must be possible to trace the impression from which it came. The test becomes a weapon: faced with a suspect metaphysical term — “substance,” “power,” “self” — Hume asks, “from what impression is this idea derived?” When we find none, the suspicion is that we are dealing not with an idea but with an empty word.
2.3 Hume’s fork
The second instrument is the famous exhaustive division of knowledge, today known as Hume’s fork. All that we can know falls into one of two categories:
- Relations of ideas: necessary truths, known a priori, whose denial is contradictory. Mathematics and logic are the model: “the square of the hypotenuse equals the sum of the squares of the other two sides” is true regardless of whether anything exists in the world. To deny it is to fall into contradiction.
- Matters of fact: contingent truths, known a posteriori, through experience. “The sun will rise tomorrow” is not contradictory to deny — it is perfectly conceivable that it should not rise.
This blade cuts deep. At the close of the Enquiry, Hume concludes that any volume containing neither abstract reasoning about quantity and number nor experimental reasoning about matters of fact should be “committed to the flames, for it can contain nothing but sophistry and illusion.” It is the antimetaphysical program in its sharpest form.
3. The critique of causation
3.1 Where is the “necessary connection”?
Here Hume strikes his most famous blow. When we say A causes B, we assume there is between them a necessary connection — that, given A, B must occur. But Hume applies the copy principle: from what impression is the idea of “necessity” or causal “power” derived?
Watch one billiard ball strike another. What do you actually perceive? One ball moves; it touches the second; the second moves. You perceive succession (one after the other), contiguity (one touching the other), but at no point do you perceive the power that forces the second to move. Necessity is not given in the impression. However closely we examine a single isolated case, there is nothing in it that authorizes us to infer the effect.
3.2 Constant conjunction and habit
So where does the idea of cause come from? From repetition. After observing many times that moving balls set in motion the balls they touch — a constant conjunction — the mind acquires a habit. When the first ball approaches, the expectation of the second’s motion arises automatically. The “necessity” we project onto the world is, in truth, a necessity felt in the mind: the propensity of habit to pass from one idea to its customary companion.
Causation, then, is not a rational relation discovered in nature but a product of habit or custom — the principle Hume calls, with deliberate irony, “the great guide of human life.” It is not reason that makes us expect the future to resemble the past, but an instinctive disposition.
4. The problem of induction
This is the heart of Hume’s skepticism, and perhaps the most persistent problem in the philosophy of science.
4.1 The argument
Every causal inference — and with it all empirical science — rests on a supposition: that the future will resemble the past, or, more generally, that nature is uniform. Call it the principle of the uniformity of nature. Hume’s question is simple and devastating: what justifies that principle?
Try to justify it. There are two routes, and only two, given by Hume’s fork:
By relation of ideas (demonstration): could we prove the uniformity of nature as we prove a theorem? No, because there is no contradiction whatever in supposing that nature should change its course. That bread which has always nourished should tomorrow poison is strange but perfectly conceivable — and what is conceivable is not logically impossible.
By matter of fact (experience): could we justify it by experience? Here is the trap. To argue that “nature has been uniform so far, therefore it will continue to be uniform” already presupposes the principle of uniformity — for inferring from the “has been” to the “will be” is precisely what was in question. The reasoning is circular: induction is used to justify induction.
4.2 What remains
There is, then, no rational justification for induction. And note carefully: Hume is not saying we should stop making inductive inferences — it would be impossible to live that way, and he himself goes on expecting the sun to rise. The point is subtler and more disquieting: we infer from the past to the future by habit, not by reason. The foundation of all empirical science is not logic but a psychological custom that nature has implanted in us. As Hume observes, nature has not entrusted so vital an operation to the uncertainties of reasoning.
5. The self as a bundle of perceptions
The same empiricist knife cuts into the notion of a substantial self. Descartes had found in the “I think” a thinking substance, permanent and identical to itself. Hume applies the test: from what impression is the idea of a continuous self derived?
His answer, in the Treatise, is famous. When he enters most intimately into what he calls himself, he says, he always stumbles on some particular perception — heat or cold, light or shade, love or hatred. He can never catch himself without a perception, and never observes anything but the perceptions. There is no “self” behind them that has them; there is only the flux of perceptions.
Hence the bundle theory: the self is not a substance but a collection of different perceptions succeeding one another with inconceivable rapidity, in perpetual flux. The personal identity we attribute to ourselves is, once again, a fiction the imagination constructs out of the relations of resemblance and causation among perceptions — not something given in the facts. Skepticism about personal identity is thus a direct and rigorous consequence of empiricism.
6. Hume’s law and the morality of sentiment
6.1 The passage from “is” to “ought”
In Book III of the Treatise, devoted to morals, appears one of the most influential observations in the history of ethics. Hume notes that moral authors begin by reasoning about what is — facts about God, about human nature — and then, suddenly and without warning, their propositions begin to link terms by an “ought” or “ought not.” This imperceptible transition, Hume says, needs to be explained and justified, for it seems inconceivable that a new relation (the “ought”) should be deduced from others entirely different from it (the “is”).
This is Hume’s law, or the is–ought problem: from purely descriptive premises no normative conclusion follows by logic. You may describe exhaustively what the world is — you will not derive from it, by deduction, what it ought to be. (The expression “naturalistic fallacy,” coined later by G. E. Moore, is related but not identical; the two should not be conflated.)
6.2 Moral sentimentalism
If reason cannot ground morality, what can? Sentiment. For Hume, reason is a faculty that discovers truth and falsehood — relations of ideas and matters of fact. But to approve or disapprove of an action is an affective reaction, not the registering of a fact. The viciousness of a crime is not a property we observe in the act as we observe its color; it is the sentiment of disapprobation it arouses in us, creatures endowed with sympathy.
It is in this context that one reads the most provocative line of the Treatise (Book II): “reason is, and ought only to be, the slave of the passions.” Hume is not advocating irrationalism; the technical point is that reason alone never moves us to action — it calculates means, but the ends are fixed by desire, passion, sentiment. Morality belongs to the domain of feeling, not of demonstration. This moral sentimentalism would directly influence Bentham’s utilitarianism and the whole of later metaethical debate.
7. The critique of religion
Hume turned his skeptical weapons against the pretensions of religion as well — always with the caution of one who knew the social price of heterodoxy, but without softening his rigor.
7.1 Miracles
In Section X of the Enquiry Concerning Human Understanding, Hume offers an epistemological argument about miracles. A miracle is, by definition, a violation of the laws of nature, and those laws are established by the most uniform and constant experience we possess. The evidence for them is therefore the strongest possible. On the other side, the evidence for a miracle always rests on human testimony, which is fallible, prone to exaggeration, deception, and the will to believe. Hume’s conclusion is a rule for weighing evidence: no testimony suffices to establish a miracle unless its falsehood would be more miraculous than the fact it purports to attest. The argument does not prove that miracles are impossible; it shows that, by the standards of empirical reason itself, we never have sufficient grounds to believe in one.
7.2 The argument from design
In the Dialogues Concerning Natural Religion — a work Hume withheld during his lifetime, which appeared only posthumously in 1779 — he submits the argument from design (the inference from the world’s order to a divine designer) to close examination. Through the character Philo, Hume exposes the weaknesses of the analogy between the universe and a human artifact: the inference is fragile because we have experience of watchmakers making watches, but no experience of worlds being made; order might arise from principles internal to matter rather than from an external architect; and the presence of evil in the world makes it problematic to infer an infinitely good and powerful author. Hume does not convert to militant atheism — the tone of the Dialogues is deliberately ambiguous — but he dismantles the pretension that natural theology is a demonstrative science.
Critical analysis
Hume’s power lies in carrying empiricism further than his predecessors dared. If all knowledge comes from experience, then causation, induction, and the self do not survive scrutiny — and traditional metaphysics collapses.
Two labels must be distinguished. Radical skepticism (“we can know nothing”) would be self-defeating, and Hume does not hold it. His position is mitigated (or academic) skepticism: we acknowledge the limits of reason, yet go on living, reasoning, and inquiring under the guidance of habit and common sense, with epistemic modesty rather than dogmatic arrogance. Nature itself, Hume observes, cures us of excessive doubt the moment we leave the study.
The inductive challenge remains the most robust and indigestible contribution. None of the exits is easy. Justifying induction by its practical usefulness seems to confuse “it works” with “it is rationally justified.” Appealing to probability only postpones the problem, since the very calculation of probabilities about future events presupposes some uniformity. It is no accident that it was Hume who, by his own confession, awoke Kant from his “dogmatic slumber”: the Kantian response was to make causation an a priori category of the understanding — a structure the subject imposes on experience rather than something read off from it. In the twentieth century, Karl Popper granted Hume that induction has no logical justification, but drew the opposite lesson: science does not proceed by induction but by conjecture and refutation — we do not confirm theories, we only try to falsify them.
Legacy
Hume’s shadow is long. Kant built the whole Critique of Pure Reason as a response to the Humean challenge regarding causation. The logical positivism of the Vienna Circle (Carnap, Schlick) and Ayer adopted Hume’s fork almost literally as a criterion of meaning, dividing statements into analytic and empirically verifiable and discarding the rest as meaningless metaphysics. Twentieth-century philosophy of science — from Popper to the confirmation debate — revolves around the problem of induction. Contemporary naturalism sees in Hume the forerunner of a philosophy that explains mind and morality as natural phenomena, without appeal to the supernatural. And metaethics still debates Hume’s law and sentimentalism: moral expressivists and anti-realists claim Hume as a direct ancestor. Few thinkers have set so many research agendas with so few principles.
Recommended reading
- A Treatise of Human Nature (1739–1740) — David Hume
- An Enquiry Concerning Human Understanding (1748) — David Hume
- An Enquiry Concerning the Principles of Morals (1751) — David Hume
- Dialogues Concerning Natural Religion (1779, posthumous) — David Hume
- Critique of Pure Reason (1781) — Immanuel Kant
- The Logic of Scientific Discovery (1934) — Karl Popper
- Hume (1977) — A. J. Ayer
- Hume’s Skepticism in the Treatise of Human Nature (1984) — Robert J. Fogelin
← Articles