Hume's Enquiry Concerning Human Understanding: Causation and Induction

Your train has arrived on time every morning this month. You leave home expecting the same service tomorrow. Asked for a reason, you point to its record. Asked why that record tells you what will happen next, you might say that a service which has been reliable will probably remain reliable. That is a useful everyday answer. It also repeats the step that needs explaining: the move from cases already observed to a case you have not observed.

This invented example introduces the question at the center of David Hume's Enquiry concerning Human Understanding. What allows experience to inform expectations about something beyond experience? Hume is interested in the ordinary confidence that makes planning possible. His challenge becomes difficult precisely because the inference feels so familiar. The train's past arrivals are evidence about its past. Tomorrow's arrival is a further claim.

The work was first published in 1748. The primary text linked throughout this chapter follows Hume's later, 1777 text, with section and paragraph references such as E 4.16. Its central discussion moves from the origin of ideas to causal inference, then from skeptical doubts to an account of how belief actually operates. That order matters. Hume asks both whether an inference has an argumentative foundation and what moves a human being to make it. Those questions can have different answers. Primary text and editions.

Hume begins his account of ideas with a distinction between experiencing something and thinking about it. Feeling heat differs from remembering heat. Anger as you undergo it differs from calmly recalling an earlier quarrel. In Section II, he calls the more forceful perceptions impressions and the less forceful ones ideas. Impressions include feelings and desires as well as sensory experiences; they are not restricted to messages from the eyes and ears. Ideas include thoughts and recollections. E 2.1–3.

The imagination can combine materials in ways that experience has never presented together. Hume's example is a golden mountain: someone acquainted with gold and mountains can join their ideas. Creative thought therefore need not copy a complete scene. His general proposal is that its simpler materials originate in impressions, even when their eventual combinations are unfamiliar. This gives him a method for inspecting obscure philosophical language. Ask which experience supplies the idea the word supposedly expresses. A fluent definition does not settle that question. E 2.5–9.

Hume also supplies a difficulty for his own proposal. Imagine someone familiar with a graduated series of blue shades except for one missing shade. He thinks the person could supply the missing idea despite never having received that precise impression. He treats this as a singular exception rather than a reason to abandon the general maxim. The exception deserves to stay visible: presenting his empiricism as a rule he allows no counterexample would misrepresent his argument. E 2.8.

For a modern illustration, imagine a discussion in which a manager explains every success by invoking a team's mysterious "energy." Hume's method invites clarification. Does the speaker mean enthusiasm, rapid responses, willingness to cooperate, or something else that can be described through experience? The question does not prove that every useful concept must name an observable object. It asks what content a speaker actually has available when a familiar word seems to explain everything.

Section IV distinguishes relations of ideas from matters of fact. Hume places arithmetic, algebra, and geometry in the first group. His arithmetic example concerns the equality between three times five and half of thirty. Discovering such a relation does not depend on finding fifteen objects in the world. A statement about an actual event has a different status: its contrary can be intelligible even when we have overwhelming evidence against it. E 4.1–2.

Consider a simple comparison. Denying the equality of fifteen and half of thirty conflicts with the arithmetic relation. Denying that tomorrow's train will arrive punctually does not contradict the concept of a train. You can describe a delay without contradicting yourself. That possibility does not make a delay likely. It shows why a proof of an arithmetic relation and evidence for a future arrival are different achievements.

Hume asks how we reason about facts beyond what we presently perceive or remember. His answer is that such inferences depend on cause and effect. A letter can support a belief about an absent friend; an artifact can support an inference about its maker. In each case, a present fact functions as evidence for another because we take them to be connected. His question therefore concerns inference to absent and unobserved facts as well as prediction of the future. E 4.3–6.

Knowing a particular causal relation requires experience, Hume argues. The visible qualities of an entirely unfamiliar object do not reveal everything it can do. Without relevant experience, seeing water would not disclose that a person could drown in it. Once a connection becomes familiar, its effects can seem obvious from the cause itself. Hume wants us to notice how learning has disappeared from that feeling of obviousness. His argument concerns what unaided inspection and reasoning can establish, rather than whether later investigation can discover useful mechanisms. E 4.6–8.

The argument usually called the problem of induction begins when Hume presses beyond the answer "experience." Suppose bread has nourished you in the past. That describes an observed pattern. Expecting apparently similar bread to nourish you again extends the pattern to another occasion. The new expectation contains something the record of earlier meals alone does not contain. Hume separates these propositions explicitly in E 4.16. Section IV, Part II.

Why take the further step? A demonstrative argument cannot establish that nature must continue in the same way, Hume argues, because a change in its course involves no contradiction. We can conceive apparently familiar objects producing unfamiliar effects. His claim is about conceivability and demonstration. He is not announcing evidence that nature is about to change or recommending that ordinary expectations receive equal weight with every imaginable alternative. E 4.18.

An argument from past experience seems more promising. We have relied on experience before, and it has often served us well. Why not trust it again? The trouble is that this answer itself moves from past cases to a new one. It assumes the very transition it was supposed to establish. To defend projection by saying projection has worked so far is to use projection in its own defense. Hume identifies that circularity in E 4.19. Section IV.

Return to the train. A detailed explanation involving timetables, staff, and maintenance can improve your prediction. It may reveal why this morning's punctuality differs from next week's circumstances. But the explanation still carries expectations about how those arrangements will operate outside the observed cases. Adding mechanisms can make an inference better informed without turning its factual premises into logical necessities. The example illustrates Hume's question; it is not an additional proof that all predictions are equally uncertain.

This last distinction prevents an overstatement. The challenge is stronger than the observation that predictions can fail: it asks for the foundation of the move that makes prediction possible. Yet Hume continues to distinguish careful reasoning from careless reasoning. Failure to provide a noncircular foundation for experiential inference does not by itself show that choosing between two evidence-based expectations is pointless. The remaining sections explain why inquiry and action continue.

Hume's Section V offers what he calls a skeptical solution. Someone encountering sequences of events for the first time would observe succession without immediately knowing what to expect. Repeated conjunction changes that person's response. The appearance of one event comes to carry thought toward another. Hume names this propensity custom or habit. He expressly says that naming it does not give an ultimate explanation of why the propensity exists. E 5.3–5.

Habit here means more than a routine someone has deliberately practiced. It is a principle through which experience shapes anticipation. The mind's transition becomes available without the person working through a fresh argument each time. Hume therefore gives a psychological account of how expectations arise while leaving the preceding argumentative difficulty in place. Explaining why someone makes an inference and providing a noncircular argument for that inference are different tasks.

This also changes how the skeptic should understand ordinary life. Hume says that nature will preserve the reasonings of common life despite abstract doubts about their foundation. Discovering that their basic transition lacks the proposed argumentative support does not extinguish the transition. The skeptical investigator still expects food to nourish and other people to respond. Hume's account includes the investigator among the human beings being investigated. E 5.2.

An imagined case makes the distinction practical. Someone learns that a colleague usually replies after lunch. After enough repetitions, an empty inbox in the morning stops being alarming. A habitual expectation has formed. Whether it is a good expectation on a particular day remains open to further evidence: perhaps the colleague is away. Hume's account of custom helps explain the expectation's formation. It should not be turned into advice to preserve it against new information.

Section VII asks a more specific question. What supplies our idea that a cause and an effect are necessarily connected? Watching one billiard ball strike another gives a sequence of events, Hume argues, rather than an impression of an intrinsic tie that makes any other result impossible. Repeating the sequence does not insert a newly visible tie between the balls. Something changes in the observer: the first event now brings an expectation of the second. E 7.6 and 7.28–30.

That customary transition supplies the impression from which Hume explains the idea of necessary connection. His account distinguishes the observed conjunction from the felt determination of thought. It does not require us to deny that scientists can observe forces through their effects or construct successful explanatory models. The disputed issue is access to a necessity that makes the effect follow from the cause through its very nature.

In E 7.29, Hume offers two definitions of cause. One describes an object followed by another, with similar objects followed by similar results. The other describes an object whose appearance carries thought to the other. The first emphasizes regularity; the second emphasizes inference. He also introduces a formulation concerning what would have happened had the first object not existed. His compressed presentation leaves questions about how these formulations fit together. It should not be advertised as a complete modern account of causal explanation. Section VII.

A difficulty remains even before those interpretive questions. Reliable sequences can occur without the first event producing the second. Imagine a station clock chiming shortly before a train arrives because both follow a schedule. The chime alone does not propel the train. This invented example raises a challenge for any simple equation of causation with regular succession. Hume's central achievement here is an analysis of what experience and mental transition contribute to our causal ideas; that achievement does not settle every question about distinguishing causes from common effects.

A reasonable objection is that Hume has substituted a description of belief for a justification of belief. Knowing why a habit developed does not show that it is trustworthy. The objection identifies a real limit of the skeptical solution. Hume's Section V does not produce the missing noncircular foundation. It explains why that missing foundation does not leave human beings unable to infer or act.

Another objection asks whether demonstrative certainty was the wrong demand from the outset. Perhaps an inference counts as reasonable because it responds well to evidence, even though it cannot justify the whole practice of responding to evidence without using that practice. This is a proposed philosophical response, rather than a claim that Hume explicitly resolves the difficulty in these terms. Its strength is that it distinguishes good local reasoning from a foundation for all empirical reasoning. Its unresolved task is to explain those standards without simply assuming the conclusion under dispute.

Hume's own discussion of testimony shows why his skepticism should not flatten evidential differences. In Section X, he recommends proportioning belief to evidence and distinguishes stronger from weaker assurance. He examines the circumstances that support or undermine testimony, including disagreement between witnesses and their interests in what they report. He then applies the comparison to miracle reports. His argument weighs testimony against experience supporting natural regularities; it is not merely the claim that anything unusual is impossible. E 10.3–13.

That discussion raises an objection of its own: how should an investigator distinguish an inadequately supported report from evidence that a supposed regularity needs revision? Hume's procedure requires a comparison of evidence, and the result depends on the strength attributed to each side. Reporting his criticism of miracle testimony as a definition that automatically excludes every challenge to an accepted law would hide the epistemic question rather than answer it.

Hume's distinction can improve a discussion about evidence when its limits are kept clear. Suppose, in an invented workplace example, a new process has improved delivery times in three small trials. It is reasonable to ask whether the trials resemble the work ahead. Perhaps the later tasks are larger or depend on a different team. These questions concern how well a specific inference is supported. They do not resolve Hume's more general question about why any observed pattern should extend beyond its observed cases.

The difference matters because philosophical skepticism can become an excuse for poor judgment. A manager cannot infer that every forecast deserves the same confidence just because no forecast is a demonstrative proof. The relevant practical work remains comparing observations, examining differences between cases, and responding to contrary evidence. Section X supports that continuing attention to degrees of evidence. The workplace procedure here is a contemporary application of that distinction, not a method Hume designed for managers.

In personal relationships, consider the difference between trusting someone because they have kept promises and treating that record as a guarantee. The first allows experience to guide expectation while leaving room for new information. The second demands more than a past record establishes. This distinction does not decide whom you should trust. It helps identify the inference you are making, which is useful when confidence has become so familiar that you mistake it for necessity.

Section XII develops mitigated skepticism. Hume recommends caution and modesty about our powers, together with a restriction of inquiry to subjects suited to those powers. He contrasts this with an excessive skepticism that cannot remain effective in the occupations of common life. The result is neither universal certainty nor durable disbelief in everything. E 12.21–25.

His final paragraph challenges works that contain neither abstract reasoning about quantity or number nor experimental reasoning about fact and existence. The incendiary language targets speculative pretensions to knowledge. Reading that ending alongside the preceding defense of cautious inquiry matters: the book has spent its central sections examining the limits of the reasoning on which common life depends, while continuing to use and discriminate among forms of that reasoning. E 12.34.

The question about tomorrow therefore survives a useful answer about what to do. You can check the timetable and make a sensible plan. You can revise it when evidence changes. You can also recognize that the expectation reaches beyond the past arrivals you observed. Hume makes that additional step visible, then asks us to understand the habits that make it possible.

• Hume, An Enquiry concerning Human Understanding. Read Sections IV and V together, then Section VII; the skeptical argument and the explanation of belief answer different questions.
• Hume, A Treatise of Human Nature, Book I, Part III. A further primary-text route into his treatment of causal inference.
• Hume, Enquiry, Section X. Read this discussion of testimony alongside Section XII's account of mitigated skepticism to test how the foundational doubt coexists with standards of evidence.