Contribution
Two things that did not exist before. First, an argument from the theory of ideas that there is no idea of a simple, persisting self. Second, a genetic account of why we believe in one anyway: the belief is a fiction the imagination produces from resemblance and causation among perceptions, exactly as it does for the identity of changing objects. The Appendix adds a third thing, rarer: an author’s published statement that his own account of what unites the mind fails, with the two principles that make it fail named.
Key insight
The identity of a person is the same kind of thing as the identity of a ship rebuilt plank by plank. In both, the mind passes so easily from one related object to the next that it feels like surveying one unchanging object, and it covers the difference by inventing something that stays the same. Nothing in the perceptions themselves is the self; the unity is supplied by how the imagination runs over them.
Assumptions
- Copy principle. “It must be some one impression, that gives rise to every real idea” (1.4.6.2). Without an impression of self there is no idea of self.
- Separability. Perceptions “are different, and distinguishable, and separable from each other … and have no need of any thing to support their existence” (1.4.6.3); restated as App. 12.
- No perceived real connexion. “The understanding never observes any real connexion among objects”, and causation “resolves itself into a customary association of ideas” (1.4.6.16). This is carried over from Book 1, Part 3.
- Association. Resemblance, contiguity and causation are “the uniting principles in the ideal world” (1.4.6.16). Contiguity is set aside for persons (1.4.6.17).
- The question is restricted. Personal identity “as it regards our thought or imagination” is the subject; identity “as it regards our passions or the concern we take in ourselves” is left to Book 2 (1.4.6.5).
Key results
- No idea of self (1.4.6.1–3).
- Against “some philosophers” who say we are “intimately conscious” of a self, its identity and simplicity, no impression is “constant and invariable” through life. “Consequently there is no such idea.”
- “I never can catch myself at any time without a perception.” In sound sleep “so long am I insensible of myself, and may truly be said not to exist”.
- Anyone who claims to perceive “something simple and continu’d” may be right “as well as I”, but “we are essentially different in this particular”.
- The bundle (1.4.6.4). “Nothing but a bundle or collection of different perceptions … in a perpetual flux and movement.” The mind is “a kind of theatre”, but “the comparison of the theatre must not mislead us. They are the successive perceptions only, that constitute the mind”.
- The mechanism of the fiction (1.4.6.6–7). The ideas of identity (one invariable object) and of a succession of related objects are “perfectly distinct, and even contrary”, but the acts of the imagination that consider them are “almost the same to the feeling”. So we “feign” the continued existence of perceptions, and “a soul, and self, and substance, to disguise the variation”. The dispute is not “merely a dispute of words”, because the fiction comes with it.
- The cases (1.4.6.8–14).
- A mass of matter: a small addition “absolutely destroys the identity of the whole, strictly speaking”, yet we call it the same. Smallness is judged by proportion: a mountain does not change a planet (1.4.6.8–9).
- Gradual change preserves identity better than sudden (1.4.6.10).
- A ship “of which a considerable part has been chang’d by frequent reparations” is the same by the parts’ “common end” (1.4.6.11).
- Animals and plants add “a sympathy of parts”, mutual cause and effect: an oak with “not one particle of matter, or figure of its parts the same” (1.4.6.12).
- A church rebuilt in free-stone is “the same church” by its relation to the parish; we confuse numerical and specific identity, as with a recurring noise (1.4.6.13).
- A river changes all its parts within a day and remains the same, because change is what is expected of it (1.4.6.14).
- Application to persons (1.4.6.15–20).
- “The identity, which we ascribe to the mind of man, is only a fictitious one, and of a like kind with that which we ascribe to vegetables and animal bodies” (1.4.6.15).
- Identity is “merely a quality, which we attribute” to perceptions “because of the union of their ideas in the imagination, when we reflect upon them” (1.4.6.16).
- Memory makes past perceptions resemble present ones and so “not only discovers the identity, but also contributes to its production” (1.4.6.18).
- Causation: the mind as “a system of different perceptions … link’d together by the relation of cause and effect”. The soul is like “a republic or commonwealth”, which may change its members, laws and constitutions and stay the same, as a person may change “his character and disposition” (1.4.6.19).
- Against the memory theory: who remembers what he did “on the first of January 1715”? Causation lets us extend identity past memory, so memory “does not so much produce as discover personal identity” (1.4.6.20).
- No answers in hard cases (1.4.6.21). Relations “may diminish by insensible degrees”, so there is “no just standard” for when they “acquire or lose a title to the name of identity”. The questions “are to be regarded rather as grammatical than as philosophical difficulties”.
- Simplicity (1.4.6.22). The same account, “with little or no variation”, explains why we ascribe simplicity to the mind.
- The Appendix (App. 10–21).
- App. 10: “upon a more strict review of the section concerning personal identity, I find myself involv’d in such a labyrinth, that, I must confess, I neither know how to correct my former opinions, nor how to render them consistent.”
- App. 11–19 restate the case against a simple self, with new arguments: the table and the chimney (App. 13); a mind “reduc’d even below the life of an oyster” to one perception gives no notion of self (App. 16); annihilation at death is the extinction of perceptions, so they “must be the same with self” (App. 17).
- App. 20: “If perceptions are distinct existences, they form a whole only by being connected together. But no connexions among distinct existences are ever discoverable by human understanding.” So “the thought alone finds personal identity”. “Most philosophers seem inclin’d to think, that personal identity arises from consciousness”; but “all my hopes vanish, when I come to explain the principles, that unite our successive perceptions in our thought or consciousness”.
- App. 21: “there are two principles, which I cannot render consistent; nor is it in my power to renounce either of them, viz. that all our distinct perceptions are distinct existences, and that the mind never perceives any real connexion among distinct existences.” If perceptions inhered in something simple, or the mind perceived real connexions, “there wou’d be no difficulty”. He pleads “the privilege of a sceptic”, but does not “pronounce it absolutely insuperable”.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | There is no idea of a simple, persisting self | strong given the copy principle; the introspective premise is first-person | 1.4.6.2–3; App. 11, 15–16; the opponent is allowed to perceive otherwise |
| C2 | The mind is a bundle of perceptions and nothing besides | moderate | 1.4.6.4; rests on separability (1.4.6.3) and C1 |
| C3 | The identity ascribed to changing objects, minds included, is a fiction produced by an easy transition of the imagination | moderate as psychology; argued by cases | 1.4.6.6–15; the cases are observations of usage, not experiments |
| C4 | Resemblance (via memory) and causation are what produce the transition for persons | moderate | 1.4.6.17–20 |
| C5 | Memory discovers rather than produces personal identity | strong as an objection to a pure memory criterion | the forgotten dates (1.4.6.20) |
| C6 | Disputes about personal identity in borderline cases have no fact to settle them | moderate, given C3 | 1.4.6.21 |
| C7 | The account in C3–C4 of what unites the perceptions is defective | the author’s retraction | App. 10, 20–21; he names the two principles that produce the difficulty |
Concepts
- impression / idea — livelier and fainter perceptions; every real idea derives from an impression (1.4.6.2).
- perception — any content of the mind; each is a distinct existence.
- bundle — the mind as a collection of successive perceptions, with no owner among them (1.4.6.4).
- fiction — what the imagination supplies to make related succession pass for invariable identity: a soul, self or substance (1.4.6.6).
- numerical / specific identity — being one thing / being exactly alike; sometimes confounded (1.4.6.13).
- imperfect identity — the identity granted to a mass with trivial change, constituted by “the uninterrupted progress of the thought” (1.4.6.9).
Connections
- Locke’s “Of Identity and Diversity”, filed beside this reading. The section takes Locke’s analogy between persons and plants and animals and turns it against him: it holds for both, and in both the identity is fictitious. “Of late years in England” (1.4.6.15) is Locke’s debate. The argument from forgotten days (1.4.6.20) answers a memory theory that readers attribute to Locke. The Appendix’s “most philosophers … personal identity arises from consciousness” is Locke’s thesis, and Hume says it would be promising if he could explain the unity of consciousness.
- Floridi (LIT-136, NOTE-130) quotes the Appendix at length as the statement of the unity problem, and argues that Lockean and narrative accounts both inherit it.
- Dennett (LIT-442, NOTE-364) closes on Hume’s failure to find a self by introspection, and treats the self as an abstraction posited by interpretation. NOTE-364 records that the essay does not take up persistence. Hume does both: the bundle answers what a self is, and the fiction answers why we think it persists.
- Minsky (LIT-432, NOTE-348). The Single-Self as a useful fiction over a society of resources is Hume’s fiction with a mechanism named.
- Clark and Chalmers (LIT-097, NOTE-180, its C14). They argue that a self of occurrent states only would be “a mere bundle”, threatening “its deep psychological continuity”. Hume’s bundle is exactly that, and he restores continuity through causation among perceptions, not through standing dispositions.
- Parfit’s “Personal Identity”, filed beside this reading, agrees that identity is indeterminate in hard cases (1.4.6.21) and closes on a quotation from Hume’s “The Sceptic”.
Bearing on the record
It supports none of the record’s THEORY documents directly. It is a precedent for THEORY-043’s question: Hume’s commonwealth analogy (1.4.6.19) is used for persistence, not for mentality, so it does not bear on whether a commonwealth could be a mind. It carries no instruction for machine-learning practice.
Note of 2026-10-16: THEORY-681 now cites LIT-451 as a source.
Limitations
What is shown and what is asserted:
- Shown, given the copy principle: that no single constant impression could be the source of an idea of a self that is the same through life (C1). The argument is as strong as the copy principle.
- Reported, not shown: the introspective premise (1.4.6.3). Hume himself concedes another may perceive otherwise.
- Asserted with illustrations: the mechanism of the fiction (C3). The cases show how we talk about ships and churches; that the same mechanism produces belief in a self is argued by analogy (1.4.6.15) and by the “closer” argument from separability (1.4.6.16).
- Withdrawn: the account of what unites the perceptions (C7). The Appendix does not say which part of 1.4.6 fails. It says the two principles cannot be reconciled with any explanation of the felt connexion. Commentators disagree about what exactly the labyrinth is; the text does not settle it.
- Deferred: identity “as it regards our passions” (1.4.6.5), left to Book 2.
Open questions
- What would a “real connexion” among perceptions be, if one were perceived? The Appendix names two ways out (inherence in a simple substance, or perceived real connexion) and takes neither.
- Is the fiction of identity itself a perception, and if so whose? The section’s language (“we feign”, “our propension”) presupposes the subject the bundle is meant to replace. The section does not address this.
Corrections
None to the brief’s citation. One point of scope: the Appendix appeared with vol. III in 1740, a year after the section, and the edited text on the site folds the Appendix’s instructed changes into Book 1. The page’s markup shows two in 1.4.6: an ERRATA insertion (“For what is the memory”, 1.4.6.18) and one Appendix change in 1.4.6.23 (“intellectual and natural world” for “intellectual and moral world”). Neither touches the argument; otherwise the text read is the 1739 text.