Personal Identity Over Time

What It Is

A person at age six and the "same" person at age sixty share almost no matter. The atoms have turned over many times. Most of the memories are unshared; the personality has shifted; the body has grown and aged. In what sense, then, are they the same person? The question of personal identity over time asks for a relation $R$ such that for any two person-stages $P_1$ at $t_1$ and $P_2$ at $t_2$, $P_1$ and $P_2$ are stages of one and the same person iff $R(P_1, P_2)$ holds.

The candidates fall into three large families.

• Psychological criteria locate identity in continuity of memory, personality, beliefs, intentions, and other psychological states. Locke, Parfit, and Sydney Shoemaker stand in this tradition.
• Somatic criteria locate identity in continuity of the body or organism. Animalism (Eric Olson, Paul Snowdon) is the contemporary version.
• No-identity views deny a substantive relation does the work. The bundle / no-self view (Hume, Parfit in one mood, the Buddhist Abhidharma tradition) holds that the person is a temporally extended bundle of states with no underlying self that grounds identity.

The debate is not merely metaphysical: prudential reasoning ("should I save for retirement?"), moral responsibility (am I now responsible for what I did at twenty?), survival in unusual cases (teletransportation, brain swaps), and AI continuity (are uploaded copies "me"?) all turn on which $R$ is right.

This page assumes What Is Philosophy?. The intersection with Mind-Body Problem (does psychological continuity require physical continuity?) and with Plato (the Phaedo's soul-continuity argument) are taken up in those entries.

Locke's Memory Criterion

John Locke, in the second edition of the Essay Concerning Human Understanding (1694), defined a person as "a thinking intelligent being, that has reason and reflection, and can consider itself as itself, the same thinking thing, in different times and places."¹

The identity claim: a person at $t_2$ is the same person as a person at $t_1$ iff, at $t_2$, the person remembers experiencing what the $t_1$ person experienced. Memory is the relation that constitutes personal identity over time.

Locke explicitly separated the person from the human animal. An immortal soul that swapped bodies overnight would, on Locke's view, be the same human in the morning and a different person in the morning (if it remembered nothing). The person is a forensic concept: it picks out the unit of moral responsibility and prudential concern, and what makes it a unit is the connection of memory.

Reid's Objection: The Brave Officer

Thomas Reid in 1785 produced the canonical objection to Locke's criterion.²

Consider three stages of a single human being.

• Stage 1: a boy who is flogged for stealing apples.
• Stage 2: a brave young officer who captures an enemy flag in battle; the officer remembers being flogged for stealing apples.
• Stage 3: an old retired general who remembers capturing the flag in battle, but no longer remembers the apple-stealing flogging.

Apply Locke's criterion.

• Stage 2 = Stage 1 (officer remembers being flogged).
• Stage 3 = Stage 2 (general remembers capturing the flag).
• Stage 3 $\neq$ Stage 1 (general does not remember being flogged).

Identity must be transitive: if $a = b$ and $b = c$, then $a = c$. But on Locke's criterion, the officer is the boy, the general is the officer, and the general is not the boy. The criterion makes identity non-transitive, which is incoherent.

The Reid objection is taken to show that memory as a one-step relation cannot constitute identity. Two responses are available. (i) Strengthen the relation: identity is constituted by the ancestral of one-step memory (the transitive closure). (ii) Abandon identity-talk: what matters in survival is psychological connectedness, not identity, and we should describe the problem in terms of degree rather than yes-or-no.

Parfit's Psychological Continuity

Derek Parfit's Reasons and Persons (1984)³ is the central text of contemporary personal-identity theory. Parfit refined the Lockean approach in two ways.

First refinement: connectedness vs continuity. Parfit distinguishes two relations.

• Psychological connectedness is the holding of direct psychological connections: $P_2$ remembers $P_1$'s experiences, $P_2$'s personality directly continues $P_1$'s, $P_2$'s intentions were formed by $P_1$, and so on. Connectedness comes in degrees.
• Psychological continuity is the holding of overlapping chains of strong connectedness: there is a sequence $P_1, P_{1.1}, P_{1.2}, \ldots, P_2$ such that adjacent links are strongly connected, even if $P_1$ and $P_2$ are not directly connected. Continuity is the ancestral of strong connectedness.

The general remembers being the officer, the officer remembers being the boy. Continuity (the ancestral) holds between general and boy even though direct connectedness does not. Continuity is transitive by construction. Parfit's identity claim: $P_2$ is the same person as $P_1$ iff there is overlapping strong psychological continuity between them, with no branching.

Second refinement: the no-branching clause. The clause is non-negotiable. Without it, fission cases break the criterion.

The Fission Case and the Lesson Parfit Draws

Imagine that brain hemispheres can be successfully transplanted into prepared bodies and that the resulting agents retain the donor's memories, personality, and intentions. Person $P_1$'s brain is split; the left hemisphere goes into body $A$, the right into body $B$. Both resulting persons $P_A$ and $P_B$ are strongly psychologically continuous with $P_1$.

Apply the criterion without the no-branching clause: $P_A = P_1$ and $P_B = P_1$. But $P_A \neq P_B$ (they are now in different bodies, having different experiences). This violates the transitivity of identity (if $P_A = P_1$ and $P_B = P_1$, then $P_A = P_B$).

The no-branching clause prevents this: identity holds only when continuity is a one-one relation. In the fission case, neither $P_A$ nor $P_B$ is identical to $P_1$.

Parfit's argument from here is striking. Suppose we are pre-fission and we know that fission will occur. We do not want to say that we survive, but the post-fission persons are each as continuous with us as we ordinarily are with our future selves. Parfit's conclusion: identity is not what matters. Psychological continuity is what matters in survival, and psychological continuity can branch in ways that identity cannot. The question "will I survive?" should be replaced by "is there psychological continuity?" because identity-talk forces a yes-or-no answer where the underlying facts come in degrees and can be many-one.

The lesson is sometimes packaged as: personal identity is not what matters in survival. This is one of the most influential single conclusions in twentieth-century metaphysics; it has consequences for prudential reasoning (caring less about distant future selves), distributive ethics (Parfit's reductionism about persons), and the metaphysics of uploading and replication.

Animalism

Eric Olson's The Human Animal (1997)⁴ argues that we are biological organisms: the bearer of personal identity is the human animal, and the conditions on biological continuity (continuity of life processes) are the conditions on personal identity.

Two main arguments for animalism.

The thinking-animal argument. There is a thinking, sentient organism in the chair you are sitting in. It has the right physical and biological structure to think. If you are not the animal but only a person constituted by the animal, then there are two thinkers in the chair: you and the animal. That is one thinker too many. The simplest position is that you are the animal.

The fetus argument. You were once a fetus before you had any psychological capacities. The fetus is biologically continuous with the adult. If personal identity were constituted by psychological continuity, then you were never the fetus; some other entity was the fetus and you began to exist when psychology kicked in. This is widely held to be implausible.

The animalist accepts Reid's objection against Locke (psychological continuity does not constitute identity) but does not retreat to Parfit's "identity is not what matters" conclusion; instead the animalist replaces psychological continuity with biological continuity. Personal identity does what it always did (track a single subject through time); the bearer is just the organism.

Objections to animalism. The strongest concern is the transplant intuition: if your brain were transplanted into another body, the post-transplant person with your memories and personality seems to be you, not the brainless body left behind. Animalists must either deny this intuition or argue that brain transplants are special (some argue the brain carries the organism in the relevant sense, so the transplant is the same animal in a new body, but this is controversial).

The Bundle / No-Self View

David Hume in the Treatise (1739) wrote: "when I enter most intimately into what I call myself, I always stumble on some particular perception or other... I never can catch myself at any time without a perception, and never can observe any thing but the perception."⁵ The unified self is not an additional entity over and above the bundle of perceptions; it is a fiction constructed from them.

The Buddhist Abhidharma tradition reached a similar conclusion two millennia earlier: anatta, the doctrine of no-self. There is no underlying atman (substantial self) that owns the experiences; there are only the five aggregates (form, feeling, perception, mental formations, consciousness) arising and ceasing in dependent origination.

Contemporary versions (Parfit in one mood, Galen Strawson 2008⁶) hold that what exists is the temporally extended series of momentary subjects; identity-talk about "the person" is a useful abbreviation but does not pick out a fundamental entity. The view is reductionist: facts about personal identity reduce to facts about psychological connections among momentary subjects, with no further fact about whether "the same person" persists.

The view dissolves many traditional puzzles: in the fission case, neither $P_A$ nor $P_B$ is "really" identical to $P_1$ in any deep sense; what we have is a branching tree of momentary subjects with strong continuity relations along the branches. The cost: it gives up the everyday assumption that I persist through time as a unified subject.

Worked Thought Experiment: The Teletransporter

Setup (Parfit's version). A teletransporter scans your body and brain, encodes the complete state, transmits the encoding to Mars, and constructs an atom-for-atom replica on Mars from local material. The replica has all your memories, personality, intentions, and physical features. The original on Earth is destroyed at the moment the replica is constructed.

What happens?

• Psychological-continuity view (Parfit). The Mars person is psychologically continuous with you, and there is no branching (the Earth original is destroyed). The Mars person is you. Teletransportation is a form of travel.
• Animalist view (Olson). The Mars person is a replica, not the same organism. The biological continuity is broken. The Mars person is a numerically distinct individual who happens to share your memories. You died on Earth.
• No-self view (Parfit in a different mood). The question is malformed. There is no fact about whether the Mars person is "really" you; there are facts about psychological continuity (which holds) and physical continuity (which does not), and asking which of these constitutes "real" identity assumes a deeper fact that does not exist.

Variant (the two-button case). The teletransporter is buggy. After scanning, the original on Earth is not destroyed; both the Earth original and the Mars replica exist. Now the case is structurally identical to fission. On the psychological-continuity view with no-branching clause, neither is you; on the animalist view, the Earth original is you and the Mars replica is a copy; on the no-self view, the question is again malformed.

The teletransporter case is supposed to test intuitions. Strong reactions to "I died on Earth" support animalism (or somatic continuity more broadly); strong reactions to "I am the Mars person" support psychological continuity; strong reactions to "the question is malformed" support the reductionist no-self view.

Common Misconceptions

• "Locke's memory criterion is dead because of Reid's objection." The one-step version is dead. Parfit's chained version (continuity as the ancestral of connectedness) handles Reid's case. The contemporary debate is between Parfit-style psychological continuity and animalism, not between Locke and Reid.
• "Animalism implies you cannot survive radical brain damage." Animalism implies that what survives radical brain damage is you as long as biological continuity holds. The animalist person who suffers total amnesia is still the same person; she is just an amnesiac person.
• "Parfit's view implies you should not care about your future self." Parfit himself drew weaker conclusions: rationality permits caring more about psychologically connected future selves than about distant future selves, but does not require it. The reductionist view does not automatically generate present-bias.
• "The fission case is science fiction, so it has no philosophical bite." Real cases of brain hemispherectomy (one hemisphere surgically removed) and split-brain syndromes (corpus callosum severed) show that the brain admits the relevant kind of partial duplication. The fission case is an idealization of phenomena that actually occur.
• "The no-self view denies that persons exist." It denies that persons are fundamental entities. Persons exist in the sense that tables exist: as useful coarse-grained patterns over more fundamental states. This is reductionism, not eliminativism.

Comparisons to Related Views

Go Further

• Stanford Encyclopedia of Philosophy, "Personal Identity" by Eric Olson. Author-of-record survey, with strong animalist representation but fair to other views.
• Stanford Encyclopedia of Philosophy, "Identity Over Time" by André Gallois.
• Parfit, Derek. Reasons and Persons. Oxford University Press, 1984. Part Three is the canonical contemporary reference for psychological continuity.
• Olson, Eric T. The Human Animal: Personal Identity Without Psychology. Oxford University Press, 1997. Canonical animalist text.
• Shoemaker, Sydney. Self-Knowledge and Self-Identity. Cornell University Press, 1963. Foundational defense of the neo-Lockean view.
• Martin, Raymond, and John Barresi. The Rise and Fall of Soul and Self. Columbia University Press, 2006. Historical survey of personal-identity theories from ancient to contemporary.

For the metaphysics of the subject of experience that personal-identity theories assume, see Mind-Body Problem. For Plato's soul-continuity argument (an early form of the psychological view), see Plato.