Personal Identity: The Puzzle of Whether You Are Still You

Plutarch, in his Life of Theseus, describes the ship the Athenians kept. The sentence, in the Loeb Classical Library edition of 1914 whose hosting page states the text is in the public domain and names no translator, runs: "They took away the old timbers from time to time, and put new and sound ones in their places, so that the vessel became a standing illustration for the philosophers in the mooted question of growth, some declaring that it remained the same, others that it was not the same vessel." Two details in it are worth keeping. Plutarch is not posing the paradox that later philosophy named after him; he is reporting that the ship had already become a stock example in an argument about growth. And the disagreement he reports, some saying it remained the same and others that it did not, is the same disagreement the rest of this article is about.

Our research file credits the classic paradox to that passage, and records the sharper version Thomas Hobbes added in 1655, in De Corpore, Part II chapter 11, Of Identity and Difference. Keep the discarded timbers. Reassemble them into a second ship. Now two vessels stand in the water, one with continuous history and new matter, one with original matter and no continuous history, and each has a claim. No amount of further looking at either hull decides between them, because nothing about the hulls is in doubt. No identifier was resolved for the Hobbes passage in the research pass behind this article, so it is cited by work and chapter and by nothing more.

Our research file states that human neurons were long believed to be non-dividing and lifelong, and that the contents of those neurons are nonetheless replaced continuously. Synaptic proteins have half-lives measured in hours to days. The file says the molecular composition of a synapse turns over on the order of weeks, and that phrase is a summary rather than a measured constant, so it is given here as an order of magnitude and appears nowhere else in this article. The lipid components of the cell membrane are replaced continuously as well. The file cites nothing whatever for any of this. The measurement of synaptic protein turnover is real and published, by Cohen and colleagues in 2013, and it is listed in the sources below because our own file does not list it.

The popular version of this is wrong, and our own file says so in its own voice. The claim that every cell in a body is replaced every seven to ten years is an oversimplification. Some cells persist for a lifetime, and the file names certain neurons, the crystallins of the lens and the cells of tooth enamel among them, while the molecules inside even those long-lived cells are constantly recycled. The method behind the modern results is carbon dating applied to human tissue, published by Spalding and colleagues in 2005 and cited below. This matters here more than it would anywhere else, because the Ship of Theseus is exactly the place where a reader arriving with the seven-year belief would have it confirmed instead of corrected.

Whether the adult human brain makes new neurons at all is a live dispute and it has not been settled. Eriksson and colleagues reported new neuron formation in the adult human hippocampus in 1998. Spalding and colleagues put the rate at roughly 700 new hippocampal neurons a day in 2013. Sorrells and colleagues challenged the finding in adults in 2018, reporting that the process drops sharply during childhood and is undetectable later. All three papers are cited below and none of the three appears in our own file's bibliography. The 2013 figure does not travel without the 2018 challenge attached to it, and neither number appears anywhere else in this article, including in its summary blocks.

Our file draws the philosophical conclusion that the pattern rather than the matter is what grounds identity: memories, personality and skills persist while their substrate is exchanged, which favours psychological or functional continuity over material continuity. The file then supplies its own counter, and both belong on the page at full strength. Molecular turnover is what an organism does. Metabolism is not evidence against bodily continuity; it is bodily continuity, watched closely. On that reading the exchange of a body's molecules is the organism persisting rather than the organism being replaced, and the biological answer set out in section 08 absorbs the whole phenomenon without strain. Which of the two readings is right is precisely what this article does not settle.

John Locke published An Essay Concerning Human Understanding in 1689, and the relevant chapter is Book II chapter 27, Of Identity and Diversity. Our research file states that Locke was the first major philosopher to separate personal identity from the identity of a substance, meaning the body, and from the identity of a human being, meaning membership of a biological species. That priority claim is our file's own, it is carried here as the file's and never as a finding, and nothing in this article depends on it. What is not in dispute is the criterion the chapter became famous for: personal identity consists in continuity of consciousness, and specifically in memory. A person is the same person as the being who had some past experience if that experience can be remembered from the inside.

One correction belongs here immediately, because the rest of the section is built on it. The Stanford Encyclopedia of Philosophy presents the memory criterion as a proposal commonly attributed to Locke, while saying it is uncertain whether Locke actually held it in that strong form. Our file states it flatly as Locke's criterion. The accurate description is that the memory criterion is the view standardly drawn from Locke, and that whether Locke held it in that shape is disputed. The argument below is unaffected either way: what the criterion says, what breaks it and what repairs it are the same whoever first wrote it down.

Those are the only words of Locke's quoted in this article, and they were taken from the encyclopedia's text rather than from a page of the Essay, which is why they are attributed to Locke and cited to the Essay rather than presented as a transcription. The last part of the definition carries the whole criterion. A thing that can consider itself as itself is a thing that can reach back into an earlier moment and claim it, and on Locke's account that reaching back is not evidence of identity. It is what identity consists in.

The repair that saved the theory is the neo-Lockean move, credited by our file to Sydney Shoemaker in 1970 and to Parfit in 1984. Replace direct memory with overlapping chains of psychological continuity: memories, intentions, beliefs, desires and personality traits, each link connected to the next, no single link required to span the whole life. The general is connected to the boy through the officer, without any direct memory reaching from the general to the flogging, and transitivity survives. One citation caution goes with this. Our file's only Shoemaker entry is a book from 1963, its attached identifier resolves to a review of that book rather than to the book, and Shoemaker's 1970 paper is not in the bibliography at all, so it is named here by author and year and carries no identifier.

The second objection is that the criterion is circular. Memory can only reach experiences the rememberer actually had, so defining identity by memory looks like defining it by something that already presupposes it. Our file credits Shoemaker with the repair, in 1959: distinguish genuine memory from quasi-memory, a memory-like state that preserves both the content of an experience and its causal origin without presupposing that the rememberer is the person who had it. That distinction looks technical and it is load-bearing. Without quasi-memory the teletransporter case in section 05 cannot even be stated, because the replica's memories are quasi-memories by construction, and calling them memories would settle the question in the first sentence.

The strongest beat in the Hume material is the one he supplied against himself. In the Appendix to the Treatise, published in 1740, he admits that he cannot give a satisfactory account of what binds the bundle together, that is, of what connects one perception to the next so that they form one mind rather than a crowd. Our file says he calls this a labyrinth from which he cannot extricate himself. That single word is not quoted here either, for the same reason as the rest. The founder of the view published its unsolved problem himself, one year after publishing the view, which is a standard of intellectual honesty this article is in no position to improve on.

Kant's reply, from 1781, is the transcendental unity of apperception, and it is not a quibble. The I think that must be able to accompany all my representations is a condition of there being experience at all, so it cannot show up inside experience as one more object among the pains and colours. Our file's own gloss is that Hume was looking in the wrong place: he searched the contents of consciousness for the thing that makes those contents one person's, which is like searching a field of view for the eye. Whether that answer is a discovery or a redescription is exactly the sort of question the literature has never closed, and the Hume section is not a debunking without it.

Derek Parfit's position is reductionist. Personal identity consists in physical and psychological continuity and there is no further fact: no Cartesian ego, no soul, no deep metaphysical entity standing behind those relations, and nothing left over once every physical and psychological fact has been stated. Our file dates the position to Part III of Reasons and Persons, published in 1984, and calls that part the most influential work on personal identity in the twentieth century, which is our file's own superlative and is carried here as its judgment rather than as a finding. One date is worth correcting while the position is being introduced. Parfit published the core claim as a journal paper called Personal Identity in The Philosophical Review in 1971, thirteen years before the book, and that paper's identifier resolves exactly. Our file never mentions it. It is also the piece of this literature a reader can most easily go and check.

The teletransportation paradox, as our file states it. A traveller steps into a scanner on Earth. It records the position of every molecule in the body, destroys the original in the process, and transmits the information to Mars, where a replica is assembled from local matter. The replica has all the traveller's memories, personality, beliefs and intentions, wakes up believing without any doubt that it is the traveller, and would pass every test anyone could devise. Is the replica the traveller? Parfit's answer is that the question is the wrong one. Once every physical and psychological fact has been specified there is nothing further to discover, and asking whether the replica is numerically the same person is asking an empty question.

The plain case is not the argument. The Branch Line is. The scanner malfunctions and fails to destroy the original, so the traveller walks out of the booth on Earth while an equally continuous replica wakes on Mars. Both are psychologically continuous with the person who stepped in, and both have exactly the same claim. They cannot both be that person, because identity is one to one and nothing can be numerically identical with two different things. That is why this article gives the Branch Line the weight rather than the plain teletransporter: the plain case invites an answer, yes or no, and lets a reader keep whichever intuition they arrived with. The Branch Line makes both answers unavailable at once.

Bernard Williams published The Self and the Future in The Philosophical Review in 1970, and it was collected in Problems of the Self in 1973. He argued against psychological continuity and for a bodily criterion, using one scenario told twice. The first telling is a body swap. Two people, A and B, are told that their memories, personalities and psychological traits will be exchanged, after which the person in A's body, now carrying B's psychology, will be tortured, and the person in B's body will be rewarded. Told that way, A should hope to end up as the B-body person, and psychological continuity looks obviously like what matters.

The second telling is the same operation described from the inside. A person is told, first, that their memories will be erased. Second, that they will then be given a set of false memories belonging to somebody else. Third, that they will then be tortured. At every step the prospect is terrifying, and it is hard to argue that the terror is a mistake. Nothing in the physical description has changed between the two tellings. What has changed is which end of the process the description starts from, and the intuition reverses completely: told the second way, the body looks like what matters.

Williams's conclusion is not that the body wins. It is that our intuitions are inconsistent, that the same case described differently elicits opposite judgments, and that this casts doubt on any simple criterion of identity. He argues further that anticipating a future experience is essentially bodily: a person can anticipate their own future pain only if it is going to be felt in their own body, which makes the body the primitive locus of self-concern rather than a container the self happens to be riding in. This is the argument in the article a general reader can test on themselves, inside a single paragraph, without owning a teletransporter.

Robert Nozick's closest continuer theory, from chapter 1 of Philosophical Explanations in 1981, is a serious answer to the Branch Line rather than a dodge around it. A person at a later time is the same person as someone at an earlier time if two conditions hold: the later one is sufficiently continuous with the earlier one, psychologically or physically or both, and no other candidate at that later time is as continuous or more continuous. Our file's own illustration is the one to use. If a perfect replica is created while the original continues to exist, the replica is not the original, because the original is a closer continuer of itself. If the original is destroyed and only the replica remains, the replica is the original. Same replica, same properties, different verdict, decided by what else exists.

That is also the theory's problem, and our file states it plainly. On Nozick's account whether a person survives depends not only on their own relation to their past but on whether anybody else happens to have a better claim, which makes identity extrinsic. Parfit's objection is that this leaves survival hostage to distant events: whether the replica on Mars is the traveller turns on whether somebody on Earth lived through the scan, even though the replica's own properties are identical in both cases. He takes it as further evidence that identity is not what matters. Nozick's theory is the price of holding on to identity as one to one, and Parfit's objection is the bill.

Animalism, argued by Eric Olson in The Human Animal in 1997 and What Are We? in 2007, holds that each of us is identical with a human animal, a biological organism of the species Homo sapiens, and persists exactly as long as that organism persists. Its main positive argument is the thinking animal. There is a human animal sitting in the reader's chair right now. That animal has a working brain and is thinking the thoughts being thought there. If the reader is not identical with that animal, then two thinkers are sitting in the chair having the same thoughts, and there would be no way to tell which of the two the reader is. Since that is absurd, the reader is the animal.

Our file lists the consequences and they are not small. A person came into existence when the organism began, not when psychological capacities emerged. A person could survive in a persistent vegetative state, because the organism persists without consciousness. And uploading a mind to a computer does not preserve anybody, because the upload might well be a person but it would be a new one. That last consequence is why animalism belongs directly beside the teletransporter rather than in a list of also-rans: it is a complete, currently defended answer on which the replica on Mars is somebody else who is sincerely mistaken about their own past.

The transplant intuition is animalism's major objection, and our file states it: if a brain is transplanted into another body, most people judge that the person goes with the brain, while animalism says the person stays with the now brainless body and the recipient is a new person carrying somebody else's memories. Our file asserts what most people judge and cites nothing for it. There is a real study of exactly these intuitions, by Shaun Nichols and Michael Bruno in Philosophical Psychology in 2010, and it is cited below. Its record was verified for this article and the paper itself was not read, so this article says that the question has been tested empirically and does not say what the test found. That distinction is the difference between a citation and a decoration.

Four-dimensionalism, which our file traces through Quine in 1950, Lewis in 1976 and Sider's Four-Dimensionalism in 2001, holds that persons are four-dimensional entities extended through time in the way physical objects are extended through space. On that view a person has temporal parts, person-stages, one at each moment, exactly as a road has spatial parts, one at each location. Our file's road analogy is the clearest handle available: nobody asks which mile of a road is really the road.

David Lewis showed in 1976 what this does to fission. Before a split there are two overlapping four-dimensional persons who share their earlier temporal parts and diverge at the moment of division, so there is no crisis about which one survived: both did, and they shared a stretch. Two roads that run together for a while and then part raise no question about which road the shared stretch belonged to, because it belonged to both. Our file supplies the objection itself, and it is the right one. This multiplies people implausibly. On Lewis's account there were two persons all along, sitting in one chair, eating one breakfast, and neither of them had any way of knowing.

The version of fission this article uses is the Stanford Encyclopedia's, and it is hypothetical from beginning to end. If both cerebral hemispheres were transplanted into different bodies, the two recipients, which the encyclopedia calls Lefty and Righty, would each be psychologically continuous with the original, and they cannot both be that person because one thing cannot be numerically identical with two different things. The encyclopedia also states Parfit's position in the form worth carrying: what a person has reason to want is that there be somebody in the future who is psychologically connected or continuous with them, whether or not that somebody is them, and the further facts about numerical identity have no practical importance.

Underneath all of this sits an older metaphysical quarrel that our file names. Perdurantists hold that objects perdure, having different temporal parts at different times. Endurantists hold that objects endure, being wholly present at every moment they exist. Our file states that the persistence of persons is a key battleground in that wider dispute. It is mentioned here and not adjudicated, because the useful question for a reader is not which camp is right but what follows for a person either way, and what follows is this: on one view a life is a thing with parts, and on the other a life is one thing that keeps happening.

Narrative identity, which our file traces through Paul Ricoeur's Oneself as Another in 1992, Alasdair MacIntyre's After Virtue in 1981 and Marya Schechtman's The Constitution of Selves in 1996, holds that personal identity is constituted by the narrative a person builds about their own life, integrating past, present and projected future into a coherent whole. MacIntyre's version makes a human life a narrative quest: to understand an action it must be located in the story of a life, and to understand a life it must be located in the narratives of a community, which makes identity inherently social and historically embedded rather than a private possession.

Two technical distinctions do the work here. Ricoeur separates idem-identity, sameness, the what that persists, from ipse-identity, selfhood, the who that acts and makes promises, and locates personal identity at their intersection: selfhood is maintained even as properties change. Schechtman's narrative self-constitution view holds that a person creates their identity by forming an autobiographical narrative that is more or less coherent, that they can articulate, and that fits reality. Her claim is not that the narrative is how a person knows who they are. It is that the narrative is what makes them who they are. That distinction between the epistemic and the constitutive is what makes this a rival theory rather than a description of a habit.

Our file's clinical illustration is the loss of narrative. It names Clive Wearing, whose herpes simplex encephalitis in 1985 destroyed his hippocampi and left him with virtually no episodic memory since, and patients with Alzheimer's disease, as cases where narrative identity is progressively lost, and it says Oliver Sacks explored neurological damage that fragmented or destroyed the narrative self in The Man Who Mistook His Wife for a Hat in 1985. One citation warning belongs with that. Sacks's book appears nowhere in our file's bibliography; what the bibliography carries under that title is a study guide about the book published in 2017 by a different author, whose identifier resolves exactly to the study guide. Nothing mechanical would ever have flagged it. This article cites Sacks and does not carry that identifier.

Galen Strawson's Against Narrativity, in Ratio in 2004, is the objection that lands hardest with a reader who does not recognise themselves in any of the above. Not everyone experiences a life as a narrative. Some people are episodic: they do not weave their experiences into a story, they do not feel much continuity with the person they were at twenty, and they have a perfectly robust sense of self anyway. Strawson's charge is that narrative theorists have taken a cultural norm, Western and literary, and mistaken it for a metaphysical condition. It is carried here at full strength and not converted into a caveat, because a theory of what a person is that quietly excludes some people is failing at exactly the point it was built to succeed.

Anatta, in Pali, or anatman in Sanskrit, is the Buddhist doctrine that there is no permanent unchanging self, and it is one of the Three Marks of Existence alongside impermanence and suffering. Our file's illustration is the chariot. In the Milindapanha, around 100 BCE, King Milinda asks the monk Nagasena his name, and Nagasena replies that Nagasena is a conventional designation with no permanent self behind it, using a chariot to show what he means. A chariot is not its wheels, not its axle, not its yoke, not any single part, and not something separate from the parts either. It is a label applied to parts arranged a certain way, and so, he says, is a person. Buddhist psychology analyses a person into five aggregates, form, feeling, perception, mental formations and consciousness, none of which is a self and none of which persists unchanged. The translation in which the passage can be read is listed below; the passage itself was not fetched for this article, which is why nothing here is quoted from it.

Our file lists what Parfit's reductionism and anatta share, and the list is real rather than decorative. Both deny a further fact beyond physical and psychological relations. Both treat personal identity as conventional rather than ultimately real. Both draw the same ethical direction from it, less self-concern and more impartial benevolence. Our file names Mark Siderits, in 2003, and Miri Albahari, in 2006, as having developed the comparison, and James Giles argued in 1993 that Hume's bundle theory and Buddhist anatta are structurally identical, which is Giles's thesis and not a consensus and is attributed to him here for that reason. The difference our file also records is the one that must not be dropped. Buddhism offers a practical path, meditation, ethics and wisdom, for realising no-self experientially rather than only understanding it intellectually. Parfit's route is philosophical argument and nothing else. Without that sentence a religion becomes a footnote to an Oxford book. The experiential side, and what happens when a sense of self actually dissolves, belongs to Ego Dissolution.

The exact negative of reductionism is worth one paragraph because it makes the phrase further fact concrete. In the Gnostic framework of the Nag Hammadi texts the true self is a divine spark, a fragment of the transcendent trapped in material existence, so a person's real identity is not their body, not their personality and not their memories. Parfit says there is no further fact of identity beyond physical and psychological continuity. Gnosticism says the further fact is the only fact that matters and everything else is costume. Those two are exact negatives of one another, and the pair defines the phrase better than a definition would.

Our file tabulates ten traditions and draws a pattern from them, and this article takes one observation from that table and nothing else. Virtually all of the traditions distinguish a surface self, the empirical personality with its memories and its body, from a deep self, named as soul or atman or divine spark or consciousness or qi, and the Western argument between psychological and bodily continuity maps onto that distinction rather than escaping it. Parfit's move was to deny that either level constitutes a real self. The reverse observation is worth one line as well: every doctrine about what happens after death is an implicit theory of personal identity, because what survives death and what constitutes identity are the same question asked twice. The map of those doctrines belongs to Death and the Afterlife, in this same wing, and this article does not reproduce it. And a shared distinction across traditions is not evidence that the deep self exists; it is evidence that the distinction is natural to make.

Punishment presupposes that the person punished is the person who committed the crime. If identity comes in degrees, as Parfit holds, then punishing somebody decades after an offence may be punishing a substantially different person, connected to the offender by overlapping chains but psychologically remote from them. Our file records that Parfit explicitly endorsed reduced punishment for temporally distant crimes, and names Nickel in 2009 and Shoemaker in 2012 as having explored diminished responsibility through temporal distance. Neither work appears in our file's bibliography and no identifier was resolved for either, so both are named here and nothing more is claimed about them. The separate question of whether anyone is responsible for anything at all belongs to Free Will, in this same wing.

The sharpest case is an advance directive. A competent person signs a document refusing life-sustaining treatment in the event of severe dementia. Years later the person with advanced Alzheimer's disease appears content, is not in distress, and expresses a wish to go on living. Which self has authority? Our file names two answers and both are defensible. Ronald Dworkin, in Life's Dominion in 1993, argued that the competent self's values should prevail, and called it precedent autonomy. Jeff McMahan argued in 2002 that the demented person is a different person, so that the directive binds somebody else. This article crowns neither, because families and clinicians make this decision without a settled theory and pretending otherwise would be worse than useless. Neither work appears in our file's bibliography and no identifier was resolved for either.

Our file states that the medical literature documents children, and some adults, who have had an entire cerebral hemisphere surgically removed, usually for intractable epilepsy, and that these patients retain personal identity, form new memories and often function remarkably well, which it takes as evidence that identity does not require the whole brain. The file attaches a case count and cites no source for it, and that count is not reproduced here because the research pass behind this article could not verify it. What can be offered instead is a real study of exactly this population, by Kliemann and colleagues in Cell Reports in 2019, on the functional connectivity of the brain in adults living with a single cerebral hemisphere. Its record was verified for this article and its findings were not read, so this article cites it as the place the question has been studied and says nothing about what it found.

Suppose the neurons of a brain were replaced one at a time by silicon devices, each performing exactly the input and output behaviour of the neuron it replaced, until the brain was entirely artificial and functionally identical. David Chalmers offers two arguments that experience would survive the process. Fading qualia: if consciousness dimmed as the replacement proceeded while function stayed identical, there would have to be some single swap that made an enormous difference to experience, which is implausible. Dancing qualia: if the silicon brain had different experiences despite identical function, a switch toggling one region between substrates would make experience lurch while nothing functional changed, which is also implausible. The implication drawn from the pair is that a transition from biological to silicon need not interrupt a person at all. This article uses that argument for substrate continuity and for nothing else. Whether a program can have any experiences belongs to Machine Consciousness and Qualia. The identifier cited below is for the anthology reprint of Chalmers's paper and not for its first publication.

Mind uploading is the teletransporter in engineering dress, and our file separates three cases that behave differently. Destructive uploading destroys the original during the scan, which is Parfit's case exactly. Non-destructive uploading leaves the original alive alongside the upload, so on Nozick's closest continuer theory the original body-person is the person and the upload is somebody new carrying their memories. Gradual uploading, which our file traces to Moravec's Mind Children in 1988, replaces neurons incrementally so that nobody ever dies at any step. The file also states the branching trilemma: if an upload is a person and can be copied, then either only one copy is that person, which is arbitrary, or all copies are, which the logic of identity forbids, or identity is not what matters, which is Parfit.

Where the engineering actually stands is checkable, and our file overstates it in two places. The OpenWorm project has worked since 2014 from the complete connectome of the nematode Caenorhabditis elegans, an animal with 302 neurons and several thousand connections, mapped by White and colleagues in 1986. Our file writes that the worm has been simulated. The accurate statement is that the project works from the connectome, and the two are not the same claim. Human scale is a different order of problem: a human brain holds roughly 86 billion neurons, a figure measured by Azevedo and colleagues in 2009, and the synapse count our file prints beside it could not be verified and is not reproduced here. The Blue Brain Project at EPFL has simulated cortical columns of rat brain since 2005 and is not close to whole-brain emulation.

Susan Schneider, in Artificial You in 2019, takes the other side of this in the technological register, and our file reports her position rather than endorsing it: uploading probably does not preserve identity, and an upload would be a copy that believes it is the original. Her proposal is to wait, and to enhance the biological brain with brain-computer interfaces before attempting anything wholesale. That position keeps the Tier 3 material from reading as advocacy, which matters, because a speculative section with only enthusiasts in it is a brochure.

Progressive consciousness transfer is the idea that gradualness does the work instantaneous copying cannot: extend a mind into a new substrate slowly enough, keeping subjective experience continuous at every step, and nothing is ever interrupted. Our file's stated precedent is the molecular turnover of section 02, on the reasoning that a brain already replaces itself gradually. The mechanisms it lists are brain-computer interfaces progressively integrating digital processing with biological networks, neuromorphic chips assuming more of the computational load, and bridge neurons operating in both substrates at once. Our file states plainly that no empirical test of any of this is currently possible and that the whole proposal is theoretical. The real implant work, with real patients and real numbers, belongs to Reading the Mind. The good material here is our file's own open question: is there a threshold of replacement beyond which a person is lost, or is any amount of replacement safe provided it is gradual enough?

Two further arguments in the same section, both carried as our file carries them. Quantum indistinguishability: particles of the same type are fundamentally identical rather than merely similar, and a multi-particle state must be antisymmetric under exchange, so there is no fact about which electron is which. Our file's implication is that none of the atoms in a body are anybody's in any deep sense, and that what makes a brain a particular brain is the arrangement rather than the specific atoms, which supports substrate independence. It also notes that Leibniz's Identity of Indiscernibles, from 1686, is challenged at this level, and that in Bohmian mechanics, set out by Bohm and Hiley in 1993, particles do have definite trajectories and individuality, so identity might in principle be grounded in trajectory. Our file itself calls that extraordinarily speculative, and the atom count it prints alongside was not verified and is not reproduced here.

Two subjects owned elsewhere touch this section at exactly one point each. If a person were a simulated being, their identity would already be substrate-independent, and every question about copying, pausing, restoring from backup and running at different speeds would be the same question as uploading, asked from the inside; the argument for thinking anybody is in that position belongs entirely to Are We Living in a Simulation? and is not stated here. And rebirth is an identity puzzle before it is anything else: in the Hindu atman doctrine the self transmigrates carrying karmic imprint but typically not episodic memories, which raises our file's own sharp question, in what sense a person with none of the memories is the same person. By psychological criteria they are not. By soul or substance criteria they are. That is the Locke argument in a different vocabulary. Plato's Phaedo and the Myth of Er make the same move, with souls drinking from the River Lethe and losing their memories while keeping their nature. The case files, the counts and the evidence belong to Children Who Remember and none of them appears here.

Naive soul theory, the view that a simple indivisible immaterial substance constitutes personal identity and persists unchanged through every physical and psychological change. Our file names Plato's Phaedo, Descartes's res cogitans and some interpretations of Christian, Islamic and Jewish theology, and lists five problems. Interaction: how an immaterial thing acts on a material body, to which Descartes offered the pineal gland, which explains nothing. Pairing, which our file credits to Jaegwon Kim: even granting that immaterial souls exist, nothing says which soul goes with which body, because material things are individuated by spatial location and immaterial ones have none. Identity without criteria: if the soul has no observable properties, nobody could ever discover that theirs had been swapped, which makes the identity it supplies empirically empty. Fission: a simple indivisible soul cannot split, though minds divide functionally. And Hume's objection, that there is no impression of a soul to begin with. Our file's status line is worth keeping exactly: virtually no contemporary analytic philosopher defends the simple soul view, and even most religious philosophers working on personal identity, our file names Swinburne, who is a substance dualist, adopt more sophisticated versions of it.

Genetic essentialism, the idea that a person is their DNA. Our file gives four counters. Identical twins share a genome and are plainly different people with different identities, memories and personalities. Gene expression varies enormously with environment, development and experience. Somatic mosaicism means that different cells in one body may carry different mutations, so nobody has a single genome to be identical with in the first place. And a human body carries roughly 38 trillion bacterial cells, a figure from Sender, Fuchs and Milo in 2016 and the one paper in this part of our file that is actually named, which affect mood, cognition and behaviour, raising the question of whether a microbiome is part of an identity. The status line is that DNA establishes biological lineage and supports forensic identification, and does not constitute personal identity. That figure appears once, here, and nowhere else in this article.

The claim that memory makes the person, and therefore that amnesiacs are not persons. Our file marks this a straw man of psychological continuity theory and supplies three counters. Locke himself did not make the extreme claim, which is exactly why he distinguished person, defined by consciousness, from man, defined by biological species. Real amnesiacs remain persons: our file names Clive Wearing and Henry Molaison, known in the literature as H.M., who retained ongoing consciousness, preferences, emotional lives and procedural memories and were not blank slates. And neo-Lockean theories require overlapping chains rather than total recall, which is the whole point of section 03. Molaison lost the ability to form new episodic memories while keeping his personality, his skills and his memories from before the surgery. This is the article's necessary safety rail, because the memory criterion is an idea a general reader can easily take further than anybody ever meant it to go.

The claim that brain transplants prove psychological continuity. Brain transplants have never been performed on a human being. The thought experiments generate intuitions, not evidence, and there is no empirical data whatever on whether a recipient would experience continuity with the donor. Our file adds that head transplant proposals, and it names Sergio Canavero in 2017, remain highly controversial and have not been successfully performed on humans, and that the animal experiments in this area, it names Demikhov and White, involved head transplants with no long-term survival and severe ethical problems. Our file's status line is that these are powerful philosophical tools and that claiming they prove any theory of identity exceeds the evidence. That sentence governs every case in this article, the teletransporter included, and it is the reason none of them is described here as showing anything.

The 2020 PhilPapers Survey, published by David Bourget and David Chalmers in 2023, asked 1,785 anglophone philosophy faculty what they accept across the field, personal identity included. Our file's stated takeaway is the safe and usable part and this article carries it: there is no consensus, personal identity remains among the most contested questions in the discipline, and the field is genuinely pluralistic. Our file also prints a set of percentages for the personal identity question, and this article prints none of them, for two independent reasons. One of the labels our file reports is not among the survey's own answer options for that question, so at least one figure is attached to the wrong thing. And the research pass behind this article could not reach the survey's results at all: three separate routes to them were blocked. The paper's identifier resolves exactly, so the source is real and correctly named, and only the extracted numbers are in doubt. The survey is cited below and is used qualitatively.

The Stanford Encyclopedia sorts the answers into three rather than our file's five, and the third is the one our own file gestures at only under a label the survey it cites does not use. The psychological-continuity view: persistence consists in some psychological relation, so a person is that future being which inherits their mental features, their beliefs, memories, preferences and capacity for rational thought. The brute-physical view: persistence consists in a physical relation not involving psychology, so a person is that past or future being which has their body or is the same biological organism, and animalism is the specific claim that we human people are organisms. And anticriterialism, the further-fact option: psychological and physical continuity are evidence for persistence but do not always guarantee it and may not always be required, and there are no informative, nontrivial, necessary and sufficient conditions for persistence at all. That third position is the honest home for a reader who finishes this article convinced that both of the first two are missing something.