Notes on John Perry's Dialog on Personal Identity and Immortality

The Second Night

Sandra LaFave


Miller proposes a new theory of personal identity (pi) based on Locke.

Weirob had argued that personal identity (pi) equals bodily identity (bi).

Miller’s counterargument (in modus tollens form):

P1: If pi = bi, then we would need to inspect our bodies to tell if we’re the same from day to day.

P2:We don’t need to inspect our bodies in this way. (We can imagine being the same person in a different body.)

C: Therefore, pi does not = bi.

Rather (25), "a person is just a whole composed of such stretches [of consciousness] as parts, not some substance that underlies them ... and not the body in which they occur." The concept "same person" is analogous to concepts like "same river" or "same baseball game." We understand these notions even though there is no "underlying substance" that remains the same.

So survival after death is understandable as a continuation of these stages, the heavenly stages in the appropriate relation to the earlier earthly ones. That "appropriate relation" is MEMORY. Miller gets this argument from the English empiricist philosopher John Locke (1632-1704)

The memory theory states that Person A (later in time) is the same person as Person B (earlier in time) iff Person A has Person B’s memories. ["Iff" is called the biconditional, and means that the two sides are both necessary and sufficient for each other.] In the case at hand, "person A" would be heavenly Getchen, and "person B" earthly Gretchen. So according to Miller, heavenly Gretchen would be the same person as earthly Gretchen if and only if heavenly Gretchen has earthly Gretchen's memories.

Let's use P to represent the statement "Person A has Person B's memories."

We'll use Q to represent the statement "Person A is the same person as Person B."

The memory theory thus states that P guarantees Q AND that Q guarantees P. In other words, "same memories" guarantees "same person" AND "same person" guarantees "same memories."

Refuting the memory theory means showing it's possible for Person A to have Person B's memories and yet not BE the same person (same memories do NOT prove same person); and/or possible for Person A and Person B to be the same person and yet not have the same memories.

Weirob refutes the memory theory both ways.

Refutation 1: "Same memories" does not guarantee "same person."

Remember the memory theory is committed to the view that "same memories" guarantee "same person." In other words, according to the memory theorist, the way to tell if a person is the same person is to inspect that person's memories and see if they're the same as (or a superset of) another person's memories. If so, then that sameness of memories is what entitles us to CONCLUDE that these two people are really the same person. According to the memory theorist, sameness of memories is sufficient EVIDENCE to establish sameness of person.

Weirob points out an obvious apparent counterexample. A person can remember things he didn’t really experience (e.g., a mental patient may think he is Napoleon and "remember" the Battle of Waterloo). So it seems memory is not sufficient to establish identity. "Same memories" do not guarantee "same person."

But the memory theorist might reply, "That's no refutation. There’s a difference between true memories and false ones. If the mental patient remembers the Battle of Waterloo, the mental patient simply has false memories. The mental patient's memories are false since he's NOT NAPOLEON."

Does that answer salvage the memory theory? NO. It takes for granted that we already know how to determine sameness of person, when that's exactly the thing we're looking for and CANNOT take for granted. Person A has a "false memory" of an experience in the first place precisely because Person A is not the same person who had the experience.The mental patient's memory of the Battle of Waterloo is false precisely because he's not the same person as Napoleon. Because of problematic cases like this one, the memory theorist must revise the original claim: the memory theorist must now say "same TRUE memories" guarantee "same person." THE TRAP SPRINGS: the memory theorist's argument has become circular. On the one hand, "same true memories" guarantees "same person"; but if there's any doubt, "same person" guarantees "same true memories."

Do you see how the memory theorist's original argument has now been demolished? The original argument was that "same memories" constitutes PROOF of sameness of person. The revised argument was that "same TRUE memories "constitutes proof of sameness of person. And analysis reveals that the notion of "same TRUE memories" actually presupposes sameness of person: in other words, the argument ends up assuming exactly what it was originally intended to prove. In logical language, it begs the question.

Refutation 2: "Same person" does not guarantee "same memories."

The counter to (2) comes from Thomas Reid, an 18th-century contemporary of Hume: a person can fail to remember things he really did experience. Reid gives the example of the flogged boy remembered by the Brave Officer remembered by the Old Man, but the Old Man does not remember the flogged boy — even though all are same person. A person who fails to remember his actual memories does not have a superset containing his old memories, and yet is still considered the same person. Perry doesn’t discuss this one here.

So the memory theory, as originally stated, fails to prove either side of the biconditional.

Cohen now proposes: the difference between real memory and apparent memory is that real memory is "caused in the right way". The "right way" could include "by the appropriate experiences" or "in the same stream".

Weirob’s response: the duplication argument refutes this. The duplication argument is in reductio ad absurdum form: in other words, we show that absurd consequences result from Cohen's view.

The duplication argument

To prove: memory in the same stream is not a criterion of identity.

P1: Suppose memory in the same stream were the criterion of identity (Cohen's claim).

P2: "Same stream" must mean either "same stream of brain processes" or "same stream of information or consciousness."

P3: "Same stream" cannot mean "same stream of brain processes," since heavenly Gretchen no longer has her brain.

C1: "Same stream" must mean "same stream of information or consciousness."
(By Disjunctive Syllogism from P2 and P3)

P4: The only way heavenly Gretchen could have the same stream of information or consciousness is if God duplicates her information stream or consciousness.

P5: If God can make one duplicate, he can make more than one. If he did, then both heavenly Gretchens would remember "in the same stream." They would both be Gretchen.

P6: But surely they’re not both Gretchen, since identity implies uniqueness.

C2: In other words, the two heavenly Gretchens must be, at the same time, both identical and not identical. This is a contradiction, and it follows from our initial supposition that memory in the same stream constitutes identity. Since that initial supposition leads to absurdity, it must be false.

Q. E. D. Therefore, memory in same stream is not the criterion of identity.


Sandy's X10 Host Home Page | Sandy's Google Sites Home Page
Questions or comments? sandy_lafave@yahoo.com