Re: A calculus of personal identity

From: Stathis Papaioannou <>
Date: Fri, 30 Jun 2006 23:19:41 +1000

Bruno Marchal writes:
> > Yes, sharing the memory is *not* the same as having the original > > experience, but this applies to recalling one's own past as well.> > > Are you really sure? When two people share memories, they can only > share third person information, which will trigger their respective > unsharable first person identities/memories.> When recollecting our own memories, we do recollect (approximations) of > our unsharable first person memories, which *does*, in the present, > participate into our present first person identity.> > > > > You may argue that recalling our past is different because we have > > just the right brain structure, other associated memories and so on to > > put it all in context, but in principle all of these might be lacking > > due to illness or the passage of time, or might be duplicated in a > > very good simulation made for someone else to experience.> > > Yes. Note that from a first person memory POV, perfect quasi memories > are not distinguishable from "real memories" (if that means anythings: > assuming comp "real" memories and artificial quasi memories are just > equivalent).> > > > > > The only way to unambiguously define a first person experience is to > > make it once only; perfect recollection would be indistinguishable > > from the original experience, and it would be impossible for the > > experiencer to either know that he was recalling a memory or to know > > how close to the original the recollection was.> > > I agree.> > > > > > The postulate of a first person entity persisting through time > > violates the 1st person/ 3rd person distinction,> > > I am not sure, although it makes sense, but only because eventually it > is the whole idea of objective time which is "illusory". Subjective > time, I would say, cannot be illusory, nor can subjective pain be.> > > > > > since it assumes that I-now can have 1st person knowledge of > > I-yesterday or I-tomorrow, when in fact such knowledge is impossible > > except in a 3rd person way.> > > I disagree. I do have a first person account of I-yesterday, and some > first person feelings about possible first person feelings of myself > tomorrow, all of which are non describable in any third person way. > Again we could be in agreement here. If you want I have no doubt about > my "I-yesterday", even if I don't believe at all in some absolute third > person describable notion of "yesterday". But I do "feel" I-yesterday: > I cannot separate it from "I-now" and "I-tomorrow". This is completely > independent of the fact that I may well die in a second.
My recollection of what I did yesterday is itself a first person experience, which is not shareable. If I imagine what you did yesterday, that - meaning my imagining - is also a first person experience, also not shareable. Of course, I am more confident that I can "capture" the experience of what I was doing yesterday better than I can "capture" the experience of what you were doing yesterday, but the difference between the two is one of degree, not kind. Whether I think about my own past or about someone else's experiences, I am making an extrapolation: the only thing I can know in a first person way is my own *present* experience. This is also shown by the idea that any person might become any other person by gradual mental change over a sufficiently long period of time, in which case any imagining of someone else's experiences would be a recollection of one's own experiences from the distant past. I can only know about my own past in a 3rd person way, albeit in a more intimate 3rd person way than I can know about someone else's experiences.
> > I believe it is this confusion which leads to the apparent anomaly of > > 1st person indeterminacy in the face of 3rd person determinacy in > > duplication experiments.> > > I don't see any anomaly, to be sure. Only weirdness, relative to > probable prejudices.> > > > > Let us assume as little as possible and make our theories as simple as > > possible. I *have* to accept that there is something special about my > > experiences at the moment which distinguish them from everyone else's > > experiences: this is the difference between the 1st person POV and the > > 3rd person POV.> > > OK, but just remember that in the UDA thought experiment, the first > person is almost defined by the content of a personal diary/memory. And > what makes those experiences personal here is that they are destroyed > together with the body during destructive teleportation or duplication. > But the memories refers, in the present, to subjective (first person) > past and future. We cannot have illusions about that, only about third > person extrapolation *from* that.
Yes, but when the subject reads his own diary or examines his own memory of events, that is also third person extrapolation. Do you remember what your third birthday was like? Even if you have some memory of what happened on the day, it is likely that your brain has changed so much in the intervening years that it is actually impossible to come anywhere near capturing what the experience was actually like for the child. Another three year old may actually be able to come closer to the actual experience by imagining what it would have been like than you are able to by recollection.
> > It is tempting to say that my 1st person POV extends into the future > > and the past as well, explaining why I think of myself as a person > > persisting through time.> > I would say it is in the nature of the first person to persist in > *subjective* time. I have more problem with (naive?) notion of time and > space. So again I would agree it is an "illusion" that "1-I" persists > through some notion of 3-time and 3-space, but somehow the first person > is inextricably linked to a notion of 1-time and 1-space. The illusion > would consist in believing in a sort of describable existence of myself > in some describable notion of absolute space-time.
I have the subjective experience of being a person persisting through time because I feel that I know in a 1st person way what I did in the past. If I really did know in a 1st person way what I did in the past I could not possibly doubt it, just as I cannot possibly doubt that I am having my *present* experience. However, I cannot be sure of my memories of the past just as I cannot be sure about someone else's experiences: I can only have 3rd person knowledge in either case. It could be, for example, that I have been brainwashed and my memories of the past are partly or completely false memories.
> > However, this latter hypothesis is unnecessary. It is enough to say > > that the 1st person POV is valid only in the present,> > > But what could that means? Which present? Certainly a lasting present: > it is part of my first person identity that I feel a sort of > "continuous" time, which gives sense to all my first person willingness > to to do anything. When I do a cup of coffee, I strongly believe that I > will drink and taste that very coffee: it is part of "me-doing coffee". > I would have the feeling to lie to myself taking this as an illusion, > again independently of the fact that I will "really" (in case that > means something) drink and taste that coffee.> I still think I agree with you, or that I understand what you say, but > in general what you call "illusory" is for me just living "imagery" > (just *hopefully* well founded). Only third person extrapolations can > be considered as being "really illusionary", but then, this just means > "false and hopefully third person refutable". Then it is science > (meaning doubts, a-la Descartes/Popper).
Of course, I don't actually live my life worrying about what it really means to be me and whether I can really be sure of my remembered past; like everyone else, I live my life as if I am a person in the usually understood sense of the word. But there is a distinction to be made between the normative and the philosophically rigorous.
> > and when I consider my future and past that is only 3rd person > > extrapolation.> > > I agree if you mean by "future" and "past" 3-future and 3-past. 1-past > and 1-future is not extrapolation thy are feelings continuously lived > in a lasting present. I can no more doubt of my feeling of past than I > can doubt of a headache (say). Even if time by itself does not exist at > all (which is the case with comp). The extrapolation would reside only > in some third person projection of that time, space, ... (I think we > agree, the problem could just be the term "illusion").
I'm not sure if you're saying what I was saying above by distinguishing between 1-future/past and 3-future/past.
> > What I consider myself to be as a person is then explained as the set > > of 1st person experiences related in a particular way,> > > OK. the word "related" is quite important here. One of the goal is to > isolate that relation. Technically this is done by finding the (modal) > logic of the (comp) first person and studying its Kripke or more > general "multiverse" structure.
The relationship between different stages in a person's life - how far apart two different experiences can be and still belong to the same person - is complicated and necessarily vague. If we allow that in principle anyone can change into anyone else, how can you pin down this relationship with any rigour?
> > such as believing themselves to be moments in the life of a single > > individual, having memories or quasi-memories in common, and so on.> > If I split into two that presents no problem for the 3rd person POV > > (there are two instantiations of Stathis extant where before there was > > one) nor for the 1st person POV (each instantiation knows it is > > experiencing what it is experiencing as it is experiencing it).> > > OK.> > > > > A problem does arise when I anticipate the split (which one will I > > become?) or look back at the split (*I* was the original!); there is > > no correct answer in these cases because it is based on 3rd person > > extrapolation of the 1st person POV, which in addition to its other > > failings assumes only a single entity can be extant at any one time > > (only a single 1st person exists by definition, but multiple 3rd > > persons can exist at the one time).> > > This is a little weird. You say there is no correct answer, and then > you give the comp-correct answer.> The first person is indeed just NOT first person-duplicable (unless > some added artificial telepathic trick, but in general I talk only on > the usual simple teleportation or duplication).
There is an unambiguous 3rd person descriptive answer, but no such unambiguous 1st person answer. We can still talk about 1st person expectations, which I agree is the important thing for the subject.
> > This is not to say that my mind can or should overcome [Lee Corbin > > disagrees on the "should"] the deeply ingrained belief or illusion > > that I am a unique, one-track individual living my life from start to > > finish,> > > Here you really talk about the third person extrapolation, so I agree > with you. But the first person is not deceive by its feeling of living > uniquely in time and space. It could be dangerous to say so, because it > leads to (materialism) eliminativism which eventually conclude that the > whole first person thing is an illusion. This leads to a deeply wrong > sense of "human"-irresponsibility. Well, it is a negation of the first > person. I can be sure it is wrong, as I bet you can too.
I would say that the 1st person experience is *not* an illusion in any sense of the word. It is the very opposite, in a way: the most real thing, which cannot be doubted. But extrapolating to other people or other selves in the past, future, coming out of the teleporter or whatever, that is another matter.
> > which is why in symmetrical duplication experiments I anticipate that > > I will become one of the duplicates with equal probability.> > > So you follow at least the UDA up to the third step included.> (For possible others we refer to the eight steps presentation of UDA > summarized in the following slide (just one page PDF) and paper:>>> (The paper is available in pdf too:>> > > > In asymmetrical duplication experiments with partial memory loss or > > merging, it becomes very difficult to know what to expect.> > > OK but the fourth step of UDA is just asymmetrical duplication without > memory loss.> I don't count the destruction of the original+its personal diary as a > memory loss, giving that the memory is thoroughly conserved in the > reconstitution (by the hypothesis + default assumptions on the > rightness of the substitution level and the reliability of the doctor's > work, etc.)> What about that?> Oh. I see you answer this below ...
What I meant was something like this. You are duplicated via destructive teleportation so that two copies are produced in separate locations. One copy has 40% of its pre-duplication memories missing, while the other has 30% of its pre-duplication memories missing and 30% of a stranger's memories implanted. What is your expectation of what is to happen to you as you enter the machine?
> >> All right then. Except that I am not sure I really agree with the idea> >> that the persisting "I" is an illusion. Here, it is hard for me not> >> taking account of the main UDA conclusion: the reversal> >> physics/bio/compscience/theology/numbers. So let me tell you what I> >> believe, accepting comp and the UDA reasoning: The persisting "I" is> >> not an illusion, or is less an illusion than its body or time itself.> >> What could also be an illusion is the feeling that "I" and "you" are> >> absolutely different, when the difference is only relative.> >> > See above. Perhaps "illusion" is not the right word. Is a motion > > picture an illusion? It is a series of still pictures presented in a > > particular way giving the effect of motion from the POV of a human > > observer; better to be descriptive than use words like illusion and > > delusion.> > > Yes. "illusion" is a dangerous word, especially with comp. "Illusion" > is the first person basic bread, and are real from its/her point of > view. The first person error (the real illusion) consists in its third > person extrapolation, from the movie or even from "real life".> > > > Some of the above is very much tied up with the way we use language, > > like the difference between saying there is motion on the screen or > > there is only the illusion of motion on the screen.> > > There is no 3-motion, but there is 1-motion. The first person really > perceives motion, it cannot even doubt it intellectually, even if she > believes there is no 3-motion at all here.> With comp, it will be highly undecidable that there is anything more > than numbers and their relations so that even "physical motion", even > though possible first person *plural* sharable one, is "illusionnary" > or let us say 3rd extrapolate. In that setting I use often the word > "dream" but it convey also that "wrong" aspect of the word "illusion". > The whole point is that comp forces those illusions/dreams to obey to > non trivial mathematical laws making them overlapping, glueing, etc. > The physical realities (should) emerge from that (UDA's last > conclusion).> > > > > I falter at step 7 of the UDA as given in your 2004 SANE paper, > > especially from the point after "We are almost done..."
I should clarify, what I meant was not that I disagree with step 7, but that I find it difficult to understand. Going over everything you have said in this thread, I think the only thing I really disagree with is your insistence that we can have 1st person knowledge of our past. I don't know that it makes a big difference in the final analysis, but I think it is neater, simpler and still in keeping with all the facts to say that the 1st person is necessarily tied to the present.
> Perhaps you have not enough appreciate the importance of the invariance > of the first person experience when reconstitution delays are added. It > did help people to explicitly recall the delays in the drawing 5), 6) > 7) (and 8)).> Unless I am deadly missing something, it seems to me that what follows > "We are almost done ..." can easily (?) be deduced from the six first > steps. My experience with people who got problems in the seventh step > is that they don't have really appreciate step 5).> It would help me if you could tell precisely where is the problem at > the end of step 7; so let me quote it:> > > <<We are almost done. Indeed, let us try a simple “physical experiment” > like dropping a pen. With comp, when we are in the state of going to > drop the pen, we are in a Turing emulable state.>>> > > All right? This follows directly from the comp hyp.
Yes, that part is fine.
> <<Our more probable consistent extension is undetermined by the 1-comp > indeterminacy on all the “reconstitution” of that similar states > appearing in UD* (the infinite trace of the UD).
Sorry, I don't understand this sentence.
> All right? This follows from the definition of the UD. And the UD > exists once we assume Church Thesis, and UD* (the trace of the UD) > exists thanks to Arithmetical realsim. Rememeber I defiune comp by > mainly CT and AR (+ the "yes doctor").
I think I understand this, but not the sentence to which it is referring. Am I right in thinking that the UD is just a way to generate all the computable functions (in view of the fact that they are not enumerable)?
> This follows from 6, and the invariance of the uncertainty measure, > notably for the arbitrary delay---including the null one, and the > infinite set of states appearing with a arbitrarily large delay in the > running of the UD[14]. This gives a huge set. >>
This I understand.
> All right? You almost just said exactly this in your today's > (29/june/2006) message to Brent Meeker.> And from this you can already conclude that comp entails a reversal > between physics and computer science/number theory. The next sentences > are there just for making easier the interface with the mathematical > interview of the lobian machine, which is needed only to make explicit > the manner we have to follow for extracting physics from computer > science.
The reversal between physics and number theory is something you have mentioned many times. Informally, I can see and accept that the universe may be Turing-emulable (if this is what you mean) but I don't really see how the computational hypothesis about minds leads to this result formally.
> <<It can be argued that finite computations are of measure null, and > that the only way to a measure on the states will consist in finding a > measure on the set of maximally complete computational history going > through those states, with obviously a rather hard to define > equivalence relation among computations. Still, we can show that those > (infinite) computations, as seen from some third person description of > UD*, correspond to maximally consistent extensions of our (hopefully) > actual consistent states. It is not necessary to be more precise here, > giving the non constructivism of the collection of those consistent > extensions, and the fact that we will make things utterly precise, by > directly interviewing a universal machine on those extensions, and this > by taking into account the 1/3 person point of view distinction. So, if > we grant a sufficiently robust universe, we are completely done: > physics, as the “correct” science for the concrete relative predictions > must be given by some measure on our consistent relative states. > Physics is, in principle reduced to a measure on the collection of > computational histories, as seen from some first person point of views. > We can say that in principle, physics has been reduced to computer > fundamental psychology. >>
Stathis Papaioannou
Try where your online world comes together - with news, sports, weather, and much more.

You received this message because you are subscribed to the Google Groups "Everything List" group.
To post to this group, send email to
To unsubscribe from this group, send email to
For more options, visit this group at
Received on Fri Jun 30 2006 - 09:20:43 PDT

This archive was generated by hypermail 2.3.0 : Fri Feb 16 2018 - 13:20:11 PST