Date: Mon, 18 Mar 2002 01:50:09 -0500 (EST)
From: "Keith F. Lynch" <kfl at KeithLynch.net>
To: WSFAlist at KeithLynch.net
Subject: [WSFA] Re: Waste of "Time Machine"
Reply-To: WSFA members <WSFAlist at keithlynch.net>

ronkean at juno.com wrote:
> Time travel stories would still be possible, as fantasy, but
> they would not qualify as good science fiction, even if they
> are entertaining.

No, within the story, time travel must be physically possible, or it
cannot be a time travel story.  (It could be a story about someone
*pretending* to be a time traveler.  As in Spider Robinson's 1997
novel _LifeHouse_.  (Except that that novel contains not just a fake
time traveler, but real ones too.  It's also noteworthy as the first
novel I know of to mention "spamming".))

Fantasy vs. Science Fiction is mostly about how the time travel (or
other non-mundane occurance) happens.  For instance in Tarr's and
Turtledove's _Household Gods_, a present-day woman traveled back to
ancient Rome through *prayer*.  That makes it fantasy.  Had she used
a *machine*, it would have been science fiction.

>> * Events will somehow conspire to remain consistent.  If you go
>> back to the Battle of Gettysburg, you will have been there the first
>> time. Heinlein uses this approach in several books.

> But if a time traveler goes back into the past and performs some
> action which had not been performed in history, as history existed
> before the traveler acted, then at least that one act would be
> inconsistent with the original history.

Yes.  According to this model, that simply doesn't happen.  If you go
back to kill your grandfather, either the gun will jam, or your victim
will turn out not to have been your grandfather but some other guy
whom records have always shown to have been murdered on that day.

Ted White <tedwhite at compusnet.com> wrote:
> Heinlein wrote both the alpha and the omega of time travel/paradox
> stories with "By His Bootstraps" and "All You Zombies."

Yes, those two short stories are the classic examples of this model.
In both stories, what originally looked like several people turned out
to be one person looping through time, often with amusing effects.
For instance the person who vowed not to let happen to his kid what
happened to him not only failed, but turned out to *be* his kid!
And to be both of his/her parents!  And the person who lived in fear
of being upstaged by the *real* "Dictor" when he turns up finally
realized that he *was* the real "Dictor," the person who had brought
himself forward through time.

ronkean at juno.com wrote:
> That, in turn, implies that some events are more 'important' than
> other events, in that the 'important' things may not change, while
> the 'unimportant' ones may change.  That does not make much sense,
> given that 'importance' is not a physical quality, but is rather a
> psychological quality.

Yes.  Except that what is really changing may be one's *knowledge*
of the past.  For instance in Varley's _Millennium_ it's known that
everyone in a given airplane crash was killed.  Or is it?  What if
everyone on board is rescued and promptly replaced with a brain-dead
clone?  That would perfectly account for what was observed about the
crash.  Or it would until one NTSB inspector got suspicious...  That's
where the story starts getting interesting.

>> [Multiple Worlds]

> There are enormous number of ways in which the universe changes
> unpredictably from instant to instant, on a quantum level.  Let's
> say that's at least 10^80 choices, based on the number of nucleons
> in the universe, and let's say that each instant is the Planck time.
> That's an awful lot of new universes popping up each second.  It's
> not economical,

You sound like the medieval astronomers who argued that the universe
can't really be very large since it would be a *waste of space*!

> and it seems like it might require a lot of extra dimensions.

No, none at all.  And even if it did, so what?  If that's how it
works, then that's how it works.

> And if many branching histories were true, how would we ever know?

You could play many rounds of quantum Russian Roulette.  If you
survive them all, that argues that multiple worlds is probably true.

> If there is no way to know whether a proposition like that is true,
> perhaps the proposition is not meaningful.

Spoken like a true Logical Positivist.  I suspect Multiple Worlds is
trivially true.  I'll try to explain why I think this way later, when
I get a TUIT with fewer corners.

>> * The stack theory: Change the past and history starts over from
>> that point.  The previous future is erased, except for whatever
>> parts of it the time traveler brought back with him.

> This sounds like a more economical variant of the many branching
> histories hypothesis.

Not really, since there's never any branching.  Only one timeline
exists.  It's like a videotape.  You can rewind it to any previous
point, and record again from that point forward.  But it only
contains one movie.

> So I suppose the grandfather paradox is solved by the traveler
> somehow becoming someone else, someone who was born, when he
> kills the grandfather; in other words, when he tries to kill his
> grandfather, the man he kills is not his grandfather, if he in
> fact succeeds in killing that man.

No, that's not at all what I mean.  In the Heinlein theory, he does
not succeed in killing his grandfather.  In the branching theory, he
does, and now there's a timeline in which he's never born, and there's
also the original timeline.  In this stack theory, he succeeds, and
there's only the timeline in which he was never born.  The timeline
in which he *was* born no longer exists.  But he continues to exist.

It's like wiping your computer, and then restoring it with a 1995
backup, and setting its clock to 1995.  And then also restoring a
single 2002 file from diskette.

For instance in Card's _Pastwatch_, people using a time viewer find
out why Christopher Columbus decided to sail west.  Because God told
him to!  Except that a closer look shows that it wasn't really God at
all, but a holographic video player sent back from the future of a
timeline in which he *didn't* sail west, pretending to be God.  That
timeline no longer exists.  It's not clear what happens if he doesn't
sail west, but it's evidently so awful that the people of that
timeline are willing to wipe themselves out, and their parents and
grandparents, to prevent it.  It's conjectured that in that timeline
the Aztec sailed east and enslaved Europe, Asia, and Africa, but
nobody knows for sure.  And there's no possible way to find out.

Contrast this with Hogan's branching _Proteus Operation_, in which
both a 1970s in which Nazi Germany rules all of the world except the
US, and a 2020s in which the Nazis had been an obscure party which
never had any power, have gateways to the same 1940s Germany.  People
travel back and forth through these gateways at will, visiting each
other's timelines and returning.

Or (without time travel) Asimov's _Earth is Room Enough_ in which
every family gets a parallel earth of their own, one in which life
never evolved.

> Another perspective is that if time machines are invented in
> the future, time travelers could visit the past and teach their
> ancestors how to make a time machine, or simply provide them a time
> machine.  So if backward time machines exist at anytime in the
> future, they should exist now, but they don't, so they won't.

It's been called the "timelike" Fermi paradox.  As contrasted with
the regular "spacelike" Fermi paradox which asks why aliens haven't
dismantled our solar system for parts eons ago.  Of course it's easier
to argue that time travelers don't visit because time travel is
impossible than it is to argue that space travelers don't visit
because interstellar travel is impossible.  It seems clear that
interstellar travel is not only possible, but fairly easy, given just
a few more centuries of technical progress.  And considering how many
millions of centuries old the cosmos is, and how many solar systems
there are in which life might develop, it seems unlikely that all
aliens would just so happen to be *behind* us in progress, or (as in
Star Trek) almost exactly parallel to us.

> Watching an old 'I Love Lucy' rerun, we can see events which
> happened half a century ago.

Yes, albeit fictional events.

There are (short) movies of real life made 120 years ago.  And still
pictures going back decades further.  And books, painting, and
sculpture going back thousand of years.  It's fascinating to "visit"
Herodotus' Europe, Asia, and Africa.

> When we look at a star 50 light years away, we see light that left
> the star 50 years ago.  What's the difference?

What year it was on that star when the light left depends on your
frame of reference.  There's no universally valid way to synchronize
clocks at a distance.

>> Unless there are major flaws in special relativity (which seems
>> unlikely) faster than light communication implies an easy way to
>> also get back through time communication, which leads back to the
>> classic paradox.  Imagine being able to enter a location and a
>> *future* date and time on your PC and being shown what will appear
>> then and there.

> It does not make sense to me that it could be possible to view the
> future.  Suppose that Enron stockholders could have found out last
> summer that Enron would collapse in December.  They would have sold
> their stock last summer, and Enron would have collapsed last summer.

Actually, stock prices in a world of perfect precognition have been
studied.  They would tend to increase slowly and smoothly, and all
at the same rate, like bank accounts.  Companies that would never be
profitable would never be founded in the first place.

> Is it now possible to communicate instanteously across space, using
> quantum entanglement?

Yes, but only if all you want to transmit is a random string of bits.

> If so (or even if not), how is that different from communicating
> faster than light?

It isn't.  It shares with FTL communications the fact that whether
the transmission or reception happens first depends on one's frame of
reference.  So it really is perfectly possible to send information
back through time.  If, as I said, you don't mind the fact that the
information consists entirely of random bits.  And no, you can't
even tell whether the other station is turned on or off, or has been
destroyed or not built yet, or whatever.  As far as anyone knows, it's
perfectly useless, except for sending cryptographic one-time pads
around, the one application where pure randomness is a *good* thing,
and where it doesn't matter which station is transmitting and which
receiving, so long as the information is the same at both stations.
--
Keith F. Lynch - kfl at keithlynch.net - http://keithlynch.net/
I always welcome replies to my e-mail, postings, and web pages, but
unsolicited bulk e-mail (spam) is not acceptable.  Please do not send me
HTML, "rich text," or attachments, as all such email is discarded unread.