|
Barney Oliver:
What do you mean by short? Short compared to what?
Monte Ross: One
nanosecond comes out to be the value that, for a given reasonable spectral
filter width, the star background photoelectrons are quite low. They are in the
range of one or less per nanosecond. This means that you have a certain number
as Jim was just talking about, you have to have a certain number of signal
photoelectrons anyway. You want to get the background down to a level below that
to compare with. You want to choose your values, from a system engineering
viewpoint, to not lose information or detectability because of the star's
background. But you do not want to go any further than that because of other
system parameters such as longer search times, etc. That's how one nanosecond
came about. There may be some better times, but short pulses are still good.
They're just like chickens soup!
Charles Townes:
Thank you. Would you like to make some comments?
Neil Tennant: I
am not a scientist. I am a philosopher, so this contribution may be very
different from the other two. I have been very interested in this session,
listening to scientists talk to one another at such high levels of technical
expertise, and getting a feel for how detailed the debate is. At the same time,
I've been quite distinctly impressed by how their being party to common
assumptions, in some crucial cases seems to leave certain assumptions still
covered; not brought out explicitly, and not examined to see whether they really
are conditioning the whole discourse. I have been reading a few more papers in
this area, and time and again I come across this assumption that communication
is possible just as soon as you detect a signal; that you can interpret what may
be said; that you will be able to tell an alien civilization by responding to
them appropriately with a message that has a content that has to be agreed on by
international protocol. It is all these philosophically loaded notions that I
want to draw to your attention as scientists. I think you move too easily from
the realm of discourse of the Shannon-Weaver theory of communication in
bits-per-second, to the area of semantically structured and semantically laden
messages. It's a very important juncture in the collaboration between scientists
and philosophers.
I am reminded of two past episodes in the
history of thought and intellectual endeavor generally. First there was the
foundation of mathematics. When mathematicians and logicians in a philosophical
way first started examining the very limits of the axiomatic and deductive
method, there was a widespread assumption at the beginning of this century, that
all the truths about certain mathematical structures could be deduced
formalistically from certain specified axioms that were clearly true about those
structures.
In a revolutionary piece of conceptual
analysis, Kurt Gödel showed the incompletability of formal arithmetic. So,
there was a mathematician doing a piece of conceptual work that showed that
arithmetical truth outstripped what could be done in principle by any machine.
This then led, because of the methods employed, into further investigations into
the notion of mechanically computable function. Once, again there was a
prevailing dogma to the effect that virtually anything could be done by a
machine so long as it was combinatorially specified, and so on. But, then the
diagonal methods involved in Gödel's proof allowed him to show that there were
certain things that machines could not do; solving the halting problem for
example, about the machines themselves. All I am suggesting is that the time may
be ripe within the SETI community for a philosophical "fronting up" to
some of the underlying issues about communication, about decodability, about
"contact" and "content". And I would like to suggest that
the search for extraterrestrial intelligence, "SETI", with a sibilant
"s" at the beginning may be a "starter" in that you can go
for detection. But when you start talking about communication with
extraterrestrial intelligences, with a hard "c", "CETI",
then you are WAY out of left field as far as most of my philosophical colleagues
would see you.
Charles Townes:
Alright, let's hear comments on any of the last three speakers. You had an
earlier question?
Floor: I want
only to comment on the definition of signal-to-noise ratio. It seems that in the
systems I have dealt with that noise usually encompasses both the background and
the signal fluctuations. In some situations the background dominates, while in
other situations it's the signal fluctuations that dominate. A good definition
of noise should include both types of fluctuations.
Charles Townes:
Any other comments? Anyone solve the question of how we communicate? I've heard
solution as to how we decide whether the signal is from an intelligent group,
but I have never heard a solution as to what it says. Somebody back here?
|
