A Priori vs A Posteriori
Benjamin J. Tilly
----------------------------------------
This is a repost of a post of mine covering certain topic of interest.
In case you missed it the first time, here it is.
In recent posts to alt.atheism Bill O'Donnell has been bringing up the
old "argument from design". The argument essentially is, "The universe
is too well-designed for life to be chance so it must have a designer,
and that designer is God." Of course the typical responses have come
up, including the ones involving "the Anthropic Principle", by now
there has been considerable discussion, and it has become obvious that
some posters on both sides lack a basic understanding of the underlying
issues, including what a priori and a posteriori probabilities are, how
you apply them, what the Anthropic Principle is, and how it can be used
to actually analyze the claims made by the argument from design.
Therefore I have decided to post a reasonably clear explanation of the
underlying issues, what the Anthropic Principle is, how it applies, and
how it applies to the above argument. I have tried to be reasonably
complete, accurate, and unbiased. I recommend that everyone who is
moderately interested in this try to understand this summary and what
it means. (Bill, this means YOU!!! :-) I have also added in
talk.origins and talk.religion.misc where I think that similar
discussions probably arise every so often. This took a considerable
amount of work to write and I sincerely hope that people try to
honestly understand what I have written...
The story starts with probability theory with the terms "a priori" and
"a posteriori". These terms have to do with how we can draw inferences
about how likely something is based on an observation. The a priori
probabilities are the probabilities that we assign to various
possibilities befre we make our observation. The a posteriori
probabilities are the probabilities that we assign to the same
possibilities after we make our observation. Usually the observation
will affect our judgement (that is, after all, the case that we are
interested in).
The a priori and a posteriori probabilities are related by the concept
of conditional probability, which you can find in any introductory
probability theory or statistics book. If A is the event that we are
trying to make estimates for, and B is the event that we actually
observe, then the a posteriori probability for A is
P(A given B) = P(A and B) / P(B)
where P(A and B) means the a priori probability of A and B both
happening, and P(B) is the a priori probability of B happening. (The
reason for this formula can be found in any standard reference, but I
want to avoid too much discussion of math.) As a corollary of this
formula if we are comparing the likelyhood of two events, A and A', the
ratio of the a posteriori probabilities is the ratio of the a priori
probabilities of A and B versus A and B'.
That is enough abstraction, now let us see how it works in a real
situation. Suppose that you are standing in front of a firing squad
that you know is well-equipped and reasonably capable. You are trying
to estimate the probability that they have been ordered (or bribed) to
give you the scare of your life, that is event A, versus the
probability that they are honestly trying to kill you. They all fire
three times and miss you. You have not a scratch. That is the observed
outcome B. Obviously after going through the ordeal you know that if
they were intending to hit you that what happened is incredibly
unlikely, but if they did not intend to hit you then it is what you
would expect. Therefore you would be inclined towards believing that A
was the case rather than A'. What does the mathematics say? We may know
that a priori A and B both being true had a small probability, but we
would estimate that A' and B happening is even less likely. Therefore
we would believe afterwards that they were not serious about killing
us.
Now this illustrates a number of basic features. One is that the math
captures a fairly common sense idea. A second point is that there is a
direct connection between the a priori and the a posteriori
probabilities. In that connection it is not what the a priori
probabilities are, but rather what the *ratios* of the probabilities
are that matters (more on that later). Thirly notice that no matter how
unlikely the observed event B is, once it is the observed event then in
all discussions of a posteriori probabilities the probability of B
happening is 1. (Mathematicians express probabilities as numbers from 0
to 1. Multiply the probability by a hundred and you get the usual %
probability. Which literally is "per cent", for "per hundred".) And a
last point that I would like to bring up is that our estimates of the a
priori probabilities matter as to what conclusions we draw. This is
generally *not* an issue that should be swept under the carpet, but
rather looked at as honest grounds for a genuine disagreement. Also in
some cases there are no real grounds for assigning probabilities. In
that case any probabilities that anyone assigns should be viewed as
suspect (in a philosophical sense there is reason to ask questions
about the validity of all probability theory arguments, though they are
just too useful to dump), but their argument can (and should) still be
evaluated according to the above principles.
Now let us go to an example where the observation of something unlikely
does not lead the the belief in some sort of interference. Let us
consider a major lottery. Let us say that I am trying to mak
conclusions about whether you are cheating on the lottery (call that A)
versus your being honest (A'). Suppose further that I do not have any
particular reason a priori to believe that you cheated, but you just
won a big jackpot (observation B). The chances that you won by cheating
(A and B) is roughly the probability that the jackpot was won by
cheating (which are fairly small--I hope:-), divided by the number of
people participating (since I know of no particular reason for them to
single you out). The a priori probability that both A and B happen is
about the probability that a random ticket wins times the number of
tickets you have bought. (I say "about" since the chance of cheating
is, I believe, small.) Roughly speaking that should be around 1/(number
of people buying in the lottery). Working out the relevant ratio
P(A and B)/P(A' and B)
I estimate that this should be around the odds that there was cheating
going on, which I think is small. Thus I am not about to jump up and
down and claim that you cheated. (Though I may feel a little bit of
envy. :-)
Let us alter things slightly in the above example...
If you were the son of the person who was in charge of the lottery,
then my a priori (before I observe B that is) estimates of how likely A
and B happening are (that is the event that you win by cheating) would
be substantially higher, and I probably would have some doubts. (That
is why family of people running lotteries are not generally allowed to
participate.)
Alternately if you were to win the lottery 10 times running then the
relevant ratios would change again and I would definitely have my
suspicions. (I will let you look into the math as a good exercise. You
will have to invent some sample figures, but as long as they are
reasonable you should find that 10 wins by the same person should raise
some eyebrows...)
Enough of the probability theory background, what is the anthropic
principle? There are many minor variations of it, however all of them
express the same idea. The one that I will look at here says, "You are
a random intelligent observer." (Publius and Sukru can disregard this
sentence.:-) Think about it for a bit. All that it makes is an
observation that you are an intelligent observer, and then goes on to
suppose that beyond that fact there is no reason to believe that you
are special. (Actually it is more accurate to say that, "You are a
random intelligent observer thinking about the Anthropic principle."
Which explains why you are likely to be better educated and more
interested in philosophical issues than most people...)
What is amazing is how many misconceptions there are about this simple
observation. First of all appeal to it when trying to draw conclusions
is not "begging the question" as many theists would claim. Rather it is
applying the standard methods of trying to work out inferences from an
observation. On the other hand it is not a phrase that "makes questions
about how things got to be this way meaningless" as many atheists would
claim. Again, it is the first step in applying standard methods of
working out inferences. It is also not "devoid of real content". It is
true that it is not an explanation per se for anything, but it is
something that can honestly be used to conclude things.
Here is a non-trivial example. The Moon is quite large in proportion to
the Earth. In the entire Solar System only Pluto has a moon that is
larger in comparison to itself (and some questions have been raised as
to whether that is something that happened more recently), and nothing
else even comes close. Now in a detailed computer analysis it was found
that if the Earth did not have a large moon then axis of the planet
would be unstable on a geological time, making it extremely likely that
over tens of millions of years, let along billions, it would develop an
extreme enough tilt at some point to make it inhospitable for life.
Thus the observation that this is a planet where conditions have been
favorable to life long enough for intelligent life to develop, along
with the fact that this would be substantially less likely if there was
not a large moon, make it significantly more likely that this planet
has a large moon. I do not remember the exact figures, but if we assume
that in 30 million years that there are even odds for a planet with no
large moon to tilt enough to be inhospitable to life, and we assume
that 2 billion years is the time needed of continuous evolution to
develop intelligent life, then a planet with no large moon would have
only around 1 chance in 10^20 of getting through that time without at
some point tilting too severely for life to survive. Given our present
ignorance about the details of the development of solar systems we
cannot really give a figure for the unlikelyhood of such a large moon,
but I would guess that it is probably less unlikely than 10^(-15). (I
used two data points, one is the solar system where the Moon seems out
of the ordinary, the other is the fact that binary stars are quite
common...)
OK, now it is time for the argument from design. The argument is that
the present world/universe is so precisely tuned for us to exist as to
be incredibly unlikely if there was no designer, therefore there must
be a designer. I would like to focus on the argument about the universe
since the arguments involving claims that this planet is special are
similar and it is the universe question that has been coming up in the
discussion.
First of all only half of the argument that needs to be presented has
been attempted here. The side that has been totally ignored is how
likely it is a priori that there is a designer, who is interested in
creating intelligent life, who would actually do it by designing a
universe like this, and who is capable of doing it. Given that the
person presenting the argument believes that there is such a designer
(indeed the individual often believes that they know what the designer
is like) this oversight is understandable. However from the point of
view of an atheist such as myself this is a major flaw since to me the
assumption of an intelligent God capable of creating the universe is a
more unlikely thing a priori than the observed universe, and then even
if there were such a being, assuming that such a being would create the
entire universe for the purpose of creating life on this particular
corner seem very low. The sheer inefficiency of that scheme is hard to
overstate, and such images as buying Yellowstone because you want to
observe the cilia on one bacteria in one pond just do not do justice to
what is being assumed. Thus for me the appropriate ratio that matters
is liable to turn out to be tremendously against my drawing the
conclusion that God exists. Of course this estimate is based on my
biases and your mileage is sure to vary, but I hope that you see that
this half of the issue is certainly neglected by those presenting the
"argument from design".
Now let us move on to a standard response. One is the fact that if the
universe was different then it could still be possible for life to come
into existence, it just would not look like us. This is true. In
standard estimates that are made about the coincidences that lead to us
it is common to forget the fact that another set of circumstances would
lead to another intelligent species that would be just as bewildered as
we are. However the event listed by the anthropic principle does not
specify "humanoid", so we should alter the estimates accordingly for
how unlikely things would be for the random event of interest (that you
are a random intelligent observer). Of course you can argue that we are
not talking about just any old sapient being, but rather humans. Fine,
but if you want to do that then in the probability of a designer you
have to work out the more restrictive probability that the designer
would want humans specifically rather than just intelligent life.
Either way this response does alter the balance against the argument
from design
Another response is that there are cosmological models that would have
this be one of a great number of different universes, possibly with
different constants in different universes. (There are reasons for
believing this, mainly the fact that once you conjecture an actual
mechanism for the universe to come into existence then there is no
reason for that mechanism to only operate once. The part about
different universes having different constants is pure speculation
though.) These models result in an increased chance that somewhere
there are intelligent life forms.
If you want to skip a technical evaluation of this response then please
skip this paragraph. When I went to analyze the effectiveness of this
argument I was in for a suprise. Do you remember that way back in the
probability theory section of this post I mentioned that differing
assumptions lead to different figures? This is one of these cases. The
innocuous phrase in my statment of the anthropic principle "chosen at
random" is not so simple. I came up with three possible meanings, of
which I fairly quickly decided that the one that was most favourable to
this response was implausible, narrowing things down to two. The one
that I consider more plausible of the two is that you are a random
intelligent being chosen out of the intelligent beings in this reality
if that is possible. Thus the probability of your existing at all is
the same as the possibility that there are intelligent beings in this
reality. Therefore the above response, which increases the chances that
without a designer there would be intelligent beings, does weaken the
argument from design. The other scenario is that you are a random
sample from all "things" in this reality and you happen to be
intelligent. With this interpretation the above argument is a total
non-sequitor and makes absolutely _no_ difference.
And a flaw in the argument itself is that in many presentations many of
the probabilities presented as amazing are totally irrelevant. For
example it is not uncommon to see the probability of a whole organism
being spontaneously formed from a bath of the appropriate chemicals
being quoted. Well, if I were to take a patch of land and dig random
channels in it my chance of coming up with something that would work as
a drainage system for a river are pretty small. However that figure is
very different from what the odds are *for* *erosion* *by* *water* to
produce a drainage system. Likewise the appropriate figure for the
creation of life is the probability of life being formed by the
processes that we think were going on. Which is a good deal higher than
the figure that was being quoted before. (See talk.origins for more on
this topic.)
Which brings me to another response to the argument from design. How do
you know that the figures that you give are really variable? For
example one figure that Bill has been quoting is the balance that has
to exist between positive charges and negative charges to keep the
electrostatic forces from overwhelming gravity. However the main models
proposed by scientists, such as the vacuum fluctuation one, all have
the feature that they predict an exact balance of charges thanks to
conservation of charge. Therefore this figure which Bill would claim is
a result of chance does not seem to have the same freedom. Furthermore
some of the theories that are being proposed within the area of
theoretical physics for how gravity and quantum mechanics fit together
have the feature that there are only a finite number of possible sets
of laws for the universe. If so, and it is still if, then there is no
need for a good deal of the "tuning" that Bill thinks is needed. All of
these possibilities mess up the probability predictions involved in the
argument from design, which weakens that argument considerably. However
note that this line of thinking makes the previous possible response
involving the possibility of different universes with different
constants weaker.
In short, the anthropic principle is a fairly natural idea and can be
used to analyze the argument from design. A proper analyses is not as
simple as some people's offhand comments would lead you to believe.
However it does reveal the fact that the standard presentation of the
argument from design totally ignores half of the argument that needs to
be presented (to me its biggest flaw). In addition some of the many
standard responses to the argument do generally show up flaws in the
argument, but in at least one case whether or not it does depends on
underlying assumptions. But overall I hope that it is clear why the
standard argument from design is not likely to convince those who are
not already convinced that God exists.
I certainly hope that this summary has proved interesting for all of
you. (It certainly took long enough for me to write! )-:
Ben Tilly
Brian D Harper
Thanks for this excellent post. Probably the best I've seen on
this topic. I am going to make several comments below which I
hope you will take in the spirit of "let's have some fun
speculating about the Anthropic Principle" ;-).
In article <3itm1j$b...@dartvax.dartmouth.edu>,
Benjamin J. Tilly <Benjamin...@dartmouth.edu> wrote:
>This is for everyone arguing about the argument from design yet AGAIN.
>
>----------------------------------------
[...]
>That is enough abstraction, now let us see how it works in a real
>situation. Suppose that you are standing in front of a firing squad
>that you know is well-equipped and reasonably capable. You are trying
>to estimate the probability that they have been ordered (or bribed) to
>give you the scare of your life, that is event A, versus the
>probability that they are honestly trying to kill you. They all fire
>three times and miss you. You have not a scratch. That is the observed
>outcome B. Obviously after going through the ordeal you know that if
>they were intending to hit you that what happened is incredibly
>unlikely, but if they did not intend to hit you then it is what you
>would expect. Therefore you would be inclined towards believing that A
>was the case rather than A'. What does the mathematics say? We may know
>that a priori A and B both being true had a small probability, but we
>would estimate that A' and B happening is even less likely. Therefore
>we would believe afterwards that they were not serious about killing
>us.
>
Firing squad analogies seem to appear frequently in discussing the
anthropic principle. Let's consider a slightly different scenario.
Suppose that there are, say, 20 sharpshooters lined up and they
really really do intend to kill you. When the order to fire is
given, all 20 rifles misfire ;-).
This scenario is given by George Greenstein to illustrate
"coincidences" which require a Strong Anthropic Principle (SAP)
as opposed to a Weak Anthropic Principle (WAP):
...The resonances between helium, beryllium, and carbon relate
to the very existence of habitable environments in the universe,
and as such they fall under the aegis of the strong Anthropic
Principle. Had those resonances not occurred, no location would
have been habitable--not the Earth and not the Moon, not anywhere
within our Milky Way Galaxy nor within any other galaxy, nor
indeed at any other epoch of cosmic history.
Furthermore, those resonances really are coincidences. They are
really remarkable strokes of luck. As emphasized above, the
Anthropic Principle provides no explanation for anything, and no
amount of anthropic reasoning can explain these coincidences away.
The point can be seen most vividly with the help of Leslie's
analogy, mentioned in the Prologue, of the firing squad. A court
has sentenced a man to be shot at sunrise. Early this morning he
was dragged before the firing squad. He stood before it blindfolded,
the commander gave the order to shoot ... and every last one of
the rifles misfired.
Had even one functioned properly he would not be alive today.
The bare fact of his existence therefore proves they must have
all misfired. That's the strong Anthropic Principle in operation.
But clearly it is not enough. It is not enough to explain why so
very many coincidences all occurred at once.
--George Greenstein, _The Symbiotic Universe_, William Morrow
and Company, 1988, pp. 58-60.
I'll discuss the difference between WAP and SAP more below.
[...]
>
>Enough of the probability theory background, what is the anthropic
>principle? There are many minor variations of it, however all of them
>express the same idea. The one that I will look at here says, "You are
>a random intelligent observer." (Publius and Sukru can disregard this
>sentence.:-) Think about it for a bit. All that it makes is an
>observation that you are an intelligent observer, and then goes on to
>suppose that beyond that fact there is no reason to believe that you
>are special. (Actually it is more accurate to say that, "You are a
>random intelligent observer thinking about the Anthropic principle."
>Which explains why you are likely to be better educated and more
>interested in philosophical issues than most people...)
>
Don't you mean that there are minor variations in how the
Weak Anthropic Principle is stated? I kind of doubt that
many atheists would be comfortable with a Strong Anthropic
Principle, remember Martin Gardner's Completeley Ridiculus
Anthropic Principle (CRAP)? ;-)
As I understand it, the main reason for delineating a weak
form of the principle was to capture the essential ingredients
of anthropic reasoning which are obvious and non-controversial.
I don't believe I've ever seen the WAP stated as you have put
it above. I prefer something like: "We expect to find ourselves
in an environment conducive to our existence". I think this is
better if we want to make sure we are talking WAP instead of SAP.
For example, later you discuss the "coincidence" of the relative
proportions of the Moon and Earth. This is a good of example of
where WAP comes to play. But, I would argue here that this is
special type of "coincidence" and that one has to be careful
generalizing the argument from this case to other types of
"coincidencs". This is the point that George Greenstein was
trying to make in the quote above. Let me try to illustrate
in terms of another "coincidence" which also falls under the
WAP. In the Feb. 9 1995 issue of _Nature_ there is an interesting
article about Jupiter:
"How Special is Jupiter?", G.W. Wetherill, _Nature_,
vol. 373, p. 470, 1995.
The results are tentative, of course, but they suggest that
intelligent life may require a Solar System that not only
contains something like our Earth/Moon but also a Jupiter in
just the right place. This takes on "anthropic implications" since
many scientists think that Jupiters are very rare in the Universe.
Again, this is tentative but let's suppose the idea pans out.
What does the WAP have to say about this "coincidence"? IMHO,
something like this: "If Jupiters are required for intelligent
life, then we should certainly not be too surprised to find
ourselves living in a Solar System that has a Jupiter".
Now, why is this a special case with respect to "coincidences"
in general? The WAP has to do with "coincidences" regarding
why we happen to be in one part of the Universe as opposed to
some other part. We are not in a Solar System that has no Jupiter
because life cannot exist in that type of Solar System. Now, at
what point would we move from WAP to SAP in our anthropic reasoning?
The basic idea of a SAP is to say that life *must* exist somewhere,
somehow. That the very existence of the Universe *requires* the
existence of intelligent observers. This is not obvious, nor is
it non-controversial ;). This is the distinction that Greenstein
was making for the "coincidence" involving the resonances between
helium, beryllium, and carbon. If these resonances did not occur,
there would be no life anywhere and at any time in the history
of the Universe. This doesn't mean we cannot say something like
"we know those resonances must have occured since we are here",
just that this observation alone doesn't explain the "coincidence"
as to why the resonances occured to begin with. In case the
distinction is not clear, let's return to the Jupiter case.
WAP explains for us the "coincidence" as to why we find ourselves
in a Solar System that contains a Jupiter as opposed to one
that does not. WAP does not explain any "coincidences" that
might be associated with why there are Jupiters at all.
[...]
>
>Which brings me to another response to the argument from design. How do
>you know that the figures that you give are really variable? For
>example one figure that Bill has been quoting is the balance that has
>to exist between positive charges and negative charges to keep the
>electrostatic forces from overwhelming gravity. However the main models
>proposed by scientists, such as the vacuum fluctuation one, all have
>the feature that they predict an exact balance of charges thanks to
>conservation of charge. Therefore this figure which Bill would claim is
>a result of chance does not seem to have the same freedom. Furthermore
>some of the theories that are being proposed within the area of
>theoretical physics for how gravity and quantum mechanics fit together
>have the feature that there are only a finite number of possible sets
>of laws for the universe. If so, and it is still if, then there is no
>need for a good deal of the "tuning" that Bill thinks is needed. All of
>these possibilities mess up the probability predictions involved in the
>argument from design, which weakens that argument considerably. However
>note that this line of thinking makes the previous possible response
>involving the possibility of different universes with different
>constants weaker.
>
Yes, these are very good points. Question: does the development
of a theory from which certain constants (previously thought to
be purely "coincidental") are obtained naturally really do away
with the Anthropic Principle? IMHO, no. It only moves it to a
different level. Now one has to explain the "coincidence" that
one has this particular law from which these particular constants
are derived and not some other law. Believe it or not, I'm not
making this up ;-), consider, for example:
This simple but powerful argument leads us to ask the question:
*how much* can I deduce about the universe from the fact that
I exist? It turns out that various elementary particle, nuclear,
atomic and molecular properties of matter have to be very finely
tuned for conditions in the universe to have permitted my
development--many examples are given by Barrow and Tipler and
elsewhere in this book. These finely tuned properties will
probably also eventually be accounted for by fundamental
theory. But why should fundamental theory _happen_ to lead
to these properties?
There seem to be three possible answers to this question:
(a) It is a complete chance.
(b) God regards me as such a desirable product of the universe
that he has fine-tuned it so as to guarantee my development.
(c) There exist other, disjoint, universes with other laws
and constants of nature.
-- D.W. Sciama, "The Anthropic Principle and the Non-Uniqueness
of the Universe," in _The Anthropic Principle_, Proceedings
of the Second Venice Conference on Cosmology and Philosophy,
F. Bertola and U. Curi, ed., Cambridge University Press,
1993, pp. 107-109.
Well, my main reason for including this quote was the question
"But why should fundamental theory _happen_ to lead to these
properties?", but I couldn't resist including his "possible
answers". I still can't believe he said (b). An argument to
"design" is one thing, a personal, caring, transcendent Creator
God is something else entirely ;-). Oh, BTW, the author opts
for answer (c) ;-).
>In short, the anthropic principle is a fairly natural idea and can be
>used to analyze the argument from design. A proper analyses is not as
>simple as some people's offhand comments would lead you to believe.
>However it does reveal the fact that the standard presentation of the
>argument from design totally ignores half of the argument that needs to
>be presented (to me its biggest flaw). In addition some of the many
>standard responses to the argument do generally show up flaws in the
>argument, but in at least one case whether or not it does depends on
>underlying assumptions. But overall I hope that it is clear why the
>standard argument from design is not likely to convince those who are
>not already convinced that God exists.
>
I agree wholeheartedly with your last sentence.
>I certainly hope that this summary has proved interesting for all of
>you. (It certainly took long enough for me to write! )-:
>
Yes, thanks again.
>Ben Tilly
--
Brian Harper bha...@magnus.acs.ohio-state.edu
Associate Professor |
Applied Mechanics | "It is not certain that all is uncertain,
Ohio State University | to the glory of skepticism" -- Pascal
Johmann
>OK, now it is time for the argument from design. The argument is that
>the present world/universe is so precisely tuned for us to exist as to
>be incredibly unlikely if there was no designer, therefore there must
>be a designer. I would like to focus on the argument about the universe
^^^^^^^^
In other words, and as you do elsewhere in your very lengthy post, you
mean the God of the Bible (gag). So, why limit yourself to this strawman?
Also, you are assuming that creation -- if there is such a thing, and you
say there isn't -- must include the universe (in keeping with the God
approach). As far as I'm concerned, the only issue is the creation of
organic life, not the universe. I don't see the universe as being
created, but I do see organic life as something that was created within
that universe (just as we create things, by using our intelligence).
>And a flaw in the argument itself is that in many presentations many of
>the probabilities presented as amazing are totally irrelevant. For
>example it is not uncommon to see the probability of a whole organism
>being spontaneously formed from a bath of the appropriate chemicals
>being quoted. Well, if I were to take a patch of land and dig random
I haven't seen any, unless by "whole organism" you mean a single cell
(and I haven't seen any for a single cell either, but I have seen many
probability analyses for known organic molecules, such as DNA -- I've
done one myself, see my sig for details).
Kurt Johmann
--
_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/
The Computer Inside You is a new book that proposes in detail an old
idea: that the universe is a virtual reality generated by an underlying
network of computing elements. In particular, the book uses this reality
model to explain the currently unexplained: ESP, afterlife, mind, UFOs
and their occupants, organic development, and such.
WWW ==> http://www.xmission.com/~gastown/the-computer-inside-you
FTP ==> ftp.xmission.com/pub/users/g/gastown/outgoing/t-c-i-y/tciy.txt
_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/