About the Shannon's Formula
grapheus
Shannon's original work by lack of mathematical knowledge, here is a
Summary of it (that J.Faucounau, another mathematician, sent to me a
few years ago).
NOTA : IGNARROGANTS are asked to ignore this thread... Anyway, they
will not understand it !...
The mathematician Claude Shannon was mainly interested in the
EFFICIENCY of a Cryptographic Code. He started from the obvious remark
that longer is a text, and more probable is its decipherment. To be
sure that a cryptogram will not be deciphered, one has then to limit
its length under a certain "number of signs", that he called the
"Unicity Distance".
This Unicity Distance is defined as follows : Supposing that the
ennemy knows (or has guessed) the TYPE of CODE and the LANGUAGE used,
the length of the text is equal with the U.D. if it is JUST ENOUGH to
allow him to reach a UNIQUE solution. If the length is LESS than the
U.D., the ennemy cannot reach a UNIQUE solution, even starting from
the correct hypotheses. If the length is greater than the U.D., he
theorically *MAY* break the cryptogram. This does'n't mean that he
will !.. Because he may spend months or years, having chosen a wrong
type of code or a wrong language!... But the POSSIBILITY does exist
for him to succeed in finding the correct solution... Moreover, if the
text length is LARGELY GREATER that the U.D., the ennemy may
INTERNALLY check the validity of his solution...
Claude Shannon has shown that the U.D. depends upon two things :
- MAINLY upon the "Type of Code". Some "types of Code" are a lot
easier to decipher than others. So that the U.D. may vary from c. 50
signs to several billions of signs, following the "Type of Code" which
is used... And, of course, Claude Shannon has given a list of the
"best modern codes" ...
- SECONDLY upon the language and the text itself. This part of the
Shannon's work is the most difficult to understand by the layman. So,
it suffices to say that, if some languages and texts are easier to
decipher than others, the difference is unimportant upon the U.D.,
compared to the influence of the "Type of Code" (One has generally to
apply a coefficient varying from c.0.5 to 2).
I hope having clarified a bit the question for people "just curious"
to know (but surely not for the IGNARROGANTS, who believe thet know
everything without opening the file!).
grapheus
Bobby D. Bryant
> For readers in this Group, who are not mathematicians and can't read
> Shannon's original work by lack of mathematical knowledge, here is a
> Summary of it (that J.Faucounau, another mathematician, sent to me a
> few years ago).
You forgot to tell us whether the "code" used in the PD is a "random
cipher", which is a prerequisite for Shannon's entire analysis.
And if you decide to answer, do tell us how you know.
--
Bobby Bryant
Austin, Texas
Jacques Guy
Philip Deitiker
news:pan.2004.01.27....@mail.utexas.edu:
The answer is Yes, because I randomly guessed it would be, maybe.
--
DNApaleoAnth at Att dot net
Dik T. Winter
> The mathematician Claude Shannon was mainly interested in the
> EFFICIENCY of a Cryptographic Code. He started from the obvious remark
> that longer is a text, and more probable is its decipherment. To be
> sure that a cryptogram will not be deciphered, one has then to limit
> its length under a certain "number of signs", that he called the
> "Unicity Distance".
He will not have used the words "to be sure". At most "to be fairly certain".
> This Unicity Distance is defined as follows : Supposing that the
> ennemy knows (or has guessed) the TYPE of CODE and the LANGUAGE used,
> the length of the text is equal with the U.D. if it is JUST ENOUGH to
> allow him to reach a UNIQUE solution.
I do not think this is the definition because there are two different
interpretations possible:
1. The smallest length such that all longer or equally long texts code
different,
or
2. The smallest length such that for all shorter texts there exists a
text such that there codings are the same.
I think he means (2).
> If the length is LESS than the
> U.D., the ennemy cannot reach a UNIQUE solution, even starting from
> the correct hypotheses.
That is, there are different source texts that give the same coding.
This does *not* mean that the enemy can not make an educated guess
which of the possible source texts is the correct one, because most
of the possible source texts will actually be fairly nonsensical.
But I see now what you mean by "correct hypothesis", i.e. the pair
"type of code" and "language".
> But the POSSIBILITY does exist
> for him to succeed in finding the correct solution...
The strange thing about this all is that if you start with an incorrect
hypothesis you might even get a sensible solution, but it will in
general be incorrect.
> Moreover, if the
> text length is LARGELY GREATER that the U.D., the ennemy may
> INTERNALLY check the validity of his solution...
Nope. You can not check the validity. You can check the validity only
when you assume your initial hypothesis to be correct. You can *not*
check your initial hypothesis, because a different language and coding
may also result in a valid source text.
The unicity distance is only for the same coding and the same language.
When you allow different codings and different languages the unicity
distance is much larger. So *only* when you have good reasons that
your initial hypothesis are correct the unicity distance is relevant.
But there is *no* way to "prove" that your initial hypothesis is indeed
correct. (I use "prove" in the mathematical sense here.)
> Claude Shannon has shown that the U.D. depends upon two things :
> - MAINLY upon the "Type of Code". Some "types of Code" are a lot
> easier to decipher than others. So that the U.D. may vary from c. 50
> signs to several billions of signs, following the "Type of Code" which
> is used... And, of course, Claude Shannon has given a list of the
> "best modern codes" ...
Granted.
> - SECONDLY upon the language and the text itself. This part of the
> Shannon's work is the most difficult to understand by the layman. So,
> it suffices to say that, if some languages and texts are easier to
> decipher than others, the difference is unimportant upon the U.D.,
> compared to the influence of the "Type of Code" (One has generally to
> apply a coefficient varying from c.0.5 to 2).
I would even be so bold to say that the unicity distance would not
change much if we allow different languages, as long as those languages
are all known.
The major problem comes when you throw in unknown languages.
But I think that with the Phaestos Disk the complete discussion about
unicity distance is irrelevant (as it is with all decypherments of
stuff written in an unknown script in a possibly unknown language).
Shannon is interesting, but only in cryptography. In decypherment
of stuff in an unknown script in a possibly unknown language the unicity
distance would be extremely large.
--
dik t. winter, cwi, kruislaan 413, 1098 sj amsterdam, nederland, +31205924131
home: bovenover 215, 1025 jn amsterdam, nederland; http://www.cwi.nl/~dik/
Peter T. Daniels
>
> For readers in this Group, who are not mathematicians and can't read
> Shannon's original work by lack of mathematical knowledge, here is a
> Summary of it (that J.Faucounau, another mathematician, sent to me a
> few years ago).
So now there are _four_ P.D. threads.
In other words, if you guess that the P.D. is in Greek, you can decipher
it as Greek.
--
Peter T. Daniels gram...@att.net
Jacques Guy
> In article <337ae51f.04012...@posting.google.com> grap...@www.com (grapheus) writes:
> > This Unicity Distance is defined as follows : Supposing that the
> > ennemy knows (or has guessed) the TYPE of CODE and the LANGUAGE used,
> > the length of the text is equal with the U.D. if it is JUST ENOUGH to
> > allow him to reach a UNIQUE solution.
> I do not think this is the definition
Doesn't matter, it's good enough for all practical purposes.
"Supposing that the enemy knows (or has guessed) the language".
John Chadwick had pretended that it was Medieval Japanese, and
he had deciphered it as a legal text on water rights.
Faucounau has guessed that the language was Proto-Ionic (which
he made up).
Padre Maricón y Cojones has guessed that it was in Quechua and
was about llama-shearing (pers.com.)
Noted Greek philosopher Polykrates of Smyrna had guessed that
it was in Amazonian and was a text of mastectomy surgery. Alas, none
of his works have survived.
Mr or Mrs Xqblwykpq, on a visit here from planet Qolob in galaxy
Feefoefiefum
told me that ukduieo pd owkjd sisuiufhe yhshe ioo ehnasovmc smeiqtyqspe.
I do not think that he, or she, or it, or ...(???) has got it right.
grapheus
> In article <337ae51f.04012...@posting.google.com> grap...@www.com (grapheus) writes:
> > The mathematician Claude Shannon was mainly interested in the
> > EFFICIENCY of a Cryptographic Code. He started from the obvious remark
> > that longer is a text, and more probable is its decipherment. To be
> > sure that a cryptogram will not be deciphered, one has then to limit
> > its length under a certain "number of signs", that he called the
> > "Unicity Distance".
>
> He will not have used the words "to be sure". At most "to be fairly certain".
>
If you want... There are, of course, very short texts that can be - by
chance or because of certain characteristics - deciphered, although
their length is less that the U.D. In one of his papers, J.Faucounau
has quoted one...
> > This Unicity Distance is defined as follows : Supposing that the
> > ennemy knows (or has guessed) the TYPE of CODE and the LANGUAGE used,
> > the length of the text is equal with the U.D. if it is JUST ENOUGH to
> > allow him to reach a UNIQUE solution.
>
> I do not think this is the definition because there are two different
> interpretations possible:
> 1. The smallest length such that all longer or equally long texts code
> different,
> or
> 2. The smallest length such that for all shorter texts there exists a
> text such that there codings are the same.
> I think he means (2).
>
NO. You misunderstood Shannon's work !..
Remember : he was interested in the SECURITY brought by the cipherment
!..
> > If the length is LESS than the
> > U.D., the ennemy cannot reach a UNIQUE solution, even starting from
> > the correct hypotheses.
>
> That is, there are different source texts that give the same coding.
> This does *not* mean that the enemy can not make an educated guess
> which of the possible source texts is the correct one, because most
> of the possible source texts will actually be fairly nonsensical.
>
> But I see now what you mean by "correct hypothesis", i.e. the pair
> "type of code" and "language".
>
YES !.. Add also : the TYPE of the TEXT...
> > But the POSSIBILITY does exist
> > for him to succeed in finding the correct solution...
>
> The strange thing about this all is that if you start with an incorrect
> hypothesis you might even get a sensible solution, but it will in
> general be incorrect.
>
> > Moreover, if the
> > text length is LARGELY GREATER that the U.D., the ennemy may
> > INTERNALLY check the validity of his solution...
>
> Nope. You can not check the validity.
Of course, you can !... If the text is 50 times longer than the U.D.,
one may decipher it using only HALF of it, then VERIFY the VALIDITY
with the second half !.
>
> The unicity distance is only for the same coding and the same language.
OF COURSE !... Are you rediscovering the wheel?..
> When you allow different codings and different languages the unicity
> distance is much larger. So *only* when you have good reasons that
> your initial hypothesis are correct the unicity distance is relevant.
> But there is *no* way to "prove" that your initial hypothesis is indeed
> correct. (I use "prove" in the mathematical sense here.)
You are MIXING UP the "Unicity Distance" which is DEFINED by the
*CIPHERER*, and the "ESTIMATED U.D." , which is EVALUATED BY THE
*DECIPHERER* !...
>
> > Claude Shannon has shown that the U.D. depends upon two things :
> > - MAINLY upon the "Type of Code". Some "types of Code" are a lot
> > easier to decipher than others. So that the U.D. may vary from c. 50
> > signs to several billions of signs, following the "Type of Code" which
> > is used... And, of course, Claude Shannon has given a list of the
> > "best modern codes" ...
>
> Granted.
>
> > - SECONDLY upon the language and the text itself. This part of the
> > Shannon's work is the most difficult to understand by the layman. So,
> > it suffices to say that, if some languages and texts are easier to
> > decipher than others, the difference is unimportant upon the U.D.,
> > compared to the influence of the "Type of Code" (One has generally to
> > apply a coefficient varying from c.0.5 to 2).
>
> I would even be so bold to say that the unicity distance would not
> change much if we allow different languages, as long as those languages
> are all known.
>
> The major problem comes when you throw in unknown languages.
>
NO, if they are NOT TRULY "UNKNOWN"!.. This was the case with Linear
B. It's also the case with the Phaistos Disk, written in
proto-ionic... Finding the Language is just a question of patience,
intuition and sound statistical considerations.
> But I think that with the Phaestos Disk
Don't call it the "Phaestos Disk" !.. Phaistos is written in Linear A
and B : PA-I-TO !!!
> the complete discussion about
> unicity distance is irrelevant (as it is with all decypherments of
> stuff written in an unknown script in a possibly unknown language).
To decipher it ?.. SURELY !.. But to bring a scientific contradiction
to all the non-mathematician kooks (John Chadwick included!) who wrote
that "it is undecipherable", NO !..
> Shannon is interesting, but only in cryptography. In decypherment
> of stuff in an unknown script in a possibly unknown language the unicity
> distance would be extremely large.
NO !.. The "ESTIMATE of the U.D." made with IMPLAUSIBLE HYPOTHESES
can !.. That's all !..
But a "reasonable and most plausible" U.D. estimate leads to the
conclusion that the PHAISTOS DISK MAY BE DECIPHERED !...
Of course, IGNARROGANTS are unable to understand that !... They
believe they know everything better than anybody !...
grapheus
grapheus
> Chante, Vraiconnot, t'auras une cage.
Here comes the CLOWN in !..
Please, A BIG APPLAUSE, folks !
grapheus
grapheus
> Dik T. Winter wrote:
>
> > In article <337ae51f.04012...@posting.google.com> grap...@www.com (grapheus) writes:
>
> > > This Unicity Distance is defined as follows : Supposing that the
> > > ennemy knows (or has guessed) the TYPE of CODE and the LANGUAGE used,
> > > the length of the text is equal with the U.D. if it is JUST ENOUGH to
> > > allow him to reach a UNIQUE solution.
>
> > I do not think this is the definition
>
> Doesn't matter, it's good enough for all practical purposes.
> "Supposing that the enemy knows (or has guessed) the language".
>
> John Chadwick had pretended that it was Medieval Japanese, and
> he had deciphered it as a legal text on water rights.
>
> Faucounau has guessed that the language was Proto-Ionic (which
> he made up).
>
> Padre Maricón y Cojones has guessed that it was in Quechua and
> was about llama-shearing (pers.com.)
>
YES !.. And more than 40 other authors have presented other
solutions...
The DIFFERENCE is that ONLY the PROTO-IONIC SOLUTION has been PROVED.
But IGNARROGANTS want to IGNORE the word PROOFS... They HATE this
word !... They prefer laughing at the most stupid and implausible
solutions, or semantically arguing about the word PROOF, than
seriously studying the existing PROOFS!...
grapheus
grapheus
YES !.. And after that, searching for PROOFS to VERIFY this solution
!..
But IGNARROGANTS want to IGNORE the word PROOFS !.. They HATE this
word !.. They prefer semantically and endlessly arguing about it, or
make fun of the UNPROVEN and RIDICULOUS solutions than SERIOUSLY
STUDYING the PROOFS concerning the ONLY PROVED SOLUTION, the
"Proto-Ionic" one !.... IGNARROGANTS are like that, and nobody can
change their attitude. They are so intelligent... or at least they
believe so!...
grapheus
Peter T. Daniels
So you finally admit that the "decipherment" is arbitrary.
Kindly show one of the PROOFS that the language is Greek and not
anything else.
Jim Webster
"grapheus" <grap...@www.com> wrote in message
> But IGNARROGANTS want to IGNORE the word PROOFS !.. They HATE this
> word !..
I think you will have to accept that other people have a higher standard of
proof than you do,
Insulting them is unlikely to either change their minds or convince other
readers that you are worth bothering with
Jim Webster
Wolf Kirchmeir
[quoting PTD:}
>> In other words, if you guess that the P.D. is in Greek, you can decipher
>> it as Greek.
>
>YES !.. And after that, searching for PROOFS to VERIFY this solution
>!..
Supporting evidence is not a proof. How many times do you have to be told
that. It's just - supporting evidence. It's quite possible for the same facts
to support divergent claims. That's why it's not a PROOF, still less a proof.
That doesn't that mean I claim that the PROOFS don't support the
decipherment. It's just that your CONTINUAL YELLING doesn't convince me,
doesn't persuade me, doesn't even tickle my curiosity.
Tough shit, as they say.
--
Wolf Kirchmeir, Blind River ON Canada
"Nature does not deal in rewards or punishments, but only in consequences."
(Robert Ingersoll)
grapheus
ALWAYS the same CONFUSION of yours between an HYPOTHESIS WITH A HIGH
PROBABILITY to be true, and an Hypothesis with LOW or INEXISTENT
probability to be cotrrect !...
From my Dictionary : "arbitrary" : "determined just by caprice".
"Plausible" : "worthy of belief".
Of course, for you, BOTH NOTIONS ARE IDENTICAL !...
IGNARROGANTS hate the words PROOFS and PROBABILITY !...
grapheus
grapheus
> "grapheus" <grap...@www.com> wrote in message
> > But IGNARROGANTS want to IGNORE the word PROOFS !.. They HATE this
> > word !..
>
> I think you will have to accept that other people have a higher standard of
> proof than you do,
I would accept that IF those "other people" KNEW about the "standard"
of the PROOFS I am talking about !...
As long as they IGNORE THEM, it JUST ARROGANCE of their part to think
that those PROOFS are of a "low standard" !...
>
> Insulting them is unlikely to either change their minds
>
I know that, and I really don't care !... IGNARROGANTS NEVER CHANGE
THEIR MINDS !.. They ALWAYS believe that they know better without
opening the file !...
grapheus
Jim Webster
> > Insulting them is unlikely to either change their minds
> >
>
> I know that, and I really don't care !... IGNARROGANTS NEVER CHANGE
> THEIR MINDS !.. They ALWAYS believe that they know better without
> opening the file !...
you can rest assured that they almost certainly regard you in the same
light, and it does your reputation no good whatsoever
Jim Webster
grapheus
> grapheus wrote:
> >
by the way, I forgot to answer your last question :
>
> Kindly show one of the PROOFS that the language is Greek and not
> anything else.
Your question shows that you don't understand neither the meaning of
the words "Probabilistic Method of deciphering", nor the meaning of
the word PROOFS. The CHOICE of a Greek language in a first step, of a
particular GREEK DIALECT in a second step, has been done by the
decipherer because of PROBABILISTIC considerations, as he has
explained in his books and papers. In other words, from the
calculations conducted by him, this particular solution was found to
be THE MOST PROBABLE...
It is then obvious that this choice was neither "arbitrary", nor 100%
sure, and anyone who knows what the word "HIGH PROBABILITY" means -
therefore NOT an IGNARROGANT !- can easily understand that !..
Once the decipherment achieved, this choice has been shown to be true
by the PROOF that the decipherment was correct indeed.. Anyone who
knows what the word "SCIENTIFIC PROOF" means and has examined ALL the
pieces of evidence - therefore NOT an IGNARROGANT !- has no trouble to
understand that if the whole decipherment is VERIFIED, the language
HAS TO BE PROTO-IONIC!...
grapheus
Peter T. Daniels
Given what was known at the time, it was highly improbable that Linear B
was Greek. For decades, they tried THE MOST PROBABLE, Etruscan, and got
nowhere.
grapheus
> On 28 Jan 2004 04:56:42 -0800, grapheus wrote:
>
> [quoting PTD:}
> >> In other words, if you guess that the P.D. is in Greek, you can decipher
> >> it as Greek.
> >
> >YES !.. And after that, searching for PROOFS to VERIFY this solution
> >!..
>
> Supporting evidence is not a proof.
This is what you think !..
> It's just - supporting evidence. It's quite possible for the same facts
> to support divergent claims.
RIDICULOUS !... HOW do you know that ?.. YOU IGNORE the PROOFS !.. Oh,
sorry ! the "supporting evidence" !..
grapheus
Wolf Kirchmeir
I know that because I know something about the logic of proofs. (You seem to
use the word PROOF to mean "supporting evidence." That's not what "proof"
means in English. I don't think that's what it means in French either, but I
don't know - I'm not fluent in French.)
Once more: To test the validity of an argument or proof, I need only to be
ably to analyse its logical structure. Supporting evidence never proves a
claim, it only _supports_ a claim. That's why in criminal law, a case must
be proved "beyond a reasonable doubt", and in science, data are presented
with probable value ranges attached, and so on. When one is satisfied that an
argument is valid, then it's worthwhile to test its soundness. At that point,
knowledge of its content becomes relevant, but not before.
I repeat: your posts have done nothing to persuade me that the decipherment
you refer to is a valid one, so I'm not going to waste my time swotting up
the facts so I can judge its soundness. That would be a waste of time. I am
interested in the PD, of course, but merely a an onlooker.
Dik T. Winter
> "Dik T. Winter" <Dik.W...@cwi.nl> wrote in message news:<Hs6Fr...@cwi.nl>...
> > In article <337ae51f.04012...@posting.google.com> grap...@www.com (grapheus) writes:
...
> > > This Unicity Distance is defined as follows : Supposing that the
> > > ennemy knows (or has guessed) the TYPE of CODE and the LANGUAGE used,
> > > the length of the text is equal with the U.D. if it is JUST ENOUGH to
> > > allow him to reach a UNIQUE solution.
> >
> > I do not think this is the definition because there are two different
> > interpretations possible:
> > 1. The smallest length such that all longer or equally long texts code
> > different,
> > or
> > 2. The smallest length such that for all shorter texts there exists a
> > text such that there codings are the same.
> > I think he means (2).
> >
>
> NO. You misunderstood Shannon's work !..
> Remember : he was interested in the SECURITY brought by the cipherment
> !..
Yes? Your point being? I still think (2) is meant, because that gives
the greatest security. And I think Shannon has it given a valid
definition rather than what your paraphrase says.
> > But I see now what you mean by "correct hypothesis", i.e. the pair
> > "type of code" and "language".
>
> YES !.. Add also : the TYPE of the TEXT...
Did Shannon really use "type of text"? I doubt it. You can not
mathematically quantify on type of text.
> > > Moreover, if the
> > > text length is LARGELY GREATER that the U.D., the ennemy may
> > > INTERNALLY check the validity of his solution...
> >
> > Nope. You can not check the validity.
>
> Of course, you can !... If the text is 50 times longer than the U.D.,
> one may decipher it using only HALF of it, then VERIFY the VALIDITY
> with the second half !.
Wrong. If you start with some hypothesis (coding method and source
language) and get out something sensible, even when the text is
50 times larger than the U.D. for that particular coding method and
source language, doing decoding on one half and seeing that the
other half comes out just as well only increases the probability
that you are right with your hypothesis. You will never be certain.
Unless the text is larger than the U.D. of the conglomerate of all
coding methods and all possible source languages. But I think in
most cases that U.D. is fairly large.
> > The unicity distance is only for the same coding and the same language.
>
> OF COURSE !... Are you rediscovering the wheel?..
Oh, you can also define it for a conglomerate, mind. You do not even
have to reword the definition of U.D., only the definition of coding
and language.
> > When you allow different codings and different languages the unicity
> > distance is much larger. So *only* when you have good reasons that
> > your initial hypothesis are correct the unicity distance is relevant.
> > But there is *no* way to "prove" that your initial hypothesis is indeed
> > correct. (I use "prove" in the mathematical sense here.)
>
> You are MIXING UP the "Unicity Distance" which is DEFINED by the
> *CIPHERER*, and the "ESTIMATED U.D." , which is EVALUATED BY THE
> *DECIPHERER* !...
In this case the only thing we are concerned with is the estimated U.D.
The point is, we are talking about a script and a language, there is
no reason to suspect that the writer had encryption in mind, so the
U.D. of the original would be extremely low. I would be pretty surprised
if there is any written language with a U.D. larger than, say, 10, and
some come as low as 2 or 3, or perhaps 1 (which is the ideal figure for
any script). Take a mythical language, which does not inflect. Assume
each word is represented by a unique symbol. The U.D. of that script is 1.
You will nevertheless have a hard time to decipher even a longer text,
and to prove that you are right is impossible.
You are too much basing yourself on Shannon, where the purpose is to hide
content. In contrast, with a script the purpose is to reveal content.
The original U.D. has *nothing* to do with the problem of deciphering.
> > The major problem comes when you throw in unknown languages.
>
> NO, if they are NOT TRULY "UNKNOWN"!.. This was the case with Linear
> B. It's also the case with the Phaistos Disk, written in
> proto-ionic... Finding the Language is just a question of patience,
> intuition and sound statistical considerations.
I doubt the soundness when only 47 symbols, or somesuch, are known.
> > But I think that with the Phaestos Disk
>
> Don't call it the "Phaestos Disk" !.. Phaistos is written in Linear A
> and B : PA-I-TO !!!
Scusi.
> > Shannon is interesting, but only in cryptography. In decypherment
> > of stuff in an unknown script in a possibly unknown language the unicity
> > distance would be extremely large.
>
> NO !.. The "ESTIMATE of the U.D." made with IMPLAUSIBLE HYPOTHESES
> can !.. That's all !..
> But a "reasonable and most plausible" U.D. estimate leads to the
> conclusion that the PHAISTOS DISK MAY BE DECIPHERED !...
How do you define a reasonable and most plausible U.D. estimate? I do not
understand that. And don't tell me that Shannon gives a definition...
> Of course, IGNARROGANTS are unable to understand that !... They
> believe they know everything better than anybody !...
Oh, apparently I am an ignarrogant. Well, so be it.
Jacques Guy
> Take a mythical language, which does not inflect. Assume
> each word is represented by a unique symbol. The U.D. of that script is 1.
Actually, zero: it is not a cypher.
This notion of unicity distance/point, and the formula, were
elaborated to answer the question: "how much must we have
deciphered of this _ciphertext_ to be pretty sure that we
have cracked it?"
You can consider the Ph.D. to be a ciphertext
(1) If you know what language it is in (as distinct from making it
up as you go), and you try to figure out the phonetic values
of the 45 signs. I did write "as distinct from making it up
as you go", that makes quite a difference.
or:
(2) If you posit that each word of "Phaistian" corresponds to one
word of English, and that a word-for-word translation is valid.
In this case, you can consider that the Ph.D. is an English
text enciphered word by word. Funny English perhaps but grant
that. As I calculated quickly in another recent post, in that
case the U.D. is greater than 1000. That means that you must have
deciphered 1000 words of Phaistian into English before you
can be confident that your Phaistian-English dictionary
(the cipher) is right. If Phaistian is an alphabet, that
require at least 5,000 signs. If it is a syllabary that requires
at least 2,500 signs. Even if it were logographic, you wouldn't
have one quarter of what you need, since the disk contains only
242 signs when you need more than 1000.
grapheus
> In article <337ae51f.0401...@posting.google.com> grap...@www.com (grapheus) writes:
> > "Dik T. Winter" <Dik.W...@cwi.nl> wrote in message news:<Hs6Fr...@cwi.nl>...
> > > In article <337ae51f.04012...@posting.google.com> grap...@www.com (grapheus) writes:
> ...
> > > > This Unicity Distance is defined as follows : Supposing that the
> > > > ennemy knows (or has guessed) the TYPE of CODE and the LANGUAGE used,
> > > > the length of the text is equal with the U.D. if it is JUST ENOUGH to
> > > > allow him to reach a UNIQUE solution.
> > >
> > > I do not think this is the definition because there are two different
> > > interpretations possible:
> > > 1. The smallest length such that all longer or equally long texts code
> > > different,
> > > or
> > > 2. The smallest length such that for all shorter texts there exists a
> > > text such that there codings are the same. typo:
> >
> > NO. You misunderstood Shannon's work !..
> > Remember : he was interested in the SECURITY brought by the cipherment
> > !..
>
> Yes? Your point being? I still think (2) is meant, because that gives
> the greatest security. And I think Shannon has it given a valid
> definition rather than what your paraphrase says.
BALONEY !.. See hereafter...
> > > > Moreover, if the
> > > > text length is LARGELY GREATER that the U.D., the ennemy may
> > > > INTERNALLY check the validity of his solution...
> > >
> > > Nope. You can not check the validity.
> >
> > Of course, you can !... If the text is 50 times longer than the U.D.,
> > one may decipher it using only HALF of it, then VERIFY the VALIDITY
> > with the second half !.
>
> Wrong. If you start with some hypothesis (coding method and source
> language) and get out something sensible, even when the text is
> 50 times larger than the U.D. for that particular coding method and
> source language, doing decoding on one half and seeing that the
> other half comes out just as well only increases the probability
> that you are right with your hypothesis. You will never be certain.
> Unless the text is larger than the U.D. of the conglomerate of all
> coding methods and all possible source languages. But I think in
> most cases that U.D. is fairly large.
NONSENSE !.. See hereafter...
>
> > > The unicity distance is only for the same coding and the same language.
> >
> > OF COURSE !... Are you rediscovering the wheel?..
>
> Oh, you can also define it for a conglomerate, mind. You do not even
> have to reword the definition of U.D., only the definition of coding
> and language.
>
> > > When you allow different codings and different languages the unicity
> > > distance is much larger. So *only* when you have good reasons that
> > > your initial hypothesis are correct the unicity distance is relevant.
> > > But there is *no* way to "prove" that your initial hypothesis is indeed
> > > correct. (I use "prove" in the mathematical sense here.)
> >
> > You are MIXING UP the "Unicity Distance" which is DEFINED by the
> > *CIPHERER*, and the "ESTIMATED U.D." , which is EVALUATED BY THE
> > *DECIPHERER* !...
>
> In this case the only thing we are concerned with is the estimated U.D.
This is NOT a reason to CONFUSE BOTH NOTIONS, as you do !...
> The point is, we are talking about a script and a language, there is
> no reason to suspect that the writer had encryption in mind, so the
> U.D. of the original would be extremely low.
Well, you are beginning to understand what I mean !...
> I would be pretty surprised
> if there is any written language with a U.D. larger than, say, 10, and
> some come as low as 2 or 3, or perhaps 1 (which is the ideal figure for
> any script).
THESE FIGURES ARE TOTALLY RIDICULOUS !!!! Are you really familiar
with the Shannon's Formula?... Even in the MOST SIMPLE CASES, the
U.D. is in the range of 100 !...
>Take a mythical language, which does not inflect. Assume
> each word is represented by a unique symbol. The U.D. of that script is 1.
TOTALLY WRONG !... The U.D. is PROPORTIONAL to "Factoriel N", N being
the NUMBER of the SIGNS used by the script !... In a "WORD-SCRIPT" ,
this NUMBER is almost INFINITE !.. (Each NEW word representing a NEW
object obliges to create a NEW sign!)..
> You will nevertheless have a hard time to decipher even a longer text,
> and to prove that you are right is impossible.
>
OF COURSE !.. The U.D. is in the RANGE OF SEVERAL MILLIONS !!!!!
grapheus
grapheus
The same thing happened with the Phaistos Disk. The "Proto-Ionic
hypothesis" was considered as "impossible" in the 70ies because of the
"Risch-Chadwick Theory" saying that the Ionic dialects developped
AFTER 1200 BC. The decipherer was lead to this dialect by Statistical
considerations, and happily believe in his calculations and not in the
Risch-Chadwick Theory !... Other potential decipherers who chose Greek
as a language, were mislead by the Risch-Chadwick Theory (Benjamin
Schwartz, Steve Fischer, Louise Ollivier, etc.). Only Florence
Melian Stawell, who lived a long time BEFORE the "Risch-Chadwick
Theory's reign", has been very close to the true solution. J.
Faucounau, who dedicated his book to her, has written that "she had
been just unlucky" ...
grapheus
grapheus
If you really want not to waste your time, then JUST SHUT UP about
things YOU DON'T KNOW !..
By the same token, you will also avoid me to waste my time !...
Thank you !
grapheus
grapheus
Please read : "The U.D. is proportional to the Logarithm of "FACTORIAL N" "....
grapheus
grapheus
> Dik T. Winter wrote:
> > Take a mythical language, which does not inflect. Assume
> > each word is represented by a unique symbol. The U.D. of that script is 1.
>
> Actually, zero: it is not a cypher.
TWO TIMES WRONG !.. First, the U.D. in THAT case ("Word-Script") would
be CLOSE TO INFINITE.
Secondly, it MAY be considered as "a cypher by SIMPLE SUBSTITUTION"
!...
Read any "Manual of Cryptography" !...
>
> This notion of unicity distance/point, and the formula, were
> elaborated to answer the question: "how much must we have
> deciphered of this _ciphertext_ to be pretty sure that we
> have cracked it?"
>
> You can consider the Ph.D. to be a ciphertext
RIDICULOUS SEMANTICAL DIVERSION about the definition of "a cypher"
!...
grapheus
Peter T. Daniels
> In this case the only thing we are concerned with is the estimated U.D.
> The point is, we are talking about a script and a language, there is
> no reason to suspect that the writer had encryption in mind, so the
> U.D. of the original would be extremely low. I would be pretty surprised
> if there is any written language with a U.D. larger than, say, 10, and
> some come as low as 2 or 3, or perhaps 1 (which is the ideal figure for
> any script). Take a mythical language, which does not inflect. Assume
> each word is represented by a unique symbol. The U.D. of that script is 1.
> You will nevertheless have a hard time to decipher even a longer text,
> and to prove that you are right is impossible.
Change it to (almost) each word is represented by a pair of symbols, and
you can call it Chinese. (Are symbols allowed to recur in different
pairs for the words?)
Dik T. Winter
> "Dik T. Winter" <Dik.W...@cwi.nl> wrote in message news:<Hs8Bn...@cwi.nl>...
> > > > > This Unicity Distance is defined as follows : Supposing that the
> > > > > ennemy knows (or has guessed) the TYPE of CODE and the LANGUAGE
> > > > > used, the length of the text is equal with the U.D. if it is
> > > > > JUST ENOUGH to allow him to reach a UNIQUE solution.
> > > > 1. The smallest length such that all longer or equally long texts
> > > > code different,
> > > > or
> > > > 2. The smallest length such that for all shorter texts there exists a
> > > > text such that there codings are the same. typo:
> > > > I think he means (2).
> > >
> > > NO. You misunderstood Shannon's work !..
> > > Remember : he was interested in the SECURITY brought by the cipherment
> > > !..
> >
> > Yes? Your point being? I still think (2) is meant, because that gives
> > the greatest security. And I think Shannon has it given a valid
> > definition rather than what your paraphrase says.
>
> BALONEY !.. See hereafter...
You do not give an answer. Note that the two interpretations I wrote above
give different values, so? From a cryptographic viewpoint (2) is the best.
But I just read the paper by Shannon and found the following (I think; it is
an interpretation):
The unicity point (unicity distance) is the number of symbols you need to
have probability 1 that the decoding is correct, assuming estimated
a priori probabilities are correct.
Which is quite different from what you wrote. So with a plain substitution
cypher for English in the Latin script, when we use frequency counts of
symbols in English and in the message, the unicity distance is 27 (this is
from that paper). This means that *with these tools* the probability is 1
that a text of 27 symbols or more is decoded, with the added assumption that
the source text is a random piece of English text.
> > > Of course, you can !... If the text is 50 times longer than the U.D.,
> > > one may decipher it using only HALF of it, then VERIFY the VALIDITY
> > > with the second half !.
> >
> > Wrong. If you start with some hypothesis (coding method and source
> > language) and get out something sensible, even when the text is
> > 50 times larger than the U.D. for that particular coding method and
> > source language, doing decoding on one half and seeing that the
> > other half comes out just as well only increases the probability
> > that you are right with your hypothesis. You will never be certain.
> > Unless the text is larger than the U.D. of the conglomerate of all
> > coding methods and all possible source languages. But I think in
> > most cases that U.D. is fairly large.
>
> NONSENSE !.. See hereafter...
Yes, just out of hand, nonsense. Where is the argument? I did not find it.
> > > You are MIXING UP the "Unicity Distance" which is DEFINED by the
> > > *CIPHERER*, and the "ESTIMATED U.D." , which is EVALUATED BY THE
> > > *DECIPHERER* !...
> >
> > In this case the only thing we are concerned with is the estimated U.D.
>
> This is NOT a reason to CONFUSE BOTH NOTIONS, as you do !...
I do not think I am confused.
> > The point is, we are talking about a script and a language, there is
> > no reason to suspect that the writer had encryption in mind, so the
> > U.D. of the original would be extremely low.
>
> Well, you are beginning to understand what I mean !...
>
> > I would be pretty surprised
> > if there is any written language with a U.D. larger than, say, 10, and
> > some come as low as 2 or 3, or perhaps 1 (which is the ideal figure for
> > any script).
>
> THESE FIGURES ARE TOTALLY RIDICULOUS !!!! Are you really familiar
> with the Shannon's Formula?... Even in the MOST SIMPLE CASES, the
> U.D. is in the range of 100 !...
Lessee. With my interpretation of Shannon I gave above, even with a
substitution cypher "the unicity point is at 27 letters". And this is
when we use in deciphering only the relative frequencies of the letters.
*And this is in an attempt to encrypt.* Take a look at the papers by
Shannon. But, if the unicity point is really in the range of 100, I
am surprised that you can read some of my sentences, because they are
much shorter than that.
> >Take a mythical language, which does not inflect. Assume
> > each word is represented by a unique symbol. The U.D. of that script is 1.
>
> TOTALLY WRONG !... The U.D. is PROPORTIONAL to "Factoriel N", N being
> the NUMBER of the SIGNS used by the script !... In a "WORD-SCRIPT" ,
> this NUMBER is almost INFINITE !.. (Each NEW word representing a NEW
> object obliges to create a NEW sign!)..
This is nonsense, mathematically, for scripts. The script I described
is simply a substitution cypher. A substitution cypher of English in
the Latin script has a unicity point of 27. But in this case when you
have the correct hypothesis about language (that mythical language) and
encoding (the method by which words are transferred to script), each
symbol can be read individually.
> > You will nevertheless have a hard time to decipher even a longer text,
> > and to prove that you are right is impossible.
>
> OF COURSE !.. The U.D. is in the RANGE OF SEVERAL MILLIONS !!!!!
Depends on how you measure. When reading Shannon, I came to the
conclusion that the unicity point is dependent not only on the
coding method and the source language.
grapheus
> In article <337ae51f.04012...@posting.google.com> grap...@www.com (grapheus) writes:
> > "Dik T. Winter" <Dik.W...@cwi.nl> wrote in message news:<Hs8Bn...@cwi.nl>...
> ...
> > > > > > This Unicity Distance is defined as follows : Supposing that the
> > > > > > ennemy knows (or has guessed) the TYPE of CODE and the LANGUAGE
> > > > > > used, the length of the text is equal with the U.D. if it is
> > > > > > JUST ENOUGH to allow him to reach a UNIQUE solution.
> ...
> > > > > 1. The smallest length such that all longer or equally long texts
> > > > > code different,
> > > > > or
> > > > > 2. The smallest length such that for all shorter texts there exists a
> > > > > text such that there codings are the same. typo:
> > > > > I think he means (2).
> > > >
> > > > NO. You misunderstood Shannon's work !..
> > > > Remember : he was interested in the SECURITY brought by the cipherment
> > > > !..
> > >
> > > Yes? Your point being? I still think (2) is meant, because that gives
> > > the greatest security. And I think Shannon has it given a valid
> > > definition rather than what your paraphrase says.
> >
> > BALONEY !.. See hereafter...
>
> But I just read the paper by Shannon and found the following (I think; it is
> an interpretation):
> The unicity point (unicity distance) is the number of symbols you need to
> have probability 1 that the decoding is correct, assuming estimated
> a priori probabilities are correct.
> Which is quite different from what you wrote. So with a plain substitution
> cypher for English in the Latin script, when we use frequency counts of
> symbols in English and in the message, the unicity distance is 27 (this is
> from that paper). This means that *with these tools* the probability is 1
> that a text of 27 symbols or more is decoded, with the added assumption that
> the source text is a random piece of English text.
What Shannon's paper did you read ???????????
Your INTERPRETATION is TOTALLY WRONG !... You CONFUSED the "Unicity
Distance" with the NUMBER N of the signs USED BY THE SCRIPT (which is
27 indeed in English!).
I've been -all the long- talking about the "Shannon's Formula" given
by Shannon on page 660 in his 1949 paper in the BSTJ 28 !... Got a
look at it : As I wrote, the U.D. is PROPORTIONAL to the Logarithm of
"Factorial N" !... For N= 27, it gaves a Figure in the range of 100,
whatever the language !..
NOTA for people who have maybe read other Shannon's papers: If Shannon
has, in later papers, modified a bit his figures concerning
"Redundancy" (of English in particular), he has never changed his
1949-Formula !...
grapheus
Dik T. Winter
Communication Theory of Secrecy Systems. Original appeared in 1946 in a
confidential report: "A Mathematical Theory of Cryptography", which is now
declassified, see
<http://www.cs.ucla.edu/~jkong/research/security/shannon1949.pdf>
> Your INTERPRETATION is TOTALLY WRONG !... You CONFUSED the "Unicity
> Distance" with the NUMBER N of the signs USED BY THE SCRIPT (which is
> 27 indeed in English!).
Page 694/695 in that paper (page 39/40 in the pdf):
"With a simple substitution cipher applied to English, this calculation
gave the curves shown in Fig. 9. The key appearance characteristic in
this case was estimated by counting the number of different letters
appearing in typical English passages of N letters. In so far as
experimental data on simple substitution could be found, they all
agree very well with the curves of Fig. 9, considering the various
idealizations and approximations which have been made. For example,
*the unicity point, at about 27 letters*, can be shown experimentally
to lie between the limits 20 and 30. With 30 letters there is nearly
always a unique solution to a cryptogram of this type and with 20 it
is usually easy to find a number of solutions."
Now, what is wrong with my interpretation?
Jacques Guy
> Page 694/695 in that paper (page 39/40 in the pdf):
> "With a simple substitution cipher applied to English, this calculation
> gave the curves shown in Fig. 9. The key appearance characteristic in
> this case was estimated by counting the number of different letters
> appearing in typical English passages of N letters. In so far as
> experimental data on simple substitution could be found, they all
> agree very well with the curves of Fig. 9, considering the various
> idealizations and approximations which have been made. For example,
> *the unicity point, at about 27 letters*, can be shown experimentally
> to lie between the limits 20 and 30. With 30 letters there is nearly
> always a unique solution to a cryptogram of this type and with 20 it
> is usually easy to find a number of solutions."
> Now, what is wrong with my interpretation?
Well! How's that! This is what I had written and was about to post
when my line dropped out, and I said, fuck, I'll watch the Swedish
thriller on tv.
grapheus wrote:
> Your INTERPRETATION is TOTALLY WRONG !... You CONFUSED the "Unicity
> Distance" with the NUMBER N of the signs USED BY THE SCRIPT (which is
> 27 indeed in English!).
> I've been -all the long- talking about the "Shannon's Formula" given
> by Shannon on page 660 in his 1949 paper in the BSTJ 28 !... Got a
> look at it : As I wrote, the U.D. is PROPORTIONAL to the Logarithm of
> "Factorial N" !...
The best estimate of the unicity distance is demonstrated to be
equal to the base-2 logarithm of the number of different encipherment
schemes (K), divided by the redundancy R of the message/language
itself multiplied by the zero-order entropy, which is equal to
the base-2 logarithm of the number of different symbols (N).
UD = log2K/(R log2N)
If you assume a simple-substitution cipher with 27 symbols, then
you have K = 27! possible different encipherment schemes, and log2(27!)
is approx. 93. Divide that by the zero-order entropy, log2(27), which
is 4.75. Result: 19.6. You still have to divide that by the redundancy
of the message/language, which is one minus the infinite-order entropy
divided by the zero-order entropy. It can only be approximated.
The easiest way to do that is to compress the message with one of
the widely available compression algorithms, ZIP, RAR, or ACE for
instance.
Let's say that you get a compression ratio of 3 to 1 (30k compressed
into 10k, a reasonable figure, which will vary from "language" to
"language" (HTML code will give a much higher compression ratio
than English text) and from language to language too (Hawaiian will
compress better than English)). Then:
R = 1 - 1/3 = 2/3.
Divide the 19.6 value obtained above by that:
19.6/R = 19.6 *3 /2 = 29.4
29, not 100.
> For N= 27, it gaves a Figure in the range of 100,
> whatever the language !..
Ben justement non. Et c'est pas la peine de gueuler comme ça, coco,
t'y changeras rien tant que les andouilles n'auront pas d'ailes.
Alors là sûrement, tu seras chef d'escadrille.
Merci quand même de m'avoir fait apprendre tout ça.
grapheus
> In article <337ae51f.04013...@posting.google.com> grap...@www.com (grapheus) writes:
> > "Dik T. Winter" <Dik.W...@cwi.nl> wrote in message news:<HsA4M...@cwi.nl>...
> > > In article <337ae51f.04012...@posting.google.com> grap...@www.com (grapheus) writes:
> > > > "Dik T. Winter" <Dik.W...@cwi.nl> wrote in message news:<Hs8Bn...@cwi.nl>...
I have carefully read again Shannon's paper and found that WE BOTH
have been wrong !..
> ...
> > > > > > > > This Unicity Distance is defined as follows : Supposing that the
> > > > > > > > ennemy knows (or has guessed) the TYPE of CODE and the LANGUAGE
> > > > > > > > used, the length of the text is equal with the U.D. if it is
> > > > > > > > JUST ENOUGH to allow him to reach a UNIQUE solution.
> ...
This definition is correct.
> > > But I just read the paper by Shannon and found the following (I think; it is
> > > an interpretation):
> > > The unicity point (unicity distance) is the number of symbols you need to
> > > have probability 1 that the decoding is correct, assuming estimated
> > > a priori probabilities are correct.
NO. That the decoding is POSSIBLE, because the ennemy may reach a
UNIQUE SOLUTION.
> > > So with a plain substitution
> > > cypher for English in the Latin script, when we use frequency counts of
> > > symbols in English and in the message, the unicity distance is 27 (this is
> > > from that paper).
Here, YOU are right, and I was WRONG... The U.D. for Alphabetic
English, coded with "simple substitution cypher" is indeed c. 30.
(See Shannon's paper on page 660). Sorry about my error !..
> > > This means that *with these tools* the probability is 1
> > > that a text of 27 symbols or more is decoded, with the added assumption that
> > > the source text is a random piece of English text.
NO. It means that the decoding is POSSIBLE if one gives to the ennemy
no limitation in time, so he can try ALL the POSSIBLE solutions.
NOTA : It's your probability of 1 which have lead me to write a false
stattement. A probability close to 1 for the cryptogram to be
deciphered (in a reasonable span of time) needs a length of about 3
times the U.D., i.e c. 100 signs. All decipherers agree upon this
figure. But the U.D. itself is indeed c. 30 in the considered case.
I hope that we may agree now...
grapheus
grapheus
OK. I already recognized that it was a mistake of my part !... NO
need to show off your superior knowledge concerning this particular
calculation of the U.D. !...
Be a bit more humble, please, because this error of mine does'n't
change the errors that YOU have also written !...
grapheus
Dik T. Winter
You = grapheus
Me = me.
You:
> I have carefully read again Shannon's paper and found that WE BOTH
> have been wrong !..
Yes, in my latest message I already did write that Shannon's definition
differed from what I wrote.
You:
> This Unicity Distance is defined as follows : Supposing that the
> ennemy knows (or has guessed) the TYPE of CODE and the LANGUAGE
> used, the length of the text is equal with the U.D. if it is
> JUST ENOUGH to allow him to reach a UNIQUE solution.
But after reading the paper, this is still wrong. Let's see what he
wrote about the substitution cipher: when the text is longer than 30
symbols, the solution is unique, when it is less than 20 there are
multiple solutions, the unicity point is at 27. What this basically
means is that between 20 and 30 symbols of text the solution may be
unique or not. The figure of 27 is reached by weighing.
You:
> NO. That the decoding is POSSIBLE, because the ennemy may reach a
> UNIQUE SOLUTION.
Decoding may also be possible with shorter texts, if there is a unique
solution, and may not be possible with longer texts if there is not a
unique solution.
Me:
> So with a plain substitution
> cypher for English in the Latin script, when we use frequency counts of
> symbols in English and in the message, the unicity distance is 27 (this is
> from that paper).
You:
> Here, YOU are right, and I was WRONG... The U.D. for Alphabetic
> English, coded with "simple substitution cypher" is indeed c. 30.
> (See Shannon's paper on page 660). Sorry about my error !..
Ok. (And as most scripts can be seen as substitution ciphers...)
Me:
> This means that *with these tools* the probability is 1
> that a text of 27 symbols or more is decoded, with the added assumption that
> the source text is a random piece of English text.
Wrong indeed.
You:
> NO. It means that the decoding is POSSIBLE if one gives to the ennemy
> no limitation in time, so he can try ALL the POSSIBLE solutions.
Wrong again. If the length of the text is 30 symbols or more, there *is*
a unique solution (see Shannon). So when you state all the possible
solutions, you can only mean a single solutions. On the other hand,
when you had said all the tentative solutions, it would have been
different. And I do not remember anything about the time needed to
decode in Shannon's paper. But indeed in a substitution cipher of 26
letters, there are 26! possible encodings. However, using frequency
counts, this drops pretty fast. (I.e. if you have a text of 30 letters
with 5 times an 'x', it is highly unlikely that it represents the 'q'.)
Moreover, additional tools are used. In normal texts (and we are talking
about fairly normal texts when we try to decipher an unknown script),
there are letter combinations that are extremely unlikely, so also
encodings that lead to such combinations can be ruled out.
> NOTA : It's your probability of 1 which have lead me to write a false
> stattement. A probability close to 1 for the cryptogram to be
> deciphered (in a reasonable span of time) needs a length of about 3
> times the U.D., i.e c. 100 signs.
I do not know what you mean with "reasonable span of time". I agree
that longer texts give faster decodings, because the basic rules can
rule out many more encodings. However, the probability is 1 (yes),
when there is a unique solution, and in Shannon's paper I read that
is the case for all texts with a length of more than 30 symbols. So
with a substitution cipher, when the text is long enough (i.e. longer
than 30 symbols) the probability is 1 that a solution is found. How
long it takes entirely depends on how much power you allow on the
deciphering.
Matt Giwer
> For readers in this Group, who are not mathematicians and can't read
> Shannon's original work by lack of mathematical knowledge, here is a
> Summary of it (that J.Faucounau, another mathematician, sent to me a
> few years ago).
>
> will not understand it !...
>
> The mathematician Claude Shannon was mainly interested in the
> EFFICIENCY of a Cryptographic Code. He started from the obvious remark
> that longer is a text, and more probable is its decipherment. To be
> sure that a cryptogram will not be deciphered, one has then to limit
> its length under a certain "number of signs", that he called the
> "Unicity Distance".
I never had a problem with math but I didn't really bother to get
ahead of the course material. I read and understood it as junior in
high school. The work is insightful not difficult. (It is on the web
but I don't have the URL. Slashdot posted the link when it went online
or rely on google.)
I have often consider the applicability of information content to
dealing with untranslated languages and there is none.
What are you trying to say?
--
Despite all the elite objections populism is democracy.
-- The Iron Webmaster, 2999
Matt Giwer
> Dik T. Winter wrote:
>>In article <337ae51f.04012...@posting.google.com> grap...@www.com (grapheus) writes:
>> > This Unicity Distance is defined as follows : Supposing that the
>> > ennemy knows (or has guessed) the TYPE of CODE and the LANGUAGE used,
>> > the length of the text is equal with the U.D. if it is JUST ENOUGH to
>> > allow him to reach a UNIQUE solution.
>>I do not think this is the definition
> Doesn't matter, it's good enough for all practical purposes.
> "Supposing that the enemy knows (or has guessed) the language".
When one is guessing the language it has to be known in all of its
essential details.
> John Chadwick had pretended that it was Medieval Japanese, and
> he had deciphered it as a legal text on water rights.
>
> Faucounau has guessed that the language was Proto-Ionic (which
> he made up).
>
> Padre Maricón y Cojones has guessed that it was in Quechua and
> was about llama-shearing (pers.com.)
>
> Noted Greek philosopher Polykrates of Smyrna had guessed that
> it was in Amazonian and was a text of mastectomy surgery. Alas, none
> of his works have survived.
>
> Mr or Mrs Xqblwykpq, on a visit here from planet Qolob in galaxy
> Feefoefiefum
> told me that ukduieo pd owkjd sisuiufhe yhshe ioo ehnasovmc smeiqtyqspe.
> I do not think that he, or she, or it, or ...(???) has got it right.
And the Rosetta stone was "correctly" deciphered twice before it was
done correctly -- at least we think it is correct.
--
Whenever you hear someone accused of mass murder or genocide you
can be certain there are holohuggers making up the stories.
-- The Iron Webmaster, 3023
grapheus
Fine !.. You are discovering the importance of the PROOFS after the
decipherment !...
grapheus
grapheus
> Let me rearrange a bit:
> You = grapheus
> Me = me.
>
> You:
> > I have carefully read again Shannon's paper and found that WE BOTH
> > have been wrong !..
>
> Yes, in my latest message I already did write that Shannon's definition
> differed from what I wrote.
>
> You:
> > This Unicity Distance is defined as follows : Supposing that the
> > ennemy knows (or has guessed) the TYPE of CODE and the LANGUAGE
> > used, the length of the text is equal with the U.D. if it is
> > JUST ENOUGH to allow him to reach a UNIQUE solution.
>
> But after reading the paper, this is still wrong. Let's see what he
> wrote about the substitution cipher: when the text is longer than 30
> symbols, the solution is unique, when it is less than 20 there are
> multiple solutions, the unicity point is at 27. What this basically
> means is that between 20 and 30 symbols of text the solution may be
> unique or not. The figure of 27 is reached by weighing.
No. The figure of 27 is an APPROXIMATION. And as all the
approximations, there is a "zone of uncertainty" around it.
Theorically, the U.D. is the length which is JUST ENOUGH to reach the
UNIQUE solution (my definition). But, as it's impossible to know it
WITH PRECISION (would it be because of the ciphered text itself : as
you noticed it, repeating 3 times a 10 signs sentence is not the same
as 3 different sentences with a length of 30 signs !), one has to deal
with an APPROXIMATION.
>
> You:
> > NO. That the decoding is POSSIBLE, because the ennemy may reach a
> > UNIQUE SOLUTION.
>
> Decoding may also be possible with shorter texts, if there is a unique
> solution,
This is CORRECT. I remember that J. Faucounau gave in one of his
papers an example concerning a 6 letters text which could be easily
deciphered !..
> and may not be possible with longer texts if there is not
> a unique solution
By DEFINITION, if the U.D. is reached, the UNIQUE SOLUTION is there !
I agree on almost everything, but your "probability of 1".
Because it depends of the TIME you give to the decipherer to decipher
a cryptogram, of which length is equal to the U.D.
You write : "I don't remember anything about the time needed in
Shannon's paper" . Look at page 659, when he says that his study will
deal with the problem (I quote) :"How secure is a system against
cryptanalysis WHEN THE ENNEMY HAS UNLIMITED TIME and manpower
available for the analysis of intercepted cryptograms?" (emphasis is
mine).
Therefore, to speak of a "probability of 1" for the cryptogram to be
deciphered when there is a unique solution is implicitely supposing an
UNLIMITED TIME at the disposal of the decipherer. Such a
"probability" has NO MEANING in science.
grapheus
Peter T. Daniels
> And the Rosetta stone was "correctly" deciphered twice before it was
> done correctly -- at least we think it is correct.
What are you referring to, and what are your references?
Dik T. Winter
> > > This Unicity Distance is defined as follows : Supposing that the
> > > ennemy knows (or has guessed) the TYPE of CODE and the LANGUAGE
> > > used, the length of the text is equal with the U.D. if it is
> > > JUST ENOUGH to allow him to reach a UNIQUE solution.
> >
> > But after reading the paper, this is still wrong. Let's see what he
> > wrote about the substitution cipher: when the text is longer than 30
> > symbols, the solution is unique, when it is less than 20 there are
> > multiple solutions, the unicity point is at 27. What this basically
> > means is that between 20 and 30 symbols of text the solution may be
> > unique or not. The figure of 27 is reached by weighing.
>
> No. The figure of 27 is an APPROXIMATION.
Eh? Not by the definition of Shannon.
> And as all the
> approximations, there is a "zone of uncertainty" around it.
> Theorically, the U.D. is the length which is JUST ENOUGH to reach the
> UNIQUE solution (my definition).
Still not enough for a proper definition. Do you mean my (1) or (2) as
I gave them in a previous post?
> > and may not be possible with longer texts if there is not
> > a unique solution
>
> By DEFINITION, if the U.D. is reached, the UNIQUE SOLUTION is there !
Ah, you mean my (1), which you classified as BALONEY.
> I agree on almost everything, but your "probability of 1".
> Because it depends of the TIME you give to the decipherer to decipher
> a cryptogram, of which length is equal to the U.D.
> You write : "I don't remember anything about the time needed in
> Shannon's paper" . Look at page 659, when he says that his study will
> deal with the problem (I quote) :"How secure is a system against
> cryptanalysis WHEN THE ENNEMY HAS UNLIMITED TIME and manpower
> available for the analysis of intercepted cryptograms?" (emphasis is
> mine).
Yup. The reason is that in cryptanalysis it is difficult to also include
the time factor, if not impossible.
> Therefore, to speak of a "probability of 1" for the cryptogram to be
> deciphered when there is a unique solution is implicitely supposing an
> UNLIMITED TIME at the disposal of the decipherer. Such a
> "probability" has NO MEANING in science.
Probability is not concerned with time, unless you can quantify it with
that factor. So, as I am not able to factor in the time needed, I do
not know what to do with the probability if the time is limited. How
do you factor in the time needed? Just by saying the probability is
close to 1? But in that case how close to 1, and in what way does that
depend on the time needed. Now? 20 years ago? 20 years hence?
grapheus
It depends of course of the manpower used !...
Why don't you forget for a while "Brute Logic" ?.. Remember that in
Mathematics, the "Calculus of Probabilities" is a branch of
Mathematics, which developped FOR HELPING HUMAN DECISIONS or CHOICES
!... The CONCLUSION of a calculation of Probability is generally :
"Therefore I MAY REASONABLY take THIS DECISION" or : "THERE IS TOO
MUCH RISK of being wrong if I take THIS decision"... The Calculus of
Probabilities is the MOST HUMANIST part of Mathematics !... As Shannon
himself wrote (on page 664) : "A cryptanalyst might be willing to
classify the possible messages into the categories "reasonable",
"possible but unlikely" and "unreasonable", but feel that finer
subdivision was meaningless"... The Calculus of Probabilities is a
MAGNIFICENT TOOL. Only IGNARROGANTS try to pollute it with their
"Brute Logic" in black and white!... TRUE MATHEMATICIANS don't...
grapheus
Jacques Guy
[snip]
Instead of spending here your hard-earned bangs and CAPS,
whip up an English translation of what Faucounau had to say
in his L'Harmattan book, and load it up on some web site.
www.geocities.com will be happy to host megabytes of it
for free.
There is the problem of copyright of course. Did you
you sign it off to L'Harmattan?
grapheus
Stupid question !.. When will you understand that only kooks like
Franz G. believe that I am J. Faucounau ?.... But, even not being the
author does'n't give me the right to translate a book without the
authorisation of the Editor. I don't intend to get into trouble for
breaking the Copyrights Law, just for sparing to a few ignarrogants
the effort of reading it !... If you know an editor interested in
publishing a translation of the book, ask him to get in touch with
L'Harmattan in Paris, who is, I suppose, the owner of the
copyrights!...
grapheus
Wolf Kirchmeir
>The Calculus of Probabilities is a
>MAGNIFICENT TOOL. Only IGNARROGANTS try to pollute it with their
>"Brute Logic" in black and white!... TRUE MATHEMATICIANS don't...
Probability calculations aren't guess work - they are are subject to the same
"brute logic" as any other branch of mathematics.
Just a gentle reminder.
BTW, you still haven't explained how you arrived at the conclusion that the
probability of Faucounau's work being true is {1 - 10^(-50)}, or in decimals
0.99....... to 50 decimal places.
Jacques Guy
> BTW, you still haven't explained how you arrived at the conclusion that the
> probability of Faucounau's work being true is {1 - 10^(-50)}, or in decimals
> 0.99....... to 50 decimal places.
No point in that: IGNARROGANTS wouldn't understand the explanation.
I took the liberty to correct the subject line. I've posted enough
about Phoque Engleesh.
Matt Giwer
> Matt Giwer wrote:
>> And the Rosetta stone was "correctly" deciphered twice before it was
>>done correctly -- at least we think it is correct.
> What are you referring to, and what are your references?
Off hand I don't have any. I read a couple articles on it and they
refered to other's claiming to have deciphered it but were wrong. I
don't remember reading any saying how wrong they were. I'm sure you
can find mention of it with a google search. I was just reminding. I
thought it was common knowledge.
--
If you wish to murder people find your own reasons against
the people next door. Do not depend upon your government
to do it for you. The government may not want you to kill
when you feel like killing. Do it yourself.
-- The Iron Webmaster, 3015
Matt Giwer
I had no need to discover it.
--
Why the code of silence among the Democrat candidates
over Bush's lies about Iraq?
-- The Iron Webmaster, 3013
grapheus
>
> >The Calculus of Probabilities is a
> >MAGNIFICENT TOOL. Only IGNARROGANTS try to pollute it with their
> >"Brute Logic" in black and white!... TRUE MATHEMATICIANS don't...
>
> Probability calculations aren't guess work - they are are subject to the same
> "brute logic" as any other branch of mathematics.
>
Surely, they are'n't GUESS WORK !... But they aren't neither subject
to Brute Logic, which PUTS INTO THE SAME BAG an event having a one out
of a million chances to be right and an event having one out of a
million chances to be wrong under the pretext that "BOTH ARE POSSIBLE"
!...
This is the point that IGNARROGANTS cannot understand !..
> Just a gentle reminder.
>
> BTW, you still haven't explained how you arrived at the conclusion that the
> probability of Faucounau's work being true is {1 - 10^(-50)}, or in decimals
> 0.99....... to 50 decimal places.
BE INFORMED about the PROOFS which have been presented, and you will
understand the figure I gave !...
grapheus
Dik T. Winter
> "Dik T. Winter" <Dik.W...@cwi.nl> wrote in message news:<HsDwt...@cwi.nl>...
...
> > Probability is not concerned with time, unless you can quantify it with
> > that factor. So, as I am not able to factor in the time needed, I do
> > not know what to do with the probability if the time is limited. How
> > do you factor in the time needed? Just by saying the probability is
> > close to 1? But in that case how close to 1, and in what way does that
> > depend on the time needed. Now? 20 years ago? 20 years hence?
>
> It depends of course of the manpower used !...
Amongst others.
> Why don't you forget for a while "Brute Logic" ?.. Remember that in
> Mathematics, the "Calculus of Probabilities" is a branch of
> Mathematics, which developped FOR HELPING HUMAN DECISIONS or CHOICES
> !...
That is new for me.
> The CONCLUSION of a calculation of Probability is generally :
> "Therefore I MAY REASONABLY take THIS DECISION" or : "THERE IS TOO
> MUCH RISK of being wrong if I take THIS decision"...
Nope, that is *never* a conclusion of a calculation. That is a human
interpretation of the conclusion of a calculation. Mathematics (and
in fact statistics) can at most say what the probability is that you
go wrong with a particular decision. Whether that is "TOO MUCH RISK"
is something else, and will mainly depend on the person who has to
make that decision.
> The Calculus of Probabilities is a
> MAGNIFICENT TOOL. Only IGNARROGANTS try to pollute it with their
> "Brute Logic" in black and white!... TRUE MATHEMATICIANS don't...
Well, I have to look back at my statistics course books from when I
was at University studying mathematics, it appears...
Peter T. Daniels
>
> Peter T. Daniels wrote:
> > Matt Giwer wrote:
>
> >> And the Rosetta stone was "correctly" deciphered twice before it was
> >>done correctly -- at least we think it is correct.
>
> > What are you referring to, and what are your references?
>
> Off hand I don't have any. I read a couple articles on it and they
> refered to other's claiming to have deciphered it but were wrong. I
> don't remember reading any saying how wrong they were. I'm sure you
> can find mention of it with a google search. I was just reminding. I
> thought it was common knowledge.
I'm considered something of an authority on the history of decipherment
(they keep asking me for encyclopedia articles on it), and I've never
heard of such a thing. It would be nice to know what you're talking
about.
Wolf Kirchmeir
>BE INFORMED about the PROOFS which have been presented, and you will
>understand the figure I gave !...
>
>grapheus
I doubt it, since a probability statement of that high a level of certainty
is nonsensical. It exceeds all possible probabilities of events in the
material universe, eg, the probability that a gas will remain uniformly
dispersed within a container of a given size.
grapheus
> Matt Giwer wrote:
> >
> > Peter T. Daniels wrote:
> > > Matt Giwer wrote:
>
> > >> And the Rosetta stone was "correctly" deciphered twice before it was
> > >>done correctly -- at least we think it is correct.
>
> > > What are you referring to, and what are your references?
> >
> > Off hand I don't have any. I read a couple articles on it and they
> > refered to other's claiming to have deciphered it but were wrong. I
> > don't remember reading any saying how wrong they were. I'm sure you
> > can find mention of it with a google search. I was just reminding. I
> > thought it was common knowledge.
>
> I'm considered something of an authority on the history of decipherment
> (they keep asking me for encyclopedia articles on it),
>
> about.
It would be also nice if you applied this elementary scientific
requirement to yourself when you are discussing about the J.F.'s
"Proto-Ionian Theory" !...
grapheus
Bobby D. Bryant
> BE INFORMED about the PROOFS which have been presented, and you will
> understand the figure I gave !...
We still haven't even seen your _summary_ of the proofs, let alone the
proofs themselves.
Can't you pick just one, say "the astronomical proof", and give us a short
summary of the evidence and argument? You invoke it so often that we can
only assume you know and understand it well.
--
Bobby Bryant
Austin, Texas
Matt Giwer
So if I do a simple google search and find a recitation of prior
incorrect translations you will retract your claim of "something of an
authority" no? I suggest you do the simple search before you reply.
Miguel Carrasquer
<bdbr...@mail.utexas.edu> wrote:
>On Mon, 02 Feb 2004 01:17:51 -0800, grapheus wrote:
>
>> BE INFORMED about the PROOFS which have been presented, and you will
>> understand the figure I gave !...
>
>We still haven't even seen your _summary_ of the proofs, let alone the
>proofs themselves.
>
>Can't you pick just one, say "the astronomical proof", and give us a short
>summary of the evidence and argument?
The paragraph in Faucounau '99 is easy to quote [italics rendered as
CAPITALS, to stay in style]:
4- La preuve astronomique
La preuve qu'existait au milieu du Troisième Millénaire une civilisation
grecque (proto-ionienne) et cycladique a été découverte bien avant le
déchiffrement du disque de Phaistos, puisque les premiers travaux de
l'astronome anglais E.W. Maunder remontent à 1909. L'idée, généralement
admise depuis la découverte du grec mycénien, que les Grecs n'étaient
arrivés que tardivement en Egée (au plus tôt vers 1900 avant JC) avait
conduit les historiens à laisser prudemment de côté les conclusions des
astronomes s'étant occupés de la question, à savoir que l'observatoire des
"inventeurs des Constellations" se trouvait, vers 2500 avant notre ère, en
un lieu situé sur un parallèle à 100 km. environ au nord de la Crète.
Ces CONCLUSIONS SONT INDISCUTABLES. Elles ont été établies à partir de
trois méthodes différentes, par plusieurs astronomes réputés (E.W. Maunder,
A. Crommelin, M. Ovenden, Archie Roy, etc.). Mais, en l'absence de la
théorie proto-ionienne, elles posaient un problème insoluble (Voir
l'historique de la question dans KADATH 83, 1994). Le déchiffrement du
disque de Phaistos a permis de résoudre l'énigme, acquérant du même coup
une nouvelle preuve, INDISCUTABLE ET DÉCISIVE, de son exactitude.
For the benefit of non-francophones:
4- The astronomical proof
The proof that in the middle of the Third Millennium a Greek (proto-Ionian)
and Cycladic civilization existed, was discovered long before the
decipherment [i.e. Faucounau's decipherment --mcv] of the Phaistos Disk, as
the first works by the English astronomer E.W. Maunder go back to 1909. The
idea, generally acepted after the descovery of Mycenean Greek, that the
Greeks had only arrived in the Aegean late (ca. 1900 BC at the earliest)
had led historians to prudently put aside the conclusions of the
astronomers who had studied the question, to wit that the observatory of
the "inventors of the Constellations" had been located, around 2500 BC, at
a place situated at a parallel 100 km. north of Crete. These CONCLUSIONS
ARE INDISPUTABLE. They have been established based on three different
methods [for an explanation of these methods, see John Pratt's website
http://www.johnpratt.com/items/docs/constellations_date.html (but disregard
the last sentence) --mcv], by several astronomers of repute (E.W. Maunder,
A. Crommelin, M. Ovenden, Archie Roy, etc.). But, without the Proto-Ionian
theory, they posed an insoluble problem (see a history of the question in
KADATH 83, 1994). The decipherment of the Phaistos Disk has allowed the
mystery to be resolved, at the same time adding another proof, INDISPUTABLE
AND DECISIVE, of its [the decipherment's] correctness.
For an overview of the Constellation question, see
http://users.macunlimited.net/ianrid/startales/startales1a
As I read that article, the "inventors of the Constellations" were the
Sumerians. The Greeks acquired their list of constellations from Egypt
(Eudoxus), and the Egyptians had acquired it from the
Sumerians/Babylonians, perhaps through Crete (as suggested by Archie Roy),
perhaps directly (as seems more likely to me). In any case, the question
is obviously of no relevance whatsoever to the decipherment of the Phaistos
Disk.
=======================
Miguel Carrasquer Vidal
m...@wxs.nl
Peter T. Daniels
>
> Peter T. Daniels wrote:
> > Matt Giwer wrote:
> >
> >>Peter T. Daniels wrote:
> >>
> >>>Matt Giwer wrote:
> >>
> >>>> And the Rosetta stone was "correctly" deciphered twice before it was
> >>>>done correctly -- at least we think it is correct.
> >>
> >>>What are you referring to, and what are your references?
> >>
> >> Off hand I don't have any. I read a couple articles on it and they
> >>refered to other's claiming to have deciphered it but were wrong. I
> >>don't remember reading any saying how wrong they were. I'm sure you
> >>can find mention of it with a google search. I was just reminding. I
> >>thought it was common knowledge.
>
> > I'm considered something of an authority on the history of decipherment
> > (they keep asking me for encyclopedia articles on it), and I've never
> > heard of such a thing. It would be nice to know what you're talking
> > about.
>
> So if I do a simple google search and find a recitation of prior
> incorrect translations you will retract your claim of "something of an
> authority" no? I suggest you do the simple search before you reply.
So you don't, in fact, have anything to offer.
grapheus
TOTALLY WRONG !... Why you don't ALSO quote the KADATH 83 (1994)
issue, in which this problem has been examined IN DETAIL, and in which
THE LINGUISTICAL PART of the "Astonomical Proof" has been added ?...
In short : the NAMES of the CONSTELLATIONS are NEITHER EGYPTIAN, nor
SUMERIAN, nor CRETAN !.. THEY ARE PROTO-IONIAN GREEK !..
This is ONCE AGAIN the way IGNARROGANTS try to DISPARAGE a theory they
WANT to IGNORE !... By QUOTING JUST A FEW LINES or an INCOMPLETE
SUMARY of a COMPLEX DEMONSTRATION !... You should be ashame of
yourself, Miguel !..
grapheus
Jim Webster
> Sumerians. The Greeks acquired their list of constellations from Egypt
> (Eudoxus), and the Egyptians had acquired it from the
> Sumerians/Babylonians, perhaps through Crete (as suggested by Archie Roy),
> perhaps directly (as seems more likely to me). In any case, the question
> is obviously of no relevance whatsoever to the decipherment of the
Phaistos
> Disk.
to quote"Astronomy before the Telescope" (the chapter of interest is written
by John Britton and Christopher Walker)
"Comparison with Ptolemy's catalogue which. it must be admitted probably
served as an essential backgroud to the task of identification, suggests
that the names and delimiting of many of the Greek ZOdiac constellations are
of Mesopotamian origin"
Jim Webster
Peter T. Daniels
>
> Miguel Carrasquer <m...@wxs.nl> wrote in message news (translating Faucounau):
> > the observatory of
> > the "inventors of the Constellations" had been located, around 2500 BC, at
> > a place situated at a parallel 100 km. north of Crete. These CONCLUSIONS
> > ARE INDISPUTABLE. They have been established based on three different
> TOTALLY WRONG !... Why you don't ALSO quote the KADATH 83 (1994)
> issue, in which this problem has been examined IN DETAIL, and in which
> THE LINGUISTICAL PART of the "Astonomical Proof" has been added ?...
> In short : the NAMES of the CONSTELLATIONS are NEITHER EGYPTIAN, nor
> SUMERIAN, nor CRETAN !.. THEY ARE PROTO-IONIAN GREEK !..
What "names OF THE constellations"? They are named in each language
according to the language. We have Egyptian, Sumerian, Akkadian, etc.,
constellation names.
As I mentioned quite a while ago, the parallel of latitude 100 km north
of Crete goes all the way around the world, so the astronomy could have
been done anywhere at that latitude.
Miguel Carrasquer
>TOTALLY WRONG !... Why you don't ALSO quote the KADATH 83 (1994)
>issue, in which this problem has been examined IN DETAIL, and in which
>THE LINGUISTICAL PART of the "Astonomical Proof" has been added ?...
>In short : the NAMES of the CONSTELLATIONS are NEITHER EGYPTIAN, nor
>SUMERIAN, nor CRETAN !.. THEY ARE PROTO-IONIAN GREEK !..
The names of the constellations are Egyptian in Egyptian, Sumerian in
Sumerian and Greek in Greek, bien sūr. I haven't seen Kadath 83, so why
don't *you* tell us what makes the Greek names "Proto-Ionian Greek". Could
it be that Eudoxus of Cnidus, who first invented the Greek names, was an
Ionian?
>This is ONCE AGAIN the way IGNARROGANTS try to DISPARAGE a theory they
>WANT to IGNORE !...
If I wanted to ignore it, I would ignore it. In fact, I think you're
right. I *should* ignore it. Faucounau's book is perfectly ignorable, as
it contains nothing of interest. And your antics as Faucounau's Usenet
buffoon are even more ignorable still, from a scientific point of view, as
you have managed to reduce the signal/noise (information/SHOUTING) ratio
from very low to absolutely zero. The only reason I'm not ignoring your
stupidity is that I find it so annoying.
>By QUOTING JUST A FEW LINES or an INCOMPLETE
>SUMARY of a COMPLEX DEMONSTRATION !...
I quoted a everything Faucounau says about the question in his book. It's
not my fault the book is worthless crap.
grapheus
First Question : What is the DATE of the "Ptolemy's Catalogue" ?...
(Answer, in the case you don't know it : c. 150 A.D. , i.e MORE THAN
2500 YEARS AFTER the "Invention of the Constellations" !!!!)
Second Question : What is the DATE of the "Mesopotamian documents"
quoted by J. Briton & C. Walker ?...
(Answer, in the case you don't know it : they go back to the "time of
Babylon", i.e. c. 1800 BC or later, what is approximately one
MILLENIUM AFTER the "invention of the Constellations" !!!!!
Third question : When J. Briton & C. Walker wrote this sentence ?..
(Answer, in the case you don't know it : At the time when the
"Risch-Chadwick Theory" was ACCEPTED as A DOGMA !!!!!
grapheus
Jim Webster
"grapheus" <grap...@www.com> wrote in message
news:337ae51f.04020...@posting.google.com...
so what?
> Second Question : What is the DATE of the "Mesopotamian documents"
> quoted by J. Briton & C. Walker ?...
> (Answer, in the case you don't know it : they go back to the "time of
> Babylon", i.e. c. 1800 BC or later, what is approximately one
> MILLENIUM AFTER the "invention of the Constellations" !!!!!
no, the earliest we have found that relate specifically to astronomical
observations go back to 1800BC. Cuniform goes back to 3000BC. It is possible
that purely by chance we have found the oldest astronomical observation, but
given the technical competence of it, this is exceedingly unlikely
> Third question : When J. Briton & C. Walker wrote this sentence ?..
> (Answer, in the case you don't know it : At the time when the
> "Risch-Chadwick Theory" was ACCEPTED as A DOGMA !!!!!
The book was conceived in 1991, so by definition they would have been asked
to write this after this point.
Except that Britton and Walker were not discussing mycenaean or greek
studies but Astrology and Astronomy in ancient mesopotamia
Jim Webster
>
> grapheus
grapheus
> On 3 Feb 2004 06:53:59 -0800, grap...@www.com (grapheus) wrote:
>
> >TOTALLY WRONG !... Why you don't ALSO quote the KADATH 83 (1994)
> >issue, in which this problem has been examined IN DETAIL, and in which
> >THE LINGUISTICAL PART of the "Astonomical Proof" has been added ?...
> >In short : the NAMES of the CONSTELLATIONS are NEITHER EGYPTIAN, nor
> >SUMERIAN, nor CRETAN !.. THEY ARE PROTO-IONIAN GREEK !..
>
> The names of the constellations are Egyptian in Egyptian, Sumerian in
First Remark : WHY are you quoting EGYPT, a country which CANNOT be
the land of the "Inventors of the Constellations" because of its
latitude ?...
Second Remark : Why are you mentioning SUMER, a country which had not
the same astronomical knowledge as the BABYLONIANS ?.. That the
BABYLONIAN astronomical knowledge goes back to the SUMERIANS is JUST
AN HYPOTHESIS , based upon NOTHING but the desire to find in
Mesopotamia the "inventors of the Constellations" !.. (Of course !..
As long as scholars did'n't know the existence of the Proto-Ionians,
there was no other possible "reasonable" solution !!!!)
Third Remark : Having NOT made any SERIOUS RESEARCH about the
"astronomical proof", you IGNORE of course that the Constellations are
NOT EXACTLY THE SAME in Egypt/Babylon and Phoenicia/Greece. The first
countries essentially knew ZODIACAL Constellations -- the second knew
the constellations which are ALONG THE ECLIPTIC !.. An essential
DIFFERENCE that you IGNORE !... It's in this second group of
constellations that one may find the NAMES of HERAKLES, PERSEUS,
CENTAUR, HYDRA, PEGASUS, which BELONG to the GREEK MYTHOLOGY, not to
the Babylonian one, which know only the ZODIACAL ones!... Concerning
the first group, most names are DIFFERENT from a Civilization to
another : The Greek SWAN is a PANTHER in Babylonia, the Greek GREAT
BEAR is a CROCODILE in Egypt, etc. But of course, you don't care !..
Fourth Remark : You also IGNORE that the calculations made by the
astronomers cited in J.F.'s book concern BOTH GROUPS, and it's for
the second group that the NAMES are important because of that!...
> I haven't seen Kadath 83,
Of course, you hav'n't read KADATH 83, and you have NOT read neither
the J.F.'s book "Les Proto-Ioniens..." in spite of what I wrote : that
it's in this last book that one may find the DETAILS of a few PROOFS,
not in the one you cited... WITHOUT, of course, having the HONESTY TO
CITE also page 15 : "Le chapitre 7 traite de l'importante question des
preuves externes..; Ces diverses questions ayant été déjà développées
dans des articles, un simple résumé est fourni... " English
translation : "Chapter 7 deals with the important question of the
external proofs... As the detail of these has already been given in
(previous) papers, JUST A SUMMARY IS PRESENTED (here)"... (emphasis is
mine).
> so why don't *you* tell us what makes the Greek names "Proto-Ionian Greek"
Here is one reason amongst others : The names in -AD- (Pleiades,
Hyades, etc.) belong to Ionic-attic or to Homeric language : e.g.
lampad-, orgad-, ikmad-, logad-, nomad-, stibad-, etc. (See P.
Chantraine, "La formation des noms en grec ancien").
> Could
> it be that Eudoxus of Cnidus, who first invented the Greek names, was an
> Ionian?
>
If you were INFORMED, you would know that obviously Eudoxus DID NOT
INVENT the Greek Names!.. He USED an old, PROTO-IONIAN TRADITION which
had probably been kept IN EGYPT !... (YES !!!!! And there are strong
reasons to believe that)..
The question of the "Sphere of Eudoxus" has been examined in KADATH
83. But asking IGNARROGANTS to make the necessary scientific work of
READING ALL THE BIBLIOGRAPHICAL REFERENCES before emitting THEIR
judgement is, of course, asking too much !.. IGNARROGANTS are so
intelligent, that they don't need making the effort !... They KNOW
everything in advance !...
> >This is ONCE AGAIN the way IGNARROGANTS try to DISPARAGE a theory they
> >WANT to IGNORE !...
>
> If I wanted to ignore it, I would ignore it.
BUT YOU IGNORE IT !.. You have just read ONE book in which it is
EXPLICITELY ASKED the reader to READ A FEW OTHER PAPERS if he wants to
know the details of the "astronomical Proof", and you EMITTED A
JUDGEMENT FROM a few lines SUMMARY !... THIS IS TYPICAL of
IGNARROGANCE !...
>
> I quoted a everything Faucounau says about the question in his book.
BUT WITHOUT quoting the lines saying that the reader would have
sometimes to read other papers to be fully informed !..
>It's not my fault the book is worthless crap.
But it's surely J.F.'s fault if he did'n't think that BAD FAITH
IGNARROGANTS would take the pretext of his SUMMARIES to emit
judgements like the hereabove!...
In this, he is guilty, I agree !..
grapheus
Matt Giwer
Right out of the box.
http://www.chesco.com/~cslice/aurora/rosetta/rosetta.html
--
Bush claims he is going to eradicate evil from the world.
Even his god could not do that.
-- The Iron Webmaster, 3003
Miguel Carrasquer
>Of course, you hav'n't read KADATH 83, and you have NOT read neither
>the J.F.'s book "Les Proto-Ioniens..." in spite of what I wrote : that
>it's in this last book that one may find the DETAILS of a few PROOFS,
>not in the one you cited... WITHOUT, of course, having the HONESTY TO
>CITE also page 15 : "Le chapitre 7 traite de l'importante question des
>preuves externes..; Ces diverses questions ayant été déjà développées
>dans des articles, un simple résumé est fourni... " English
>translation : "Chapter 7 deals with the important question of the
>external proofs... As the detail of these has already been given in
>(previous) papers, JUST A SUMMARY IS PRESENTED (here)"... (emphasis is
>mine).
[...]
>> I quoted a everything Faucounau says about the question in his book.
>
>BUT WITHOUT quoting the lines saying that the reader would have
>sometimes to read other papers to be fully informed !..
>
>>It's not my fault the book is worthless crap.
>
>But it's surely J.F.'s fault if he did'n't think that BAD FAITH
>IGNARROGANTS would take the pretext of his SUMMARIES to emit
>judgements like the hereabove!...
>
>In this, he is guilty, I agree !..
Let me quote p. 12: "Pour des raisons dues à l'histoire du déchiffrement,
la plupart de ces preuves ont déjà été publiées, mais de façon éparse, dans
des revues diverses. Un ouvrage reprenant l'ensemble du dossier, de
manière à le soumettre au feu de la critique scientifique s'imposait donc.
C'est cet ouvrage que nous proposons aujourd'hui."
Well, if you don't give the details about the decipherment, and then only
the weakest of summaries of the other "proofs", then I'd say you have
failed spectacularly to "submit yourself to scientific criticism". The
only appropriate scientific response is to ignore Faucounau's
"decipherment". Which is what I'm going to do now.
grapheus
Please, EXPLAIN the difference with what I wrote and the change of a
YES into your NO !!!!
> Cuniform goes back to 3000BC. It is possible
> that purely by chance we have found the oldest astronomical observation, but
> given the technical competence of it, this is exceedingly unlikely
>
Irrelevant !... The DATE of the "invention of the Constellations" is
c. 2500BC, not 1800BC... Moreover, it has been shown that these
Constellations were invented by a MARITIME people.
But, of course, how could you know ?.. You refuse to read the
referenced papers!...
>
> > Third question : When J. Briton & C. Walker wrote this sentence ?..
> > (Answer, in the case you don't know it : At the time when the
> > "Risch-Chadwick Theory" was ACCEPTED as A DOGMA !!!!!
>
> The book was conceived in 1991, so by definition they would have been asked
> to write this after this point.
>
> Except that Britton and Walker were not discussing mycenaean or greek
> studies but Astrology and Astronomy in ancient mesopotamia
What gave them the right not to mention the works published in KADATH
and other journals ...
But *YOU* have not the same excuse !....
grapheus
Jim Webster
"grapheus" <grap...@www.com> wrote in message
> >
> > > Second Question : What is the DATE of the "Mesopotamian documents"
> > > quoted by J. Briton & C. Walker ?...
> > > (Answer, in the case you don't know it : they go back to the "time of
> > > Babylon", i.e. c. 1800 BC or later, what is approximately one
> > > MILLENIUM AFTER the "invention of the Constellations" !!!!!
> >
> > no, the earliest we have found that relate specifically to astronomical
> > observations go back to 1800BC.
>
> Please, EXPLAIN the difference with what I wrote and the change of a
> YES into your NO !!!!
the two sentences below do this adequately
>
> > Cuniform goes back to 3000BC. It is possible
> > that purely by chance we have found the oldest astronomical observation,
but
> > given the technical competence of it, this is exceedingly unlikely
> >
>
> Irrelevant !... The DATE of the "invention of the Constellations" is
> c. 2500BC, not 1800BC... Moreover, it has been shown that these
> Constellations were invented by a MARITIME people.
> But, of course, how could you know ?.. You refuse to read the
> referenced papers!...
You have referenced papers concerning the constellations being invented by
Maritime people. Obviously they are different constellations to those used
by the Greeks because theirs are known to have been taken from Mesopotamia.
Also the main use of constellations in this period was not navigation but
maintaining a calendar and astrology.
>
> >
> > > Third question : When J. Briton & C. Walker wrote this sentence ?..
> > > (Answer, in the case you don't know it : At the time when the
> > > "Risch-Chadwick Theory" was ACCEPTED as A DOGMA !!!!!
> >
> > The book was conceived in 1991, so by definition they would have been
asked
> > to write this after this point.
> >
> > Except that Britton and Walker were not discussing mycenaean or greek
> > studies but Astrology and Astronomy in ancient mesopotamia
>
> What gave them the right not to mention the works published in KADATH
> and other journals ...
I suggest you read the book yourself.
> But *YOU* have not the same excuse !....
I am not the one trying to brow beat people into accepting a dubious theory
with no argument more intellectually compelling than a jammed caps lock key.
Your case would be somewhat stronger if you actually quoted supporting
evidence from the work of more than one author
Jim Webster
grapheus
I am not surprised of that !... It's the USUAL WAY dogmatic people
act : "If you cannot disprove the theory, JUST IGNORE IT !"... It's
called in French "le Mur du Silence"...
You are not the first to use this trick. You will not be the last !..
Thanks for the demonstration !..
grapheus
Jim Webster
> I am not surprised of that !... It's the USUAL WAY dogmatic people
> act : "If you cannot disprove the theory, JUST IGNORE IT !"...
Actually it doesn't work like that. Those putting forward a new theory have
to provide evidence and defend it against questions.
Normally defence of a theory does not revolve round an inability to master
the shift key
Jim Webster
Peter T. Daniels
What is "chesco.com"? Doesn't look like a site for a Department of
Egyptology to me.
I don't click links, because I can't sit around waiting ten minutes for
them to start to appear, or for them to demand cookies, or for them to
turn out to be crackpots, or for them to crash Netscape or even freeze
the computer.
grapheus
> "grapheus" <grap...@www.com> wrote in message
> > >
> > > > Second Question : What is the DATE of the "Mesopotamian documents"
> > > > quoted by J. Briton & C. Walker ?...
> > > > (Answer, in the case you don't know it : they go back to the "time of
> > > > Babylon", i.e. c. 1800 BC or later, what is approximately one
> > > > MILLENIUM AFTER the "invention of the Constellations" !!!!!
> > >
> > > no, the earliest we have found that relate specifically to astronomical
> > > observations go back to 1800BC.
> >
> > Please, EXPLAIN the difference with what I wrote and the change of a
> > YES into your NO !!!!
>
> the two sentences below do this adequately
>
> >
> > > Cuniform goes back to 3000BC. It is possible
> > > that purely by chance we have found the oldest astronomical observation, but
> > > given the technical competence of it, this is exceedingly unlikely
> > >
> >
> > Irrelevant !... The DATE of the "invention of the Constellations" is
> > c. 2500BC, not 1800BC... Moreover, it has been shown that these
> > Constellations were invented by a MARITIME people.
> > But, of course, how could you know ?.. You refuse to read the
> > referenced papers!...
>
> You have referenced papers concerning the constellations being invented by
> Maritime people. Obviously they are different constellations to those used
> by the Greeks because theirs are known to have been taken from Mesopotamia.
That is what *YOU* say, without knowing the complete file !..
>
> Also the main use of constellations in this period was not navigation but
> maintaining a calendar and astrology.
>
Same remark as hereabove...
> >
> > >
> > > > Third question : When J. Briton & C. Walker wrote this sentence ?..
> > > > (Answer, in the case you don't know it : At the time when the
> > > > "Risch-Chadwick Theory" was ACCEPTED as A DOGMA !!!!!
> > >
> > > The book was conceived in 1991, so by definition they would have been asked
> > > to write this after this point.
> > >
> > > Except that Britton and Walker were not discussing mycenaean or greek
> > > studies but Astrology and Astronomy in ancient mesopotamia
> >
> > What gave them the right not to mention the works published in KADATH
> > and other journals ...
>
> I suggest you read the book yourself.
>
I already did about two years ago, when it was published, but I
certainly will read it again as soon as possible...
> > But *YOU* have not the same excuse !....
>
> I am not the one trying to brow beat people into accepting a dubious theory
> with no argument more intellectually compelling than a jammed caps lock key.
> Your case would be somewhat stronger if you actually quoted supporting
> evidence from the work of more than one author
>
I'm sorry, but the Proto-Ionian Theory has been proposed by one
scholar, with no publicity and against the "general opinion"... So,
it's easier to quote other scholars who, not even being aware of it,
have supported other theories (like the theory that the Greeks took
their astronomical knowledge from Mesopotamia) than scholars
supporting the theory... A second problem is that even when these
(rare) supporters are cited, like M. Ovenden for instance,
IGNARROGANTS don't pay any attention to the reference !...
grapheus
Brian M. Scott
<gram...@worldnet.att.net> wrote in
sci.archaeology,soc.history.ancient,sci.lang:
>> http://www.chesco.com/~cslice/aurora/rosetta/rosetta.html
It's a fairly harmless recitation of the history of the
decipherment of the Rosetta Stone mostly cribbed from articles in
the Britannica and Encarta. MG is lying, as usual: it makes no
mention of earlier 'decipherments'.
[...]
Brian
Jim Webster
well if they were not from Babylon then they have to be different, some
evidence of this would be interesting.
>
> >
> > Also the main use of constellations in this period was not navigation
but
> > maintaining a calendar and astrology.
> >
>
> Same remark as hereabove...
you have evidence of Constellations being used predominantly for navigation?
I would be interested to see it.
I never saw the reference, could you please repeat it
Jim Webster
grapheus
> "grapheus" <grap...@www.com> wrote in message
> news:337ae51f.04020...@posting.google.com...
>
> > I am not surprised of that !... It's the USUAL WAY dogmatic people
> > act : "If you cannot disprove the theory, JUST IGNORE IT !"...
>
>
> Actually it doesn't work like that. Those putting forward a new theory have
> to provide evidence and defend it against questions.
Yes. And those who want to critizise it have FIRST to READ ALL THE
ARGUMENTS, even if they have to do some lengthy bibliographical search
!..
grapheus
Jim Webster
"grapheus" <grap...@www.com> wrote in message
news:337ae51f.04020...@posting.google.com...
> "Jim Webster" <J...@feeswerve.spam.co.uk> wrote in message
news:<bvqqif$k8o$1...@news7.svr.pol.co.uk>...
> > "grapheus" <grap...@www.com> wrote in message
> > news:337ae51f.04020...@posting.google.com...
> >
> > > I am not surprised of that !... It's the USUAL WAY dogmatic people
> > > act : "If you cannot disprove the theory, JUST IGNORE IT !"...
> >
> >
> > Actually it doesn't work like that. Those putting forward a new theory
have
> > to provide evidence and defend it against questions.
>
> Yes. And those who want to critizise it have FIRST to READ ALL THE
> ARGUMENTS, even if they have to do some lengthy bibliographical search
> !..
I asked for sources and so far you have suggested just the one book by the
author of the theory which is hardly lengthy bibliographical searching
Jim Webster
>
> grapheus
grapheus
> "grapheus" <grap...@www.com> wrote in message
> news:337ae51f.04020...@posting.google.com...
> > "Jim Webster" <J...@feeswerve.spam.co.uk> wrote in message
> news:<bvqqif$k8o$1...@news7.svr.pol.co.uk>...
> > > "grapheus" <grap...@www.com> wrote in message
> > > news:337ae51f.04020...@posting.google.com...
> > >
> > > > I am not surprised of that !... It's the USUAL WAY dogmatic people
> > > > act : "If you cannot disprove the theory, JUST IGNORE IT !"...
> > >
> > >
> > > Actually it doesn't work like that. Those putting forward a new theory
> have
> > > to provide evidence and defend it against questions.
> >
> > Yes. And those who want to critizise it have FIRST to READ ALL THE
> > ARGUMENTS, even if they have to do some lengthy bibliographical search
> > !..
>
> I asked for sources and so far you have suggested just the one book by the
> author of the theory which is hardly lengthy bibliographical searching
>
If you would mind BEGINNING by reading the two books (not just one) I
asked you to read, you would have found ALL the needed references in
them !..
But as usual, IGNARROGANTS prefer endlessly discussing without knowing
the file !..
grapheus
Jim Webster
> asked you to read, you would have found ALL the needed references in
> them !..
>
to the best of my knowledge you mentioned only one, and that in French
Jim Webster
Bobby D. Bryant
> In any case, the question is obviously of no relevance whatsoever to the
> decipherment of the Phaistos Disk.
I'm so surprised!
Bobby D. Bryant
> Irrelevant !... The DATE of the "invention of the Constellations" is c.
> 2500BC, not 1800BC... Moreover, it has been shown that these
> Constellations were invented by a MARITIME people.
Even if you were right (for a change), as Miguel has already pointed out,
that wouldn't tell us diddley about what the disk says.
Bobby D. Bryant
> I'm sorry, but the Proto-Ionian Theory has been proposed by one
> scholar, with no publicity and against the "general opinion"...
Small surprise he's singing solo, given how irrelevant your favorite
supporting argument for his claims is.
grapheus
> On Tue, 03 Feb 2004 11:16:06 +0100, Miguel Carrasquer wrote:
>
> > In any case, the question is obviously of no relevance whatsoever to the
> > decipherment of the Phaistos Disk.
>
> I'm so surprised!
But *I* am not surprised to see your bad faith, by quoting Miguel
Carrasquer's remark WITHOUT MENTIONING my answer, saying that it is
TOTALLY WRONG by lack of information !..
grapheus
grapheus
Of course, not !.. But it would tell us a lot about the EXISTENCE
c.2500BC of PROTO-IONIANS in the Aegean !...
But you are of TOO MUCH BAD FAITH to recognize it, right?....
grapheus
grapheus
But there is no surprise to see YOU joining the pack of the IGNARROGANTS !
grapheus
Bobby D. Bryant
No, I'm just rational enough to see that even if the argument is correct,
it's not a "proof" of your interpretation of the disk.
So. What's your *second* best "proof"?
grapheus
No. I mentioned two books from J. Faucounau : "Les Proto-Ioniens :
Histoire d'un peuple oublié" and "Le dechiffrement of the Phaistos
Disk. Preuves et Consequences" . It's true that these books are in
French, as there is no traduction of them in English for the moment.
But in these two books, one may find a large bibliography, half in
English, half in French or German. There is generally not difficult to
read these two basic books in an University Library (as most
librarians know about their existence) and to take notice of the
Bibliography, or to buy them in any e-bookshop selling French books.
Finding all the books or papers cited is a more difficult task, I must
recognize it, and I've had personally some trouble to do it (I've had
sometimes to write to the author for a photocopy). The reason of that
is, that as these referenced papers were going against the "general
opinion", they have been printed in some "half-confidential journals".
Such is, for instance, the very important Michael W. Ovenden's paper
on "The origin of the Constellations", published in "The Philosophical
Journal" of Jan. 1966 ... Some University Libraries have it, others
don't... Other essential papers have never been printed, like
A.Goetze's "History of the Arzawan Kingdom", REFUSED by C. Gadd for
the "Cambridge Ancient History" because this study of some 25 Hittite
unpublished Tablets was showing GREEK NAMES in Western Anatolia around
1900BC (what was contradicting the "Risch-Chadwick Theory), and lost
since...
grapheus
Hagen
>>judgements like the hereabove!...
>>
>>In this, he is guilty, I agree !..
>
>Let me quote p. 12: "Pour des raisons dues à l'histoire du déchiffrement,
>la plupart de ces preuves ont déjà été publiées, mais de façon éparse, dans
>des revues diverses. Un ouvrage reprenant l'ensemble du dossier, de
>manière à le soumettre au feu de la critique scientifique s'imposait donc.
>C'est cet ouvrage que nous proposons aujourd'hui."
>
>Well, if you don't give the details about the decipherment, and then only
>the weakest of summaries of the other "proofs", then I'd say you have
>failed spectacularly to "submit yourself to scientific criticism". The
>only appropriate scientific response is to ignore Faucounau's
>"decipherment". Which is what I'm going to do now.
>
>=======================
>Miguel Carrasquer Vidal
>m...@wxs.nl
Don't give up Miguel Carrasquer Vidal, if this topiic really takes
your interest.
Now when you too have had a lesson in absurdity, don't you think, you
should go to my homepage, and make your own judgement of my
enumerations, and you'll realise: the inscription in question was
solved twenty years ago, as an obvious prehistoric calendar.
Do me the favour, please!
http://home.gvdnet.dk/~hagen/phaistos.htm
Best wishes
Ole Hagen
grapheus
> On Thu, 05 Feb 2004 01:14:44 -0800, grapheus wrote:
>
> > "Bobby D. Bryant" <bdbr...@mail.utexas.edu> wrote in message news:<pan.2004.02.05....@mail.utexas.edu>...
> >> On Wed, 04 Feb 2004 01:38:24 -0800, grapheus wrote:
> >>
> >> > Irrelevant !... The DATE of the "invention of the Constellations" is c.
> >> > 2500BC, not 1800BC... Moreover, it has been shown that these
> >> > Constellations were invented by a MARITIME people.
> >>
> >> Even if you were right (for a change), as Miguel has already pointed out,
> >> that wouldn't tell us diddley about what the disk says.
> >
> > Of course, not !.. But it would tell us a lot about the EXISTENCE
> > c.2500BC of PROTO-IONIANS in the Aegean !...
> > But you are of TOO MUCH BAD FAITH to recognize it, right?....
>
> No, I'm just rational enough to see that even if the argument is correct,
> it's not a "proof" of your interpretation of the disk.
>
But it shows that ONE UNEXPECTED CONSEQUENCE of the J.F.'s
decipherment is VERIFIED BY FACTS having nothing to do with the Disk
itself !.. I say "UNEXPECTED" because when J.F. (not me!) has
reached the decipherment in the 70ies, the "general opinion" was that
the existence of a Proto-Ionic dialect c.1650 "WAS TOTALLY IMPOSSIBLE"
(See the "Risch-Chadwick Theory")... To the point that if numerous
would-be-decipherers have proposed a GREEK SOLUTION since 1950, NONE
of them (but J.F. , lead by statistical reasons) has supposed that
the dialect might be "Proto-Ionic" !...
> So. What's your *second* best "proof"?
The "Second Best" Proofs are a)- the "Linguistical Proof", which
shows the EXISTENCE of numerous "Proto-Ionisms" in the Mycenaean
Tablets, "Proto-Ionisms" which have the SAME PHONETICAL
CHARACTERISTICS as the Disk's language found by the decipherment b)-
the "epigraphical proofs" which shows that ALL the "epigraphical
anomalies" on the Disk can be EXPLAINED in a very satisfactory and
logical way by the decipherment (Why he obliterated THIS sign, why the
signs are crowded here and not there, etc.).
grapheus
grapheus
Don't worry, Ole !... Miguel Carrasquer Vidal *IS* convinced that the
Disk *IS* a Minoan Calendar...
He has NEVER seen the difference between an hypothesis having one out
of a million chances to be right and one having one out of a million
chances to be wrong... And he HATES the word "PROOFS"!
grapheus
Peter T. Daniels
I think he was hoping for more than the _names_ of the proofs ...