Humanity must assume that the codex has a single, incontrovertible meaning. What throws me off is when you say that we can start with a single string that can have that meaning. Absent a full translation, and absent a hint such as pictograms or numbers, the only way to verify that a single string is decoded correctly is if that string is part of a larger body of text that is decoded correctly. For instance, of the first ten pages read like alien War and Peace, or highly coherent scientific or philosophical discourse, it is somewhat probable that each of the strings is decoded correctly. If the grammar used to decode the first ten pages is applied successfully to more pages, the likelihood grows. But a single string? I don't see how you can get anywhere with that.

That is to say that if we could, across the distribution of meanings the codex could take on, narrow down the likelihoods of certain interpretations over others, there is probably one that is most likely, although I don't know to what degree, or what degree to which it would have to be the case to be considered the correct interpretation. — ToothyMaw

I think this is right, and I think cohernece is the only criteria we can use to decide likelihood (the fact that these are aliens means we have to make huge allowances for things that don't make sense, which makes this evaluation much more difficult).

Humanity must assume that the codex has a single, incontrovertible meaning. — hypericin

I don't understand this assumption. Does every novel have a single incontrovertible meaning? Take for instance idioms/metaphors, which bring forth the issue/conflict of literal versus figurative meanings. Both the following passages are coherent on two levels (?), but they have two different meanings based on whether or not you have knowledge of what the idiomatic content actually means.

I decided to bite the bullet and hit the road early, hoping to beat the clock, but when push came to shove, traffic was a whole different ball game. By the time I made it to the office, I was running on fumes, yet I still had to jump through hoops to get the project off the ground. At the end of the day, though, we pulled it off by the skin of our teeth. — ChatGPT paragraph in Idioms

It gets even more bizarre if you translate foreign idioms:

I woke up feeling like I had an octopus on my face, but I decided to tie my stomach and head to work. The meeting was chaos — everyone was watering their salad while the boss was trying to give birth to a mountain. When it was my turn to speak, I almost dropped my face, but somehow I managed to hang noodles on everyone’s ears. By the end, we were all pressing the cucumber, pretending everything was fine. — ChatGPT paragraph in foreign idioms

If alien codex were an idiomatic prank that was deciphered at a literal level, the meaning would still be lost. This would be compounded by the gulf between what is universal between species and what is hopelessly local and perhaps untranslatable.

I don't understand this assumption. Does every novel have a single incontrovertible meaning? — Nils Loc

Really I should have said "translation", not "meaning". And it is true, not every earth-language translation is the same. What I really meant was, the assumption has to be that the thing isn't War and Peace (in spaaace) and a dietetic guidebook.

Humanity must assume that the codex has a single, incontrovertible meaning. What throws me off is when you say that we can start with a single string that can have that meaning. — hypericin

Where did I say that? I suppose that my method would, ideally, approach creating a single string of meaning if it were applied over and over again, but I don't think we start with that translation or that it would be absolutely incontrovertible. Furthermore, it could arise out of analysis of the coherence of various possible translations.

So yes, given enough time and computing power, a meaning can be imposed on the codex, I think. — ToothyMaw

Couldn't it be possible that there are actually hundreds to billions of variations of meaning that can be imposed on the codex that satisfy the level of coherence hypericin/humanity is looking for. If this was known to be the likelihood, the meaning of any can be disputed within/against that set of all possibilities. What exactly makes the manufactured meaning of the text incontrovertible? Are we assuming only one meaning can fit the codex?

So yes, given enough time and computing power, a meaning can be imposed on the codex, I think.
— ToothyMaw

Couldn't it be possible that there are actually hundreds to billions of variations of meaning that can be imposed on the codex that satisfy the level of coherence hypericin/humanity is looking for. If this was known to be the likelihood, the meaning of any can be disputed within/against that set of all possibilities. What exactly makes the manufactured meaning of the text incontrovertible? Are we assuming only one meaning can fit the codex? — Nils Loc

Like I said in an earlier post:

I would say that any endeavor to interpret the text in a meaningful way probably has to assume that the codex could theoretically have a discoverable, incontrovertible meaning, even if it cannot possibly be truly identified - because it is the limiting case.

Thus, even if we cannot say there is definitely an incontrovertible meaning, I would say that we can approach it from a probabilistic standpoint that might get us close to virtual incontrovertibility. That is to say that if we could, across the distribution of meanings the codex could take on, narrow down the likelihoods of certain interpretations over others, there is probably one that is most likely, although I don't know to what degree, or what degree to which it would have to be the case to be considered the correct interpretation. — ToothyMaw

Couldn't it be possible that there are actually hundreds to billions of variations of meaning that can be imposed on the codex that satisfy the level of coherence hypericin/humanity is looking for. If this was known to be the likelihood, the meaning of any can be disputed within/against that set of all possibilities. — Nils Loc

Good point. If one coherent (whatever that means) interpretation can be produced it seems likely innumerable can be. This will call the legitimacy of all of them into question. There might be advocates of each of them.

This is one logical outcome. However I still intuitively feel that no coherent (whatever that means) translation can ever be produced.

That is to say that if we could, across the distribution of meanings the codex could take on, narrow down the likelihoods of certain interpretations over others, there is probably one that is most likely — ToothyMaw

The likelihood of arriving at one meaning might be a consequence of how difficult it is to make the codex coherent though. If you had the set of all possible coherent meanings, which might be numerically staggering, what exactly would help you to pick the "one that is most likely"?

That is to say that if we could, across the distribution of meanings the codex could take on, narrow down the likelihoods of certain interpretations over others, there is probably one that is most likely
— ToothyMaw

The likelihood of arriving at one meaning might be a consequence of how difficult it is to make the codex coherent though. If you had the set of all possible meanings, which might be numerically staggering, what exactly would help you to pick the "one that is most likely"? — Nils Loc

We could just do rote textual analysis by a reader, I guess. Although, that is hardly feasible given the potential multitudes of valid meanings, so I guess we would need some sort of efficient process or algorithm or something. I'll get back to you on that.

I suppose the one that is most likely would have to be the one that gets the closest to being incontrovertible. Every meaning imposed on the codex could be measured against this standard - the limiting case. It would be like funneling everything towards a limit and seeing how close the interpretations get to that limit.

It would be kind of like stipulating: "only really big masses can balance this scale" and then measuring various masses on a scale until we find one that gets the closest to balancing the scale and then saying that that mass qualifies as being the closest to being really big.

So, to make it as clear as possible, that means that only an incontrovertible meaning has a 100% chance of being the correct meaning, and every other interpretation has a chance of being correct that aligns with a probability assigned according to how close it is to being incontrovertible.

I guess I'm still confused as to why one would make the assumption of incontrovertible meaning.

Suppose there are 10 different civilizations similar to ours in their intelligence/knowledge/life, that all receive the same hoax codex, and the syntactical nature of the codex serves perfectly as any language emptied of original meaning might. Each of these civilizations go to work at imposing meaning onto the script in a way that achieves a compelling level of coherence such that they have, in their expert opinion, reached a stage of incontrovertible meaning, which really just means they've achieved a remarkable coherence/intelligibility that seems indisputable.

What is the likelihood that the meaning of these 10 different efforts in different parts of the universe yield the same understanding? My intuition is that every completed codex would be radically different in meaning, yet perfectly intelligible and complete. The attitude that forms as to why the text's meaning is incontrovertible comes simply from the fact that it is way too difficult to try again afresh on any planet. Therefore there is no absolute incontrovertible meaning of any version, except with regard to all the work already done. It is only deemed incontrovertible because the meaning created "out of whole cloth" works but that fails to take in mind what else could work.

Is there any way we can ground our speculation as to whether there are many possible perfect impositions of meaning of or just a few or only one that works for the codex?

So, to make it as clear as possible, that means that only an incontrovertible meaning has a 100% chance of being the correct meaning, and every other interpretation has a chance of being correct that aligns with a probability assigned according to how close it is to being incontrovertible. — ToothyMaw

"Incontrovertible" seems far from a rigorous, objective term. It is a "know it when I see it" kind of thing. At one end are completely coherent novels, or the musings of an alien Aristotle. At the other end is gibberish. But between them is a whole hazy spectrum of material that kind of makes sense, if you squint hard enough, make ample allowances for alien references and ways of thinking, and don't pay too much attention to all the contradictions. I suspect that something along these lines would be the best case scenario. Here, one person's "incontrovertible" is another's "horseshit".

But that is only half the problem. The other half is the method the transition was achieved. You can imagine a perfectly ad hoc method, like, "XYZ means ABC, when seen on page one". This might yield an "incontrovertible" text: "One million moons ago our 12-eyed ancestors first descended from the trees...", but that is meaningless because the method was bullshit. On the other end, you can have a beautiful, logical grammar. Again, in between these two lies a spectrum of complications, exceptions, and hacks.

Both translation and method have to be evaluated, not one or the other.

Is there any way we can ground our speculation as to whether there are many possible perfect impositions of meaning of or just a few or only one that works for the codex? — Nils Loc

Or none. But that is the question. Is there a linguist in the house?

"Incontrovertible" seems far from a rigorous, objective term. It is a "know it when I see it" kind of thing. At one end are completely coherent novels, or the musings of an alien Aristotle. At the other end is gibberish. But between them is a whole hazy spectrum of material that kind of makes sense, if you squint hard enough, make ample allowances for alien references and ways of thinking, and don't pay too much attention to all the contradictions. I suspect that something along these lines would be the best case scenario. Here, one person's "incontrovertible" is another's "horseshit". — hypericin

I came up with a semi-rigorous way of defining incontrovertibility: if a translation can be modeled by a one-dimensional string or series of strings that do have an incontrovertible meaning, and the linguistic content of the translation would be correct only in the case that the content of that one-dimensional string is 100% correct or realized, then it could be an incontrovertible interpretation. Other interpretations would have probabilities of being correct associated with the likelihood of the one-dimensional strings modelling them being correct or realized both generally and with reference to modeling the text itself in a coherent way.

The method behind finding these translations is beyond me.

I hate to frustrate you, but I'm just not following you here. Maybe eli5?

It would probably be a good thing for us as a species.
Even though it's wasted effort, the feeling that there are important secrets that we might figure out soon might make us more aspirational and hopeful.

I've sometimes had the same thought with [possible hijack incoming] the idea of aliens coming to earth, deciding we aren't that interesting, and leaving immediately, for good. As frustrating as it would be, human progress would skyrocket. Knowing that interstellar travel is doable, practically, and that there are species out there would be tantalyzing.

Think Maw is just considering translation from an insufficient sample of text with known (incontrovertible) meaning.

The Rosetta stone would probably be an interesting case to read up on. Modern day Coptic was a vital source for deciphering Egyptian hieroglyphics because of the strong phonetic correspondences between the two languages. If they had the Rosetta stone but spoken Coptic was extinct, would they still be able to crack the hieroglyphics? Possibly not. Coptic furnishes most of the clues to reconstructing the meaning that the Rosetta stone translation does not contain.

Trying to reconstruct a foreign dictionary with just a handful of entries sounds impossible and absurd, as would be finding meaning in the alien codex.

Good point. If one coherent (whatever that means) interpretation can be produced it seems likely innumerable can be. This will call the legitimacy of all of them into question. There might be advocates of each of them.

This is one logical outcome. However I still intuitively feel that no coherent (whatever that means) translation can ever be produced. — hypericin

The largest outcome naturally is that we aren't alone as a 512-page book with obscure writing doesn't accidentally form just by accident in the universe. The real problem simply is that there's no way of knowing just what "the book" is about or what it is meant for. It can look like to us as a book, but that is the only thing we understand. We can just guess and this makes cracking of any code difficult.

Indeed we had to have the Rosetta stone to finally crack the ancient hieroglyphs. Even before we could assume what they were telling: praising the greatness of the Pharaohs etc. What else do you write in Temples etc? In this case, people would be having argument on just what is the whole function of the "book".

Indeed we had to have the Rosetta stone to finally crack the ancient hieroglyphs. Even before we could assume what they were telling: praising the greatness of the Pharaohs etc. What else do you write in Temples etc? In this case, people would be having argument on just what is the whole function of the "book". — ssu

Yes, this is the other side of the coin that I don't think has been mentioned yet. It may be that even if the contents were perfectly meaningful, we would never be able to crack it.

I think if this book arrived today, this would be the case. We would need a Rosetta Stone, or something , to assist, beyond the text itself. This is why I appealed not just to the focused effort of all the worlds linguists, but to transhuman AI, and fully matured quantum computers that could evaluate millions of grammars in a second. Would this be enough? Not sure!

Think Maw is just considering translation from an insufficient sample of text with known (incontrovertible) meaning. — Nils Loc

But the core premise is that there is no meaning at all in the text.

[

Trying to reconstruct a foreign dictionary with just a handful of entries sounds impossible and absurd, as would be finding meaning in the alien codex. — Nils Loc

Perhaps. But it would be interesting to see how the Coptic/Egyptian case, minus the Coptic, would have played out, had even today's technology been available. With far future technology, maybe decoding even a (meaningful) alien text would be tractable. It is, after all, ultimately a computational problem: there are only so many grammars that can produce the text, and only so many meanings that can be assigned to individual semantic units (though the alien factor would certainly compound the problem, there are certainly many alien meanings that would have no earthly correspondence and would be impossible to anticipate. There would have to be allowance for a fraction of words with unknowable meanings).

I hate to frustrate you, but I'm just not following you here. Maybe eli5? — hypericin

Think Maw is just considering translation from an insufficient sample of text with known (incontrovertible) meaning.
— Nils Loc

But the core premise is that there is no meaning at all in the text. — hypericin

@Nils Loc basically has it. I am suggesting we use a string that is incontrovertible in meaning (yet meaningful independent of any meaning we might assign to the codex) to scaffold interpretations. To start, we would need to find a string that is both incontrovertible in meaning and can model the codex. By "model the codex" I mean for the string and the codex to exist such that they are arranged in an identical combination of characters (whatever they might actually look like or represent for each). Then, if this string is both incontrovertible in meaning and the content of a particular interpretation of the codex hinges on the content of this string being absolutely confirmed in reality, much like a common proposition might be considered to be true, then this interpretation is potentially making a coherent statement about reality by virtue of being both semantically and materially meaningful.

This would work because there is a sort of interface between the meaning of the string and that of the codex that gives an interpretation an indisputable meaning in a virtual sense. I don't know if that qualifies as real incontrovertibility, though.

By "model the codex" I mean for the string and the codex to exist such that they are arranged in an identical combination of characters (whatever they might actually look like or represent for each). — ToothyMaw

So if in the codex we encounter,

:smile: :hearts: :smirk: :point: :lol: :wink: :nerd: :love: :roll: :monkey: :nerd:

Are you proposing we can map this to, say, "Dogs are Cute", and then proceed from there, with more mappings?

If not, please ground with a simple example.

Yes, that mapping works, but the process would be more like modeling your string of emojis after an interpretation that says "Dogs are Cute" - although it could be done this way too, I suppose. Furthermore, if that string of emojis were to actually express, say, that "all four-legged animals that bark are cute" in emojis, then we have pretty much successfully executed what I have described and can proceed with more mappings if we so desire.

@hypericin

On average the number of unique words used only once in any written work is 40-60%. This is a problem for translators if they don't have other works in which the same words appear to help them infer meaning.

An interesting easy exercise would be to scrub any book of its hapax legomena (unique words that appear only once) and see how much meaning is lost for the reader. How much work does the remaining context do to interpret the missing 40-60%?

If the alien codex was actually a version of English gibberish with fine sytnax and was entirely original (had no other copy or translation on Earth), even with a known sentence with incontrovertible meaning, I still believe it's fully untranslatable. The ratio of known meaning to unknown is really vital to the possibility of deciphering/translating language.

J.L. Borges wrote a story inspired by the thought experiment of the set of all possible books given a certain text length and symbol set. The combination output exceeds the estimated number of atoms in the universe and that can easily grow (exponentially) by increasing the length of text and symbol set. I've always wanted to know about specific qualitative sets within the space of all possible books given those stipulations. Using the English alphabet, what percent of the set of all possible books would be complete and comprehensible for any reader today? These question is unanswerable but I intuit the proportion is tiny, maybe the number of atoms in the solar system or galaxy out of the number of atoms in the universe. The mystery makes for an itch that can't be scratched.

How big would be set of the translation variants of Moby Dick in English? Can we replace the whale with a small land animal and consider it a variant of Moby Dick?

Using the English alphabet, what percent of the set of all possible books would be complete and comprehensible for any reader today? — Nils Loc

I think this is far more answerable than my question.

Assume the set of all 100 page books, 1500 characters per page. Ignore punctuation.

In the random case, that is 26^150000, or 10^212246, vastly, vastly larger than the number of atoms in the observable universe , just 10^80!

Now the coherent subset. Assume average word length of 5 words. Thats about 150000/5 = 25000 words. Given a incomplete text in a natural language, there are roughly, on average, 35 plausible word choices that may follow (according to chatgpt, who would know!). So roughly, that's about 35 ^ 25000, or 10^38602 "coherent" books (I suspect this is generous).

That still dwarfs the number of atoms in the universe, but is utterly dominated by the number of random texts. If the number of possible books was represented by all the atoms in the universe, the number of coherent books would be far, far, far, far less than one atom's worth!

That still dwarfs the number of atoms in the universe, but is utterly dominated by the number of random texts. If the number of possible books was represented by all the atoms in the universe, the number of coherent books would be far, far, far, far less than one atom's worth! — hypericin

:up:

Good luck on gaining any insight into your original problem. Let me know when you've figured it out. :sweat:

I realized I did a horrible job conveying just how tiny the coherent subset of every possible book is. One atom vs. the whole observable universe doesn't nearly do it justice.

Every possible book is represented by all 10^80 atoms in a special universe,
Where every single atom contains a sub-universe containing 10^80 atoms,
And each of those atoms contains a sub-universe containing 10^80 atoms,
And each of those atoms contains a sub-universe containing 10^80 atoms...
(repeat this 2167 more times)
Then one of those atoms represents the coherent subset.

Even though, to enumerate the coherent subset, one atom per book, you would need a nested universe like this, 482 layers deep!

And of course, that is just for paltry 100 pagers. It gets much, much worse as the page count goes up...