With all the knowledge we now possess about human language, all the data on the languages of the world, all the great minds in the field, the computing power, the rise of AI, etc... would it be fair to say, short of finding a Rosetta Stone II, that any undeciphered language, text, or script is beyond recovery at this point? Or does the field have yet room to evolve and crack those codes even with no further material unearthed?
Something I thought about the other day.
Are yet undeciphered languages / texts hopeless by 2022?
- outcast
- Blue Belt
- Posts: 585
- Joined: Sat Dec 05, 2015 3:41 pm
- Location: Florida, USA
- Languages: ~
FLUENCY
Native: ENglish, ESpañol
Advanced: -
High Basic: DEutsch (rust), FRançais (rust), ZH中文
Basic: -
~
ACQUIRING
Formally: KO한국말, ITaliano, HI हिन्दी
Dabbling: HRvatski, GW粵語
Dormant: POrtuguês
~
Plan to learn: I BETTER NOT GO HERE FOR NOW
~ - x 679
Are yet undeciphered languages / texts hopeless by 2022?
1 x
"I can speak wonderfully and clearly in zero languages, and can also fluently embarrass myself in half a dozen others."
The End of Language learning:
The End of Language learning:
-
- Blue Belt
- Posts: 520
- Joined: Thu Jul 23, 2015 3:21 pm
- Languages: English (N), German, Japanese, Mandarin, Korean
- x 1415
Re: Are yet undeciphered languages / texts hopeless by 2022?
AI is just getting started. There's still hope.
1 x
- Iversen
- Black Belt - 4th Dan
- Posts: 4787
- Joined: Sun Jul 19, 2015 7:36 pm
- Location: Denmark
- Languages: Monolingual travels in Danish, English, German, Dutch, Swedish, French, Portuguese, Spanish, Catalan, Italian, Romanian and (part time) Esperanto
Ahem, not yet: Norwegian, Afrikaans, Platt, Scots, Russian, Serbian, Bulgarian, Albanian, Greek, Latin, Irish, Indonesian and a few more... - Language Log: viewtopic.php?f=15&t=1027
- x 15049
Re: Are yet undeciphered languages / texts hopeless by 2022?
In other areas modern technique used by humans has wrought miracles. For instance the Lidar system has demonstrated that there are far more Mayan ruins buried under the jungle than anyone in their wildest moments had expected, and the use of genetical information has wrought havoc in zoological taxonomy. In cases where there are enough data to work on and some assumptions about the way languages work I could see AI play a similar role in code breaking. But AI works on the basis of learning by trial and error, and if success (the opposite of error) is that a human thinks that a translation is meaningful then I suppose that human involvement still would be necessary.
Some old writing systems may be so scantily represented that they are beyond repair. But I wouldn't exclude the possibility that for instance the Indus writing system could be cracked, and that AI could be part of the process. My hunch is that the languages behind it could a precursor for the Dravidian languages, but even if it isn't all hope isn't gone. Sumerian can be read today even though it is a linguistic isolate with a fiendishly complicated Cuneiform writing system. The problem is something like the Phaistos stone, which only is found on one single archeological item. Unless someone comes up with a really good idea about the hidden language behind those signs I can't see how that specific writing system can be cracked. As for linear A ... well, I wouldn't call it hopeless, but we still don't know which kind of language the Minoans spoke. The weak point could be that the Myceneans - who spoke a very oldfashioned kind of Greek - took over the writing system when they took over the island Crete, and it's unlikely that they didn't reuse at least some of the phonetic values of the signs - like the Akkadians adopted the Cuneiform writing invented by the Sumerians for a completely unrelated language for which it wasn't really suited.
Some old writing systems may be so scantily represented that they are beyond repair. But I wouldn't exclude the possibility that for instance the Indus writing system could be cracked, and that AI could be part of the process. My hunch is that the languages behind it could a precursor for the Dravidian languages, but even if it isn't all hope isn't gone. Sumerian can be read today even though it is a linguistic isolate with a fiendishly complicated Cuneiform writing system. The problem is something like the Phaistos stone, which only is found on one single archeological item. Unless someone comes up with a really good idea about the hidden language behind those signs I can't see how that specific writing system can be cracked. As for linear A ... well, I wouldn't call it hopeless, but we still don't know which kind of language the Minoans spoke. The weak point could be that the Myceneans - who spoke a very oldfashioned kind of Greek - took over the writing system when they took over the island Crete, and it's unlikely that they didn't reuse at least some of the phonetic values of the signs - like the Akkadians adopted the Cuneiform writing invented by the Sumerians for a completely unrelated language for which it wasn't really suited.
8 x
- leosmith
- Brown Belt
- Posts: 1353
- Joined: Thu Sep 29, 2016 10:06 pm
- Location: Seattle
- Languages: English (N)
Spanish (adv)
French (int)
German (int)
Japanese (int)
Korean (int)
Mandarin (int)
Portuguese (int)
Russian (int)
Swahili (int)
Tagalog (int)
Thai (int) - x 3158
- Contact:
Re: Are yet undeciphered languages / texts hopeless by 2022?
Too bad this isn't a poll.
I'm going to vote "no". AI is estimated to be as smart as humans in just about every way by around 2050, and will continue to grow smarter at an accelerated pace. I believe it will crack at least some of the remaining indecipherable scripts. Some will never be cracked though because there isn't enough text.
I'm going to vote "no". AI is estimated to be as smart as humans in just about every way by around 2050, and will continue to grow smarter at an accelerated pace. I believe it will crack at least some of the remaining indecipherable scripts. Some will never be cracked though because there isn't enough text.
2 x
https://languagecrush.com/reading - try our free multi-language reading tool
- zenmonkey
- Black Belt - 2nd Dan
- Posts: 2528
- Joined: Sun Jul 26, 2015 7:21 pm
- Location: California, Germany and France
- Languages: Spanish, English, French trilingual - German (B2/C1) on/off study: Persian, Hebrew, Tibetan, Setswana.
Some knowledge of Italian, Portuguese, Ladino, Yiddish ...
Want to tackle Tzotzil, Nahuatl - Language Log: viewtopic.php?f=15&t=859
- x 7032
- Contact:
Re: Are yet undeciphered languages / texts hopeless by 2022?
The answer is maybe.
Goedel’s incompleteness proof is pretty devastating.
Goedel’s incompleteness proof is pretty devastating.
2 x
I am a leaf on the wind, watch how I soar
- rdearman
- Site Admin
- Posts: 7260
- Joined: Thu May 14, 2015 4:18 pm
- Location: United Kingdom
- Languages: English (N)
- Language Log: viewtopic.php?f=15&t=1836
- x 23317
- Contact:
Re: Are yet undeciphered languages / texts hopeless by 2022?
zenmonkey wrote:The answer is maybe.
Goedel’s incompleteness proof is pretty devastating.
Goedel’s incompleteness proof ... a nice diversion to read up on. Thanks.
1 x
: Read 150 books in 2024
My YouTube Channel
The Autodidactic Podcast
My Author's Newsletter
I post on this forum with mobile devices, so excuse short msgs and typos.
My YouTube Channel
The Autodidactic Podcast
My Author's Newsletter
I post on this forum with mobile devices, so excuse short msgs and typos.
- zenmonkey
- Black Belt - 2nd Dan
- Posts: 2528
- Joined: Sun Jul 26, 2015 7:21 pm
- Location: California, Germany and France
- Languages: Spanish, English, French trilingual - German (B2/C1) on/off study: Persian, Hebrew, Tibetan, Setswana.
Some knowledge of Italian, Portuguese, Ladino, Yiddish ...
Want to tackle Tzotzil, Nahuatl - Language Log: viewtopic.php?f=15&t=859
- x 7032
- Contact:
Re: Are yet undeciphered languages / texts hopeless by 2022?
rdearman wrote:zenmonkey wrote:The answer is maybe.
Goedel’s incompleteness proof is pretty devastating.
Goedel’s incompleteness proof ... a nice diversion to read up on. Thanks.
When I bought my copy, the book was a misprint and the last half was missing.
5 x
I am a leaf on the wind, watch how I soar
-
- Black Belt - 3rd Dan
- Posts: 3534
- Joined: Thu Jul 30, 2015 11:04 am
- Location: Scotland
- Languages: English(N)
Advanced: French,Spanish, Scottish Gaelic
Intermediate: Italian, Catalan, Corsican
Basic: Welsh
Dabbling: Polish, Russian etc - x 8810
- Contact:
Re: Are yet undeciphered languages / texts hopeless by 2022?
Humans have a tendency to discover truths, then apply them falsely, because we're not capable of processing the amount of data that the real truth requires.
For example: behaviorism. Behaviorism is a school of philosophy that says human thought is a matter of learning and associating inputs with outputs, and being conditioned to certain behaviours.
This is probably true (with the caveat that there are certain biologically-mediated mechanisms in the brain) but the problem with behaviorism was that it tried to deal with behaviours at too coarse a level. Any visible human behaviour is really the interaction of multiple units of behaviour. It's immensely complex, so psychology has gone back to more subjective models, and they're better.
Machine learning is starting to extrapolate more variables of human behaviour than psychologists can measure. Computers can reasonably accurately estimate ages, hobbies, professions, physical health etc from buckets of unsorted user data. Humans can't.
So far, so off-topic.
Humans have for generations tried to decipher texts by making certain assumptions and trying to see if they fit. This isn't an unreasonable approach per se... the problem is that it leaves us with one arbitrary hypothesis and no way of measuring of evaluating it. A human simply isn't capable of making enough hypotheses and evaluating them objectively in order to prove any of them.
Computers can create, evaluate, compare and adapt multiple models/hypotheses. They're not limitless, though, and currently that means starting with human intuition, leaving us halfway between human subjectivity and computer objectivity.
Computer capacity is growing, though, and soon I think there will be a fair chance of computers identifying patterns that we miss, and one day they'll supply a set of explanations for Indus Valley etc that will just look obvious in hindsight.
For example: behaviorism. Behaviorism is a school of philosophy that says human thought is a matter of learning and associating inputs with outputs, and being conditioned to certain behaviours.
This is probably true (with the caveat that there are certain biologically-mediated mechanisms in the brain) but the problem with behaviorism was that it tried to deal with behaviours at too coarse a level. Any visible human behaviour is really the interaction of multiple units of behaviour. It's immensely complex, so psychology has gone back to more subjective models, and they're better.
Machine learning is starting to extrapolate more variables of human behaviour than psychologists can measure. Computers can reasonably accurately estimate ages, hobbies, professions, physical health etc from buckets of unsorted user data. Humans can't.
So far, so off-topic.
Humans have for generations tried to decipher texts by making certain assumptions and trying to see if they fit. This isn't an unreasonable approach per se... the problem is that it leaves us with one arbitrary hypothesis and no way of measuring of evaluating it. A human simply isn't capable of making enough hypotheses and evaluating them objectively in order to prove any of them.
Computers can create, evaluate, compare and adapt multiple models/hypotheses. They're not limitless, though, and currently that means starting with human intuition, leaving us halfway between human subjectivity and computer objectivity.
Computer capacity is growing, though, and soon I think there will be a fair chance of computers identifying patterns that we miss, and one day they'll supply a set of explanations for Indus Valley etc that will just look obvious in hindsight.
7 x
-
- Brown Belt
- Posts: 1125
- Joined: Mon Jul 20, 2015 12:44 am
- Languages: English (N), Irish (Teastas Eorpach na Gaeilge B2), French, dabbling elsewhere sometimes
- Language Log: viewtopic.php?f=15&t=757
- x 3364
Re: Are yet undeciphered languages / texts hopeless by 2022?
Cainntear wrote:Computer capacity is growing, though, and soon I think there will be a fair chance of computers identifying patterns that we miss, and one day they'll supply a set of explanations for Indus Valley etc that will just look obvious in hindsight.
I've had semi-thoughts about trying to do something like this with what we have recorded of Iberian and Trasks's etymological dictionary of Proto-Basque to see if a computer could detect any ready correspondences that could allow us to them translate the Iberian texts (and thus pretty firmly prove they're related). Makes me wonder if anything has been tried yet.
1 x
-
- Black Belt - 3rd Dan
- Posts: 3534
- Joined: Thu Jul 30, 2015 11:04 am
- Location: Scotland
- Languages: English(N)
Advanced: French,Spanish, Scottish Gaelic
Intermediate: Italian, Catalan, Corsican
Basic: Welsh
Dabbling: Polish, Russian etc - x 8810
- Contact:
Re: Are yet undeciphered languages / texts hopeless by 2022?
galaxyrocker wrote:Cainntear wrote:Computer capacity is growing, though, and soon I think there will be a fair chance of computers identifying patterns that we miss, and one day they'll supply a set of explanations for Indus Valley etc that will just look obvious in hindsight.
I've had semi-thoughts about trying to do something like this with what we have recorded of Iberian and Trasks's etymological dictionary of Proto-Basque to see if a computer could detect any ready correspondences that could allow us to them translate the Iberian texts (and thus pretty firmly prove they're related). Makes me wonder if anything has been tried yet.
It'll need a lot of data from known languages -- the idea would be to have a computer that "knows" more languages than a human can, so would have a map of all the possibilities of what can actually happen.
It would take an operation the size of Google to do such work, I reckon.
4 x
Return to “General Language Discussion”
Who is online
Users browsing this forum: No registered users and 2 guests