Ars speaks with a linguist about the ease with which Grace and Rocky communicate.
See full article...
See full article...
Project Hail Mary is in theaters—but do the linguistics work?
- Thread starter JournalBot
- Start date
jonbob_newcastle
Ars Scholae Palatinae
It’s not contextual, the two gestures are different. The insulting gesture is palm and thumb inward, the peace sign is palm out. But the thing is that it’s not as forceful as “fuck you”, more “up yours”. We tend to use the middle finger when we want to be really insulting.
The most common explanation for giving two fingers is probably apocryphal but the gist is that it was a taunt to the French from English archers who are demonstrating that they can still use a bow and arrow. At the time, the English were known for their skill with a longbow.
Upvote
1
(1
/
0)
I think the first starting point is to realize they're two people in the same place - if you're in a spaceship around a planet, and an alien is in their spaceship around the same planet, you have a starting set of assumptions that you could build off of. First, that you have enough science and engineering background to understand things like the basic composition of the universe (the elements), that you have some way of navigating around and your spaceship is likely an environment similar to your home. You will likely have some form of mathematics because honestly, you're going to start at physics.
Every object in space behaves identically everywhere - it's one of the fundamental assumptions we have and hope it true, which means basic orbital mechanics by any space faring species will be identical. Gravity, orbits, etc., will have identical concepts. You might not be able to use the same coordinate systems (Euclidean or polar might be human concepts), but mathematics has told us that it's basically just a transform to go into any other coordinate system.
Likewise fundamental constants are still fundamental constants. Fundamental geometric shapes are still fundamental geometric shapes - a sphere is the only shape with the largest volume and smallest surface area.
Mathematics may not be the universal language, but science is - physics is pretty constant throughout the known universe. And even if you have a language that is "one, two, many", obviously they have gone beyond that to handle physics = either through a new language, or through other things.
Every object in space behaves identically everywhere - it's one of the fundamental assumptions we have and hope it true, which means basic orbital mechanics by any space faring species will be identical. Gravity, orbits, etc., will have identical concepts. You might not be able to use the same coordinate systems (Euclidean or polar might be human concepts), but mathematics has told us that it's basically just a transform to go into any other coordinate system.
Likewise fundamental constants are still fundamental constants. Fundamental geometric shapes are still fundamental geometric shapes - a sphere is the only shape with the largest volume and smallest surface area.
Mathematics may not be the universal language, but science is - physics is pretty constant throughout the known universe. And even if you have a language that is "one, two, many", obviously they have gone beyond that to handle physics = either through a new language, or through other things.
Upvote
0
(3
/
-3)
Point of order: that's not even true within the context of the story of PHM, where Rocky's people have no concept of relativity and are unable to understand why their basic physical assumptions about the universe didn't quite apply in terms of the length and time of the journey to Tau Ceti.
Or, rather, a better way to respond is that even though space and time do indeed respond similarly everywhere once the basic rules are known, one alien's understanding of space and time cannot be taken as a given when establishing a bedrock framework on which to ensure the meaning of terms is being accurately communicated. Observable physics, like so many other properties of space-time, have heavy dependencies on the observer's frame of reference.
Upvote
4
(4
/
0)
chaos215bar2
Ars Praefectus
This is probably the least plausible aspect of the story, at least from what's discussed here. (Have not read the book, and will not see the movie until later this week.)
Even orbital mechanics within our own system don't quite work without relativity. Reaching a level of development sufficient to send a manned spacecraft to another system without understanding relativistic mechanics somewhere along the way just doesn't make a lot of sense to me.
Upvote
1
(2
/
-1)
chaos215bar2
Ars Praefectus
Why? We've discovered and have technology that makes use of all sorts of phenomenon we can't directly observe.
And while echolocation may have distinct limitations in air, keep in mind that is not the environment in which Rocky's species operates. Echolocation in a high pressure liquid environment could be nearly as good as sight is in ours.
Upvote
6
(7
/
-1)
chaos215bar2
Ars Praefectus
I think you're focusing far too much on the layout, while not paying enough attention to the data it contains. Rearrange the periodic table however you want — randomize it and change all the element names, even — and it will still be recognizable as a list of elements grouped by nuclear charge.
How you go from there to full communication might be a harder problem, but reaching a baseline shared understanding of what you're looking at should not be a problem.
Upvote
-1
(0
/
-1)
No, I'm responding to the poster claiming that (one of) the layout(s) we currently use is inevitable and immutable - which is clearly very, very wrong.
Upvote
4
(4
/
0)
Though if you have nearly unlimited energy and can course correct when you seem to be going off course since you didn't take relativity into account, maybe you can get away with it.
Upvote
1
(1
/
0)
Just like in The Martian, all the cool sciency bits skipped over very briefly so the general public doesn’t fall asleep. We need geeks editions of both that are at least 6 hours long.
Upvote
3
(3
/
0)
Sure. Books have the luxury of detail, and also of a semi-omniscient narrator to explain things to the reader. The Martian novel is about 400 pages long - that would take me a solid couple of days of reading to get through, many more hours than the ~2.5 hours the film has to work with.
Doesn't mean the film is "worse," just that it's different because it uses a totally different media.
---
Edit: I mean, OK, sometimes films are much, much worse than the books they're base on. For example, the execrable, misbegotten film trilogy based on The Hobbit.
See also David Lynch's Dune.
Last edited:
Upvote
5
(5
/
0)
Where would ant colonies fit into this? Or a creature like MorningLightMountain (from Pandora's Star), in which a single central intelligence guides multiple separate mobile units that have no intelligence of their own? Or the Formics of Ender's Game? Or even the pack-like Tines from A Fire Upon the Deep? Bears and tigers aren't the only alternatives to human tribes.
Upvote
1
(1
/
0)
I have read the book and last night I went to see the movie with one of my daughters. And I have to say that I was delighted by this movie like I haven't been by a movie in a very long time. I didn't have time to think about linguistic technicalities as I was enjoying the movie too much.
Upvote
6
(6
/
0)
This is something I've personally been working on for a VERY long time. Unfortunately, I've never had the type of academic help I would have liked, or I might have been able to actually finish writing everything up by now and had it published properly... (Once I've finished I'll be sending it to everyone I can, probably including Ars, too, so...)
Unfortunately, since I'm still in the process of writing it all up as best as I can, I can't really go into too much detail. The problem is that scholars and academia have completely failed to properly, fully and consistently conduct one of, (if not THE most) fundamental studies required for their own purpose and existence: of what language truly is - of how and why it exists and functions as a matter of human creation, ultimately within the universe as a whole. (Starting with language and working all the way down to the universe (that we then perceive) and then rebuilding its descriptions etc. all the way back up.) This context includes communication (and more), but they're not the same thing, and no, even communication isn't fully understood by academia either.
I know for an absolute fact, with asolute certainty, based on everything being taught and described including some of the history involved, that scholars and academia have not done this, along how and good idea of why, because it's exactly what I myself have done in their place - (becuase I really want to talk about games, but could not do so without such a context existing as it needs to...)
That this failure has symptoms EVERYWHERE, that are not fully recognised and understood, (assuming they're recognised at all) should be completely unsurprising. (E.g. This is also the reason for why mathematics doesn't have the context it needs, which is why it's NOT a language in itself, but a subject for existing languages, which is why not understanding this context merely leads one to have to create their own in their place (see: Principia Mathematica.))
Upvote
-9
(0
/
-9)
Respectfully—and I genuinely mean this as a constructive reply, not a snarky one—"every expert in a given field is wrong/misguided/missing a fundamental that I have found" is rarely correct, regardless of of the field. The odds are not in your favor
Upvote
14
(14
/
0)
Saw movie opening night, liked it (you will too), but I, too, was disturbed by the fast and loose treatment of communication. Hard to believe (re: "willing suspension of disbelief") and kept bothering me as it progressed into values, implications, references, etc. One of several Deus Ex Machina issues that disturbed me and drew me out of the film experience into intellectual and logical considerations. Perhaps better dealt with in the book.
Upvote
0
(0
/
0)
Sort of off topic but I wonder if “Erideans” was derived from The Lensmen SciFi series’ Aridians.
Upvote
-1
(0
/
-1)
It's very obviously derived from Eridanus, the constellation.
ETA for the doubters: specifically from a triple star system in that constellation, 40 Eridani. If you can't be bothered to read the book, at least read the non-AI summary in Wikipedia.
Last edited:
Upvote
6
(7
/
-1)
I loved the book, plan to see the movie. These deep questions about communication with aliens will get a lot clearer once we have a chance to actually talk with aliens. I’m looking forward to it!
Upvote
1
(1
/
0)
Spoilers as marked.
We are thinking about language and communication a little too narrowly.
Sticking with the film version, the first series of "messages" Grace and Rocky interchange include:
Guys, there is communication happening already. No language involved.
What does the object contain?
A symbol apparently representing Grace, and another of his ship. Rocky is already telling us "he" understands sign and signifier. And, when we learn what the sperm-shaped thing is, it immediately gives them a point of commonality.
How quickly they bond?
I don't know how much time in the movie their first few sessions took, but Grace's hair and clothing show time has passed. It may seem hand-wavy to point out Rocky's audio adeptness is a giant help, along with the best version of Google Translate Project Hail Mary could buy (running locally, to boot!), of course, but I can fit that within a reasonable suspension of disbelief.
It's also worth noting from the seemingly extraneous detail of watching over each other's sleep that
Rocky's species literally evolved depending on trusting others to have your back.
It may have been a good thing for everyone involved that every other human who might have been on that ship other than Grace was dead.
I may be running some scenes together on one viewing. In a nutshell: it was a wonderful movie full of heart even though it's primary two human characters struggled in that area. That we're talking about the science in it is to its credit, but if what we take away from it is the science, we're probably missing its point.
My son loved it.
We are thinking about language and communication a little too narrowly.
Sticking with the film version, the first series of "messages" Grace and Rocky interchange include:
- Grace running away; Rocky effortlessly matching course
- Repeat, with a corresponding number of engine impulses
- Grace assuming a lobbed object might be a threat and letting it bounce off
- Grace emerging from an airlock and engaging in antics to catch a second object, tossed slower
- Another object is tossed directly at said airlock.
Guys, there is communication happening already. No language involved.
What does the object contain?
A symbol apparently representing Grace, and another of his ship. Rocky is already telling us "he" understands sign and signifier. And, when we learn what the sperm-shaped thing is, it immediately gives them a point of commonality.
How quickly they bond?
I don't know how much time in the movie their first few sessions took, but Grace's hair and clothing show time has passed. It may seem hand-wavy to point out Rocky's audio adeptness is a giant help, along with the best version of Google Translate Project Hail Mary could buy (running locally, to boot!), of course, but I can fit that within a reasonable suspension of disbelief.
It's also worth noting from the seemingly extraneous detail of watching over each other's sleep that
Rocky's species literally evolved depending on trusting others to have your back.
It may have been a good thing for everyone involved that every other human who might have been on that ship other than Grace was dead.
I may be running some scenes together on one viewing. In a nutshell: it was a wonderful movie full of heart even though it's primary two human characters struggled in that area. That we're talking about the science in it is to its credit, but if what we take away from it is the science, we're probably missing its point.
My son loved it.
Upvote
5
(5
/
0)
Funny, I was just starting to re-watch a series called SeaQuest DSV recently. It's basically a space opera set underwater instead. Their (initial!) token "alien" is a dolphin, and they have a computer to translate for it. The computer mostly works, for basic things like names, "play", "fish", etc. But in the second episode the more abstract idea of "I'm dying alone" gets absolutely mangled into "light is dark and one", which the human characters spend half the episode trying to figure out.
Upvote
4
(4
/
0)
I got the feeling that since Rocky's entire thing is interpreting complex vocal harmonies, he's uniquely capable of learning and understanding new sounds. So he's probably understanding human speech faster than the opposite, and helping the whole thing move along.
While the movie skips a lot of the work they go through, I reckon that they probably spent a fair few days just learning to communicate.
While the movie skips a lot of the work they go through, I reckon that they probably spent a fair few days just learning to communicate.
Upvote
1
(1
/
0)
I don't know what "none of the marketing around the book" is meant to include, but the back of the hardcover dust jacket for PHM says:
"And with the clock ticking down and the nearest human being light-years away, he's got to do it all alone.
Or does he?"
Pretty obvious clue Grace is going to have company.
Upvote
2
(2
/
0)
Has anyone in the thread mentioned Peter Watts ? Blindsight?
Haven't seen Project Hail Mary but I read the book and - while I enjoyed it - I couldn't help but think that the aliens should have been more alien. Rocky was pretty close to a guy in a rubber suit. Trying something like "Alien with no theory of mind" probably would have made the movie a bit too weird... though getting the idea across might have been fun.
Haven't seen Project Hail Mary but I read the book and - while I enjoyed it - I couldn't help but think that the aliens should have been more alien. Rocky was pretty close to a guy in a rubber suit. Trying something like "Alien with no theory of mind" probably would have made the movie a bit too weird... though getting the idea across might have been fun.
Upvote
3
(3
/
0)
I know that, which is why I couldn't believe it myself for a very long time - but since I now know and understand so much more, I can see very well how and why all the problems still exist, and have done for (possibly) millennia. I ran most of it past experts on the way, and all they told me to do was write it up, which is proving to be easier said than done. (The last expert being Professor Neil Mercer of Cambridge University (UK) before he retired.)Respectfully—and I genuinely mean this as a constructive reply, not a snarky one—"every expert in a given field is wrong/misguided/missing a fundamental that I have found" is rarely correct, regardless of of the field. The odds are not in your favor
That there is an obviously known problem with academia understanding both communication and language - things that have existed (and therefore obviously known and understood by humanity) long before scholars and academia - should tell you that there's something fundamental going wrong at its (academia's) root.
Nothing I've told you should actually be surprising to anyone who recognises that such problems exist.
Once I've finished writing everything up and managed to get relevant people to read it, I expect there to be the biggest academic headdesk of all time
Upvote
-10
(0
/
-10)
Andy Weir wanted to make sure he worked in the phrase "mitochondria, the powerhouse of the cell" in his interview, eh?
It's a law of the internet, perhaps the English language itself, that upon first mention of mitochondria in a conversation, someone has to inform the rest of the audience that it is, in fact, The Powerhouse of the Cell.
It's a law of the internet, perhaps the English language itself, that upon first mention of mitochondria in a conversation, someone has to inform the rest of the audience that it is, in fact, The Powerhouse of the Cell.
Upvote
3
(3
/
0)
Someone may want to consider there are multiple parties involved in communication, and it often is not the people sending the messages who are the reason the information is not being grasped.Once I've finished writing everything up and managed to get relevant people to read it, I expect there to be the biggest academic headdesk of all time
We all look forward to seeing you blow our minds.
Upvote
2
(2
/
0)
I know that, which is why I couldn't believe it myself for a very long time - but since I now know and understand so much more, I can see very well how and why all the problems still exist, and have done for (possibly) millennia. I ran most of it past experts on the way, and all they told me to do was write it up, which is proving to be easier said than done. (The last expert being Professor Neil Mercer of Cambridge University (UK) before he retired.)
That there is an obviously known problem with academia understanding both communication and language - things that have existed (and therefore obviously known and understood by humanity) long before scholars and academia - should tell you that there's something fundamental going wrong at its (academia's) root.
Nothing I've told you should actually be surprising to anyone who recognises that such problems exist.
Once I've finished writing everything up and managed to get relevant people to read it, I expect there to be the biggest academic headdesk of all time
Ah, another square-jawed, lonely paragon Hero of Science, single-handedly inventing one of the most important concepts that's somehow conspicuously missing from a field of study that's several centuries old.
Next up, you'll find an obvious truth that will change all of mathematics forever and that somehow no-one noticed before. Just a hint, before you gird your loins for this next quest: Fermat was a horrible practical joker and the fact that he turned out to be correct is not fundamentally different from the flip of a coin. Or a custard pie.
Upvote
0
(1
/
-1)
Upvote
4
(4
/
0)
To your earlier point about Rocky being the one doing the linguistic heavy-lifting, I think it's very plausible. With his eidetic memory, he should be able to retain a list of all the words he's taught Grace, and generally stick to them. Grace comments more than once, I think, about Rocky seeming to think he (Grace) is stupid, and/or treating him like a child. It stands to reason that Rocky would figure out how to confine his communication to combinations of the words he's already taught Grace wherever possible.
I agree the subject of reproduction and mates could have come up without them discussing whether either of them had a mate - though it also seems a little odd that you'd discuss how your species reproduce and not ask any follow up questions like "do you have a mate/children?"
However, there's a more egregious example in the film (it's been a while since I read the book, so I don't remember if it's present there as well): there's a scene where Grace tells Rocky he's going to sleep, which provokes a discussion where Grace has to explain what sleep is, because Eridians don't sleep (they have a hibernation behaviour which is somewhat analogous, but it's not clear if they do it regularly, or only under certain conditions). At the conclusion of that conversation, Rocky says something which is translated as "I will be here when you wake up". That really bugged me, because how could they have a shared vocabulary for "wake up" without having discussed sleep? If it had translated as "when you finish sleep" or "end sleep", that would have been fine, but "wake up" is a far more specific phrase.
The "our word for your name" thing also bugged me, because it happens early on, seemingly before they could have talked for long enough for him to explain what "Grace" even means, and there are so many possibilities - not just the religious one, or the one about movement or form, but also "grace under pressure" and even "a statement of gratitude made before a meal", and doubtless others I haven't even thought of.
I'd say there's a decent chance that calling an Eridan "Rocky" might be offensive, considering, in the reverse case, an Eridian might call a human "Fleshy". But, coming from a non-flesh based creature, I think most people would understand the logic and accept that appellation with a certain degree of... grace.
Last edited:
Upvote
1
(1
/
0)
Its obvious thematic connotations aside, of course. Like Tom Cruise's character in Risky Business being named Joel Goodson. Good son. Get it?
Ryland Grace. Wry grace...
Upvote
0
(0
/
0)
"...she threw him the 'reverse-peace' sign (I don't know what they call it)..."
It's often called, "flicking the V".
Apocryphally, it relates to a claim of superior skill with the longbow, (Of the English, over the French).
Upvote
1
(1
/
0)
While I get where you come from and to a large part agree with you, there are stories (or rather plot twists) that I feel shouldn't be known to the reader before reading the book. Two that come to mind are I am Legend and Ender's Game.
Upvote
0
(0
/
0)
yeah, wonderful (NOT). That paper is one of the bedrocks depended upon by people who deny that true AI could ever possibly be intelligent. Which puts them at odds with pretty much every computer scientist and physicist (Roger Penrose excluded. Nobelitis is a bitch).
The brain is a physical mechanism. Qualia and conscious events are patterns of physical activity in that mechanism. Nagel's argument is basically "I can't imagine it therefore it can't exist.". It's just a failure of imagination and a failure to understand the law of large numbers (see Searle and the Chinese room thought experiment, another big failure of imagination).
Fortunately we don't have to listen to these morons for a heck of a lot longer. Since this is fast approaching not being an academic argument. When we have built AGI that claims it is conscious and which we can analyze to see the patterns which represent consciousness and qualia, they'll have to shut up.
We can already see the rudiments of this in the working of current AI technology, which uses emergent "meaning" vectors to automatically develop an abstract concept of meaning as a vector in a very high-dimensional emergent space. We have a really hard time looking at those vectors from outside and perceiving them as "meaning", but they clearly are. Just hard for us to see from outside the system, partly because of the law of large numbers (current AI capabilities have "magically" come from mostly from scaling, and we have a very hard time thinking about many-dimensional vector spaces).
Upvote
-7
(1
/
-8)
You're even more dogmatic than the "idiots" you mock. Your argument is "because I can imagine it, it will be a thing."
What we are fast approaching is the law of diminishing returns. Perhaps you'll find a magic bean to overcome them.
Upvote
4
(5
/
-1)
How am I dogmatic? There is zero evidence that our minds are anything but physical systems, governed by the dynamics of physics. We are physical systems, that are conscious. Therefore physical systems can be conscious, and perceive, whether they are biological or not; biological has nothing to do with it.
The burden is on those who think there is some necessary additional "magical/nonphysical" ingredient, to prove that. And they never have, all the flailing notwithstanding. Penrose's ideas are based on a misunderstanding of Godel's Theorems. Which are outside of his field. Thus Nobelitis.
The dogma is rather on the side of those who say "computer's can't be conscious or experience qualia, because <obviously shitty reasons>". THEY need to prove that there is any process involved in consciousness that isn't physically simulable. They've tried. And failed, so far.
The diminishing returns in LLM scaling aren't any kind of evidence that AI isn't possible, or rapidly approaching. Almost all technologies go through alternating exponential growth phases, interspersed with plateaus as the exponential growth regime ends, followed by new exponential growth phases as additional breakthroughs are made. So plateaus prove nothing.
There are obvious temporary limitations to LLM architectures by themselves, but they are clearly part of the solution. The next exponential phase will likely involve a more sophisticated architecture that adds a permanent learning process that updates a persistent memory, so that actual temporal memory and persistent "self state" and the perception of self-thought can emerge. That isn't dogma, it is the consensus opinion of the scientists who actually know what they are talking about, not of bad philosophers and rando non-computer-scientists, who have proven for hundreds of years that they are not competent to talk about this.
At any rate, you can wait ignorantly sniggering on the sidelines, waiting for the "magic beans" to fail to appear. When they appear, what will you say?
Last edited:
Upvote
-4
(1
/
-5)
Okay. Show me. Make it so.
I'm not positing a soul or some mysticism. I am eminently pragmatic. You say "it is, therefore it is." But the technology is missing something that at this point can be dismissed only through semantics. The output mouths claims of self-awareness while demonstrating its utter absence.
Find what is missing. Find your magic bean.
Upvote
3
(4
/
-1)
Your fallacy is in the induction from "brains are physical systems" + "LLMs appear to be able to emulate a small part of what brains appear to do" => "conscious AGI is just around the corner". We mostly can simulate physical systems that we understand, but there are some caveats here: first, we don't really know bow consciousness emerges from what the brain does; second, we don't know if it's only the brain or the whole body that needs to be considered; and third, we.do know that LLMs are not doing anything even remotely like what happens in the brain.
We don't even have a good model of what consciousness is, nor is it obvious that consciousness is a prerequisite for intelligence in general or for AGI specifically.
Upvote
5
(5
/
0)
Observation: When someone edits a post, usually it is to tame down insults. To edit a post to add phrases like "ignorantly sniggering" while claiming that "for hundreds of years" "rando" people have been doubting computers are a path to... well, to be honest, I don't want to know what someone thinks people have thought computers couldn't do "for hundreds of years," is really just pounding the table.
He says "It can be." I say "Well, it isn't." He says "More of the same will get us there."
Okay, NOW I snigger. Magic bean time, buddy!
He says "It can be." I say "Well, it isn't." He says "More of the same will get us there."
Okay, NOW I snigger. Magic bean time, buddy!
Upvote
1
(1
/
0)
A failure to understand how and why passive communication even exists, (and is part of the reason for why language was created) is part of the symptoms.
Upvote
0
(0
/
0)