Comment by mettamage | Hacker News Reader

Comment by mettamage | Hacker News Reader

We would be able to challenge human exceptionalism way more effectively if we could fully decode the languages of other species. The first thing we'd notice is:

1. language features we have and they don't understand

2. language features we both have

3. language features they have and we don't understand

Probably in that order.

Then it's just a question of gathering a couple of different species that are seemingly intelligent. Such as: corvids, octopuses, whales, etc. And see if the species can be reasoned with. If so, then you can set up schools where you can train them on human things and vice versa. Eventually you can form interspecies groups and really test the hell out of things.

Doing it that way will really challenge human exceptionalism, as well as the exceptionalism of that particular species.

I know it sounds a bit far off, but I figured that we might be able to get there with AI. I mean, we're getting better and better at giving machines tons and tons of data, and it somehow makes some sense of it.

So far, I think it's not necessarily the human species that is exceptional. It's the revolutionary periods it went through in order to become more exceptional hunters, so we could dominate and control the world in the way we want to. Things such as: discovery of fire, agriculture (+ creating defensive settlements) and antibiotics. We couldn't kill bacteria for a long time. We still have trouble with viruses and are getting into trouble with bacteria again. Could dolphins or whales have done it too, if they were land creatures?

It doesn't help that the main conclusion of over 100 years of this kind of animal language research, is that -- and I quote -- "animals don't have language" [0]

[0] https://en.m.wikipedia.org/wiki/Great_ape_language

The most charitable viewpoint I can give is: that can be true or that can be false. Time will tell.

If it's true, then we are exceptional but how can we truly know at this point? I mean, Michael the Gorilla told us about what poachers did [1].

If it isn't true, then it shows how in the prevailing consensus we are still too arrogant. In that case, we don't understand that well what sets us apart and what doesn't.

It will probably be a nuanced discussion either way regardless of what the truth is probably due to the definition of language. But in this particular case I'd want to characterize it: some way of communicating that is about as effective as whatever it is that humans do, when they speak out loud.

Also, when you look nowadays at some of the dog/cat videos and they press those buttons, they clearly are capable of communicating something. I remember one dog inventing a word for ambulance given the words he/she knew. Ah, found it! [2]

There's clearly a lot more research to be done here. I hope AI can accelerate it. I know AI is a tricky business, but one can hope.

[1] https://www.youtube.com/watch?v=DXKsPqQ0Ycc&ab_channel=kokof...

[2] https://www.youtube.com/shorts/a93PWudceKk

I wonder if part of the nuance here is understanding what language and communication are, respectively. I am definitely not in my own lame here, but at first glance of Great Ape Language Wikipedia article and some comments, it seems like that it is not addressed? My admittedly infantile assessment is this; all language is communication, but not all communication is language.

For example, I can issue a warning to another human by using the words "step back or there will be a problem." Assuming they also speak English, my structured sentence conveys the warning in a clear way.

My neighbor's dog can also issue a similar warning by growling as I move into it's space. It's not using any sort of structure or specific messaging, rather a universally understood sound that conveys some sort of feeling related to caution and fear.

I use the "warning" example since fear/caution are extremely powerful and seem easiest to convey across species. Again, not adept at this, so my example might be dumb. I'm super interested in understanding more about this, though.

Any suggestions of further reading?

Agreeing with your distinction. Also, you could "issue the same warning" by growling like the dog, too - or by waving your hands frantically or shaking your head, etc. This would absolutely convey the same information and count as an act of communication, but we wouldn't describe it as using language.

But the information in that example is also relatively simple. There are other kinds of information that are much harder to convey without language, talking about future or hypothetical events or abstract concepts, etc.

There seems to be evidence that dolphins use sounds to identify each other and to coordinate tasks that involve a number if animals. I think this is much more interesting and strongly hints that something language-like is going on there.

https://www.ecowatch.com/dolphins-language-communication-evi...

https://pmc.ncbi.nlm.nih.gov/articles/PMC8074934/

Interesting reads, thank you. You're right in that the complexity of the information matters when it comes to how it can be communicated, although I also think of the bit, as in the smallest unit of information, and how we have accomplished communicating some wildly complex information by cramming a billion transistors into microchips. Still blows my mind while it seems like it's become so commonplace for most people.

When I first learned about the dolphin "names" whistle, I thought of the way dogs supposedly identify each other by scent in the way that there's subtle differences that our olfactory system is not able to pick up on, but theirs is. In both cases, it begs me to ask what that means for consciousness. If the dolphins and dogs can assign individual labels to each other and recognize those labels when used later on, does that imply they form some sort of a personal identity? I also have to wonder how language, as opposed to other forms of communication, promotes the birth of that identity, but so much of this is over my head in a way that leaves me grasping at how to, well, communicate my own questions so that I receive answers that promote better understanding of the phenomenon and its challenges.

So, one of my favourites is this one: https://en.wikipedia.org/wiki/A_Man_Without_Words

Not a good book by any means, but there has been surprisingly little written about people who go from having no language to their first language post-puberty and are asked about their experience of living without language.

Language is extremely important to living as a human, even outside of social relations. Memories don't form well, it is really hard to keep events in order of time, it is hard to order places in space and not get lost. It's even hard to remember people.

It's an engineer take, of course, but I like understanding systems by what happens when they break.

> I like understanding systems by what happens when they break.

Same, though I am not an engineer, more a denizen of the adjacent service and repair field, but I have been dabbling in development, lately. Fixing broken things because I grew up in a household that was too poor to think of anything as disposable is how I learned probably 90% of what I know, providing me with an understanding of the fragility of systems and the importance of anticipating maintenance the design process.

I see the forward was written by Oliver Sacks, whom I've read and enjoyed in the past, but from more of a pop-Psychology angle. But what you say about language being related to how we think of time and forming memories strikes me. I once heard that Indigenous Australian cultures had a different concept of time than Western culture does, passed down through the language used in their folklore. I posit that we can some evidence for proof of this happening in Western cultures currently with the way that we talk about decades.

With the advent of the year 2000, we lost the ability colloquially refer to the decade we were presently in. Prior to that, we were living in "the 90's," perhaps being born in "the 60's, 70's or 80's." Each of those decades even evokes a certain composite memory of the styles of clothing people may have worn, or the music we listened to, popular automobiles, etc. The character of the decade coalesced around the colloquial name.

2000 hit, and we didn't know how to refer to it. Some folks tried to make the "aughts" happen, which felt awkward or "the 2000's" which seemed to generalized and not specific enough to be referring to the current decade. Then 2010's, which we didn't really even bother calling "the Teens" as I recall, because by then we'd fallen out of the habit of referring to time by decade.

The result is seeing a movie from 2005 and feeling like that was so very long ago, versus when I watched Star Wars for the billionth time in 1997 and the 20 years that had passed since it's release didn't seem like much time at all; only two decades.

Anyway, getting a ramble on to show agreement, but you get the idea. I might have to look for some science fiction authors who have tried to tackle this, see what sort of thought-experiments we can do when we strip language as we know it from the development of the human mind.

My view is as infantile as yours. We're all armchair <insert_topic> here with these types of topics.

> all language is communication, but not all communication is language

Yea, good point!

> Any suggestions of further reading?

Other than this and people now trying to study animal communication through AI (having seen a few videos on it), I don't know much more than that, haha.

Is language still unique to humans given the rise of LLMs lol!

This is the conclusion of linguists. Linguists’ definition of language necessarily includes a productive grammar. It’s not enough to be capable of uttering from a long list of sounds or signs. To have language we must be able to regularly construct novel and unique communications composed of strings of symbols governed by grammatical rules.

All animals that we’ve studied — including great apes we’ve attempted to teach sign language — have so far failed to demonstrate the acquisition of grammar. This means the number of unique communications they can express is exactly limited to the number of utterances they know.

Ok, but why do you expect us humans to be the ones to do all of this study if you don't already recognize that we are exceptional?

Why study anything? So far I've found the broad strokes to be, because of:

1. Satisfying curiosity / because it's fun

2. To build new technology

By virtue of that, my stance on whether I think we are or aren't exceptional is irrelevant.

I don't have any hope that any of this will be researched thoroughly enough in my lifetime though.

Biological autoencoders "go brrrr" and use every side-channel available.

Reply on news.ycombinator.com