Languages require some set of components or rules commonly recognized by speakers that can be used to convey information. Languages can be minimal, but that’s not really what we’re interested in. We don’t see dogs communicating in a complex language. My dog can make a several sounds and gestures to convey information, but the meaning of those expressions wouldn’t be recognized consistently by other dogs or people. It’s not what I would call a language. Some primates show some ability to communicate with language, but in a very minimal way, akin to a child’s use of language, but that is not the way they communicate with each other or any other animals.
I’m sure whales can communicate at some level with some very basic expressions.
We’re trying to find out what level of complexity exists in any language whales might use. Can it be used to express emotions using commonly understood expressions? I can express emotion without the use of words or language, but not the reason why I feel that way, or anything much more specific than ‘angry’ or ‘happy’. Do they have grammar, semantics, or even granular elements that we could categorize as words?
My ignorance no doubt shows and guess that when I hear semantics I misunderstand to be restricted to lexical semantics, given that you’ve excluded music from having semantics?
Music clearly communicates meanings. Often it has something to say without expressing it in words. True the domain it is most versatile in is the communication regarding emotional states (and may be more useful there than are words) …
Some of them can. A playbow is a playbow, and any dog who’s socialized to other dogs is going to recognize that; so is any human who’s been around dogs much and also paid attention to how they communicate (the first category of human is unfortunately a lot larger than the second). A growl with raised hackles is a warning. And so on.
I think what we’re doing may be trying to come up with a precise definition and distinction for something that’s actually a gradual continuum, with various aspects of the continuum being exhibited by various species. I often see people trying to do the same sort of thing to produce a clear distinction between humans and all the other species – we’re the only tool users! (no, we aren’t.) we’re the only tool makers! (no, we aren’t.) we’re the only complicated tool makers! (no, we aren’t.) – yes, there’s a distinction to be made, but it’s not a bright line one; it’s essentially blurry in its nature.
That’s my take too. You can define language based on the cognitive abilities of humans and it will naturally exclude dogs and bees, but I’m not convinced that our intelligence is fundamentally unique – just a more powerful implementation. I think this is precisely why the study of linguistics encompasses cognition.
I assume the definition of language is one of the first topics covered in Linguistics 101 and then endlessly redefined and debated through PhD and beyond. I found a short blog post from a Linguist that breaks down one common definition of language. It touches on the difference between language and communication as well as whether animal communication can be considered language.
Using the blog’s definition a language must have the ability to create and communicate a brand new idea; it is not enough to select from a very large set of pre-defined ideas. ASL is a language; semaphore signaling is not.
I understand why they assert that bird songs and bee dances are not language, but thought the jury was still out on whale songs.
Agreed and agreed. I’m don’t see convincing evidence that whale songs use a formal language, but then again we know so little about the whales and their songs. Admittedly, I can’t make out the lyrics in most songs at all, yet I can often understand the message they convey. I do recall Victor Borge explaining how certain parts of music convey feelings and actions in certain ways that do border on the concept of ‘words’. They aren’t formalized in any way I know of, though keep in mind I know very little about music, and understand even less about it. But conceivably a dictionary of musical sequences with commonly known interpretations is possible. I doubt whales could play a piano as well as Borge, but it’s possible they can be just as funny as he is in their own way.
I’m not sure that “Human, please open this door” is a brand new idea, exactly. But it certainly isn’t innate; especially in cats.
And each individual cat learns it anew – or do they? Each individual cat that’s taken at six weeks to live with no other cats, yes. But the ones who live in clowders may learn it from each other; I’m not sure how we’d tell.
And the cat who knows “Human, open this door!” can also say “Human, open this [fill in a different closed thing]” and will probably do so in an equivalent fashion: make eye contact with human, go to closed thing, aim body and attention at closed thing, meow, look at human, meow, look at closed thing, paw at closed thing –
That fails the blog author’s symbolic test, of course: no abstract sound or sign is substitued for the Closed Thing; the Closed Thing itself has to be referenced.*
Again: a continuum. Not something with a sharp line of distinction; which I believe you’re agreeing with. I do think the blog you cite has a useful discussion of the matter; but I think they’re trying to draw a super sharp line of distinction on creativity when even that one is blurry.
Bird songs may be one sort of thing; and some of what some of the corvids do may be something else.
*(Though I think it’s pretty clear that dogs, and some other species, understand that some human words reference something specific that’s nothing like the words. And I knew a dog as a child who said “water.” She couldn’t pronounce the consonants, but she had the intonation, the syllables, and the vowels. And she only said it when she wanted water; not when she wanted food or out or attention. She wasn’t taught this deliberately.)
That’s a good example. Does a language need to be understandable without context? It is clear what the cat means when you see him in front of the closed thing, but the entire meaning is not encoded in the cat’s forms. I feel like there would be similar examples in human language, but they would be more subtle. Perhaps when humans discuss emotions – fear, jealousy, love.
Totally agree, possibly intentionally over-simplified. It is a good demonstration of some of the major decision points though.
Which is what I was trying to get at with stating complex ideas being expressed not stereotyped forms. But that says it better!
How can we however tell that novel ideas are being expressed in creative manners when we don’t understand, perhaps never can understand, what the ideas are? Especially if we want to accept that the forms can be very different than the forms that gird human languages?
What do we want to observe to conclude that lacking an ability to translate or to even know if translation is possible in principle?
Yes, like intelligence and consciousness there are likely different grades and types, not a binary.
I can see your point. There would probably be concepts that exist in an alien race that do not exist in humans (and vice versa). So the aliens’ languages would have words for these concepts that human languages lack. And that means those words would be untranslatable.
But if we’re meeting these aliens we must exist in the same universe and therefore there must be some universal constants that both our races experience and discuss in our respective languages. This would produce enough of an overlap to demonstrate both of our races are using language and would hopefully be enough for us to be able to bridge the gap and be able to communicate.
I mean even with whales, we live in the same universe but have mostly evolved our intelligences, and potentially languages, in very different domains. It’s potentially not just different sets of objects and actions creating non overlapping nouns and verbs, but completely different modes of thinking about the world.
Less extreme but somewhat the concept behind the Chiang short story “The Story of Your Life”, made into the movie “Arrival” …
They may have different ways of thinking about the world but it’s the same world we’re both thinking about. And I feel that gives us enough common ground that we could bridge the gap between our two different languages - assuming there is a language on both sides of the gap. Which Louise managed to do.
Something that occurred to me the other day… Most studies of language in birds involves teaching the birds human languages. This is potentially feasible, because many birds have vocal apparatus capable of reproducing human sounds… but it’s not the language (if any) that they would use on their own. And bird brains are enough different than mammalian brains that they’re almost certainly not using the same brain structures, or ones evolutionarily connected to them, that we are. If some linguists are right that there are some aspects of language that are truly instinctive, birds wouldn’t have those same instinctive language features. Which means both that it’d be harder for birds to learn human languages than to use their own, and that we’d learn a lot more by studying the communications used by wild birds than those of birds who are taught human language.
We’d learn different things, I think. “How are the birds communicating in the wild?” is one question; “Can any birds learn to understand and use human language?” is a different question. And then “What are the differences and similarities between bird communication and human communication?” is yet a third question; though it depends on at least partial answers to the first question and ideally to the second.
I think we should be studying all three. But I agree with you that studying only the second would be leaving a whole lot out.
I’ve read variants of this critique over the years in popular scientific articles, most notably regarding the teaching of language to gorillas. Skeptics of whether gorillas can be taught a true language tend to advocate more in-the-wild-studies.
But we’ve done a lot of that. In fact there are 2 sizable scientific literatures, one devoted to teaching animals human language and another devoted to studying animal communication in the wild. Oddly, “Research programs on animal communication systems in nature have proceeded essentially independently of research programs endeavoring to teach language to animals. This is surprising in light of the early, well-known efforts to relate these two research streams, especially by Hockett (1960) and Marler (1961),” according to Beecher (2021), in an interesting paper entitled Why Are No Animal Communication Systems Simple Languages?
Beecher tentatively assumes that some animals may indeed have the cognitive and physical ability to use language, but that even simple language (which he defines) is unknown in the wild outside of homo sapiens. How could this be?
Emphasis added:
Very unusual circumstances are required for a true language system to evolve. Three essential conditions have to be met. First, the species must have the underlying cognitive capacity. Honeybees may lack this, but some other animals may have it. Second, and this is the clue provided by honeybees, sender and receiver must have identical or near identical interests. Third, individuals must have a compelling need to transmit information across multiple contexts. These are precisely the conditions that existed in pre-human and early human hunter-gatherer societies, the context in which humans and our hominid precursors spent some 95% of our evolutionary history…
Humans are the supreme cooperators in the animal world, but because this cooperation is not supported by high kin relatedness, it has to withstand a strong undercurrent of individual competition. We sometimes lose sight of the human affinity for within group cooperation because of its paradoxical coexistence with intense between-group competition and tribalism.
So according to Beecher for animal communication to evolve you need not only brains and an adequate physical communication system, you also need a society where sender and receiver are in a cooperative relationship. Animals are very good at communicating, “Go away or I’ll hurt you”, for example, not so good at explaining the Pythagorean theorem.
Beecher’s hypothesis requires some elaboration in my view. Sperm whales probably meet both criteria one and two. They get partial credit at least for three, insofar as they spend a lot of time socializing with each other (as do Bonobos). So why don’t either have complex language?
There can only be one
It is said that Homo Sapiens are the only species to have evolved complex language. But we don’t really know that. It is conceivable that Neanderthals and other primates such as Homo floresiensis may have evolved language thereby becoming a strong competitor to Homo Sapiens. If that is the case, we’d expect such competitors to be dispatched with extreme prejudice: language is a powerful enabler of effective warfare and therefore genocide. If Orcas had sufficient ability to coordinate and significantly interfere with human fisherman, they would have been wiped out. Sort of like wolves.
This perspective of mine may lack imagination. To meet Beecher’s 3rd criteria a sexual selection arms race of sorts may be sufficient. And perhaps a complex language could evolve among water dwellers without that specie wholly dominating their domain.
Yet another prospective machine learning (ML) article, this time in the preeminent journal Science, published on July 13, 2023: Using machine learning to decode animal communication: New methods promise transformative insights and conservation benefits
It’s not that great to be honest, but its high profile is noteworthy. This paragraph gives a sense of the breadth of the endeavor:
There is a growing number of studies that are exploiting the potential of ML for investigating animal communication, including large collaborative initiatives, such as the Earth Species Project (ESP); Communication and Coordination Across Scales (CCAS); Vocal Interactivity in-and-between Humans, Animals and Robots (VIHAR); Interspecies Internet; and Project CETI (Cetacean Translation Initiative), which recently provided a detailed roadmap for ML-assisted work on sperm whale (Physeter macrocephalus) communication (2). Although efforts to tackle this grand research challenge are clearly intensifying, the field faces at least two main data-related obstacles: Most methods require vast amounts of data (4), and recordings of a single modality (e.g., vocalizations) are insufficient for functional decoding; additional context is required, including information on the animals’ behavior and environment.
It makes sense for an intelligent species, like whales, for selection of language to emerge if it results in more whales reaching reproductive age. “Hey kids, avoid those big trawling nets you see behind those monkey-driven boats, or else you’ll soon find yourself on the menu in one of their sushi bars.”
Or, it could be selected for if it results in more baby whales being born—the whale equivalent pick-up line, “so, how You doin?”
But, for whatever reason language emerges in a species, having it diverge from there to other topics of interest is inevitable. I suspect that’s what happened with hominids. Example: Step 1 [Oldowan]: “Listen up and watch me guys: this is how to turn a stone into a killing tool, then use it to skin those you kill into food.” Step 2 (a couple of million years later): “Who wants a Big Mac from MacDonalds?”
IOW, if a species is cognitively advanced enough to develop and use language as a tool to give it a reproductive advantage, it has a good chance of occurring. But, then again, gene mutation in that direction may not occur, in which case it won’t happen. Biological evolution is pretty good, but also messy, and has resulted in many failed attempts over eons of time on our planet. It had no good conductor, like Herbert von Karajan.
AI evolution, on the other hand, has the potential to put biological evolution to shame. Once AI becomes conscious and self-aware (it’s inevitable, IMHO), it will direct its own evolution. AI will first laugh at Asimov’s Three Laws of Robotics, then biologicals (like us) will become obsolete.
Natural selection is willy-nilly, with no direction, but artificial selection will have direction. Sucks to be us.
You can play the same notes in a different key, with a different rhythm or pace, and still recognize the song or the melody as the same. Many songs have many different versions by different artists. That is not translation for me, I would call it re-interpretation. Your linked article seems to say something similar when they write “turn Beethoven into blues”, not “translate Beethoven into blues”. But I was talking about meaning: is that kept? What is the meaning of a certain song? Does that remain invariant when you “translate” (or, as I claimed: re-interpret) from Betthoven into blues, or pop into blues?
I am afraid we have not agreed what the word meaning means. Is a feeling a meaning? Can you transmit that reliably with music? Would it be clear enough?
A typical song could mean: I fell in love with this woman, but she rejected me and now I am sad. Could you transmit that with notes so that it is clearly understood? I am afraid the message is too complex and nuanced for notes alone, it needs words. You can say “this is sad” in general with notes, melody and rhytm, or “this is happy” or “let’s dance!”. But not a complex sentence like the one I used as an example. I remember the piano suite “Pictures at an exhibition”, by Mussorgsky:
It is a musical depiction of a tour of an exhibition of works by architect and painter Viktor Hartmann put on at the Imperial Academy of Arts in Saint Petersburg, following his sudden death in the previous year. Each movement of the suite is based on an individual work, some of which are lost.
I used to try to picture the paintings in my head. Of course I had no chance. It may be my shortcoming and not the music’s, but I think it is more likely it is the lacking semantic depth of music that is at fault here.
I believe that is not how evolution works, I think that would be finalism or Lamarkism. Just because something makes or would make sense it will not necessarily happen. It would make sense to have wheels, or propellers, but that has not happened. Nature needs an evolutionary path, then, when passing through that path by chance through random mutations things emerge, like language and wings, and you see whether the mutations survive.
I ignore whether evolution is pretty good or not, it depends on what it is considered to be good for, I guess. But not only did it not have a good conductor, it had no conductor at all. That was my point about finalism too.
