Apart from humans, which animal do linguists currently think has a communication system that is closest to being considered real language?

"Language" meaning whatever linguists currently understand it to mean, as far as I'm aware that's a system of communication (via any means) that has at least infinite productivity and displacement. Please correct me if I'm wrong about this too.

  • 5
    New research claims that some whales use communication that is “compositional”: x.com/begusgasper/status/1732069051512598797 Dec 20, 2023 at 20:43
  • I've always enjoyed reading about squid communication. I can't find what I was reading, previously, but this is an interesting place to start: bigthink.com/life/cannibal-squid-speak
    – RIanGillis
    Dec 21, 2023 at 2:12
  • That is a very interesting question. I know that cross-posting is normally frowned-upon on stackexchange, but in this particular case I would be extremely interested in seeing this question asked on biology.stackexchanges.com or bioacoustics.stackexchange.com . The answers you got here are extremely human-centric. I could caricature and say they all amount to "Animals don't use the exact same grammar constructs as humans, therefore animals don't have language". I wonder how the viewpoint and the answers would change if you asked on a more animal-oriented stackexchange.
    – Stef
    Dec 21, 2023 at 10:33
  • 4
    @Stef specialists in animal behaviour often underestimate the additional structures present in human languages that do not seem to be present in (almost) all systems of animal communication (hence such bad science as things like Koko the Gorilla). There is good research being done on the linguistics of animal communication, but general animal behaviourists are no more able to give unbiased answers than general human linguists
    – Tristan
    Dec 21, 2023 at 11:25
  • 1
    @Stef "specialists in human language are prone to underestimate the complexity of animal communication" Are we? It would be absolutely incredible to identify any human-like structures in animal communication. Dec 21, 2023 at 22:28

6 Answers 6


This is a little like asking which species of elephant is closest to having the ability to fly. Animal communication is a long long way from human speech.

The closest are probably the "usual suspects": our close kin, chimps and bonobo and other apes. And dolphins or other aquatic mammals.

Chimps can produce a variety of calls which have an arbitrary association with the function of the communication. They can warn each other of danger and of the nature of the danger. That is they can say "Snake!" (with the implied "Get up a tree quick!") But they can't use that skill to talk about snakes in any way.

Dolphins undoubtedly communicate, but most of the essential features of language are absent. A lot of dolphin calls are social, and insofar as they have meaning, the meaning is "I'm here".

No animal communication system has anything like "grammar", They don't have a (large) set of "morphemes" that are combined to form words and sentences. Experiments with apes show that even in artificial learning environments, they don't learn to talk, the can learn to know and use a range of calls (or signs) but they don't put them together in a way that is instinctive for a human.

  • 2
    Wasn't there a gorilla (or orangutan), in the 1970s, who was "adopted" into a home in America and learnt sign-language and who could communicate really rather extensively..? Although, I must admit, I am not sure if it was what you would call putting the signs "together in a way that is instinctive for a human". I remember there being a TV documentary about it, years ago, possibly "Horizon". I think it was Koko Dec 20, 2023 at 23:14
  • 11
    The claims of "language" have been discussed at length. The claims of the researchers seem to have been widely criticised as cherry picking and projection of human language onto the gorilla's signs. It could do what a dog does, but better. Dogs can be trained to respond to commands or to bark for food, but they don't have "language".
    – James K
    Dec 21, 2023 at 0:20
  • 4
    Bees I haven't included at all. The reason is that bee language is entirely fixed. It is completely instinctive and hard-wired in the bees' brains. It is a remarkable adaptation, but it isn't remotely like human language. Bees are not a couple of evolutionary steps from evolving human language (chimps are, dolphins might be)
    – James K
    Dec 21, 2023 at 13:59
  • 3
    I don't think it is a grammar. I'm not sure there is a word for what it is, but not gramamr.
    – James K
    Dec 21, 2023 at 14:22
  • 2
    @JamesK doing some further googling it looks like I may have misremembered the stuff I'd read, or be remembering overly credulous pop-sci coverage. This seems like a good overview of what the waggle-dance is and why it cannot be called language: prizedwriting.ucdavis.edu/…
    – Tristan
    Dec 21, 2023 at 14:38

Human language is doubly articulated. That means that for every word (written or sounds), a regrouping of existing sounds (phonemes and morphemes) is used to produce meaning.

Double articulation refers to the twofold structure of the stream of speech, which can be primarily divided into meaningful signs (like words or morphemes), and then secondarily into distinctive elements (like sounds or phonemes). double articulation_Wikipedia

Thus, phonemes have no inherent meaning per se. They are combined with other phonemes (and morphemes) to produce meaning.

Animals cannot do this. They may produce sounds that have meaning but they do not doubly articulate language. They cannot recombine two sounds to mean other things. But many animals produce sounds that have meaning. My cat has about 5 different miaous. Each has meaning. I'm hungry. I want to play. I'm scared. I'm bored. That hurts! For example. And gesture: scratching the wall. seems to also mean: play with me.

The concept comes from André Martinet in his Eléments de linguistique générale, (Colin, Paris, 1960).

And is described in some detail here: double articulation of language

  • 3
    I'm not a linguist, but have to say this double articulation thing seems to be awfully oversimplifying things in a rather arbitrary way. It could equally well be argued that human language is triply or quadruply articulated. After all, a sound like, say, /æ/ just an abstraction over particular characteristics of sound waves, with certain spectral properties and a certain way of generating them with a mouth (which applies just as much to chimpanzee calls). And on the other side, many words are also quite meaningless, and by no means the highest level of structure found in human speech. Dec 22, 2023 at 7:47
  • @leftaroundabout Sorry, but you are mistaken. The point is that every phoneme (which has no inherent meaning) can be combined with others or another one to produce meaning. There is no articulation beyond that at that level. Many words are not meaningless. If something is actually a word, it has meaning and is made up of phonemes. And it is not about the highest level of structure. It's about the lowest meaningful structure and the recombination of phonemes.
    – Lambie
    Dec 22, 2023 at 15:21
  • 1
    Many words are not meaningless, but some are. Some phonemes have meaning, such as "aaah" or "uff". Vice versa, animals have things that could be described as words with meaning distinct from the individual phonemes, e.g. a single dog bark would likely signify simply surprise whereas a repetition of the same sound would be defensive aggression. Even "cat scratching wall" could be taken apart as "cat looking for a wall and walking towards it, then lifting paw and lowering it while in contact with the wall". There's no clear distinction you can draw between human and animal communication here. Dec 22, 2023 at 15:44
  • @leftaroundabout You said you are no linguist. Of course, onomatopeias (aahh or uff) created by humans have meaning. They can create new ones. Sure. But that is not the first thing that linguists study, is it? In the history of linguistics, it was probably much later. Tell me "a word" (and not the wonderful nonsense poetry of Lewis Carroll) an animal has. Animals do not produce words. That much is clear. The sounds they produce may signal this or that, sure. Look, I don't want to be sarcastic but I'm getting close to having my buttons pushed...I might even start to scratch the walls.
    – Lambie
    Dec 22, 2023 at 15:55
  • 1
    Granted, that might help me better understand the hallmarks of the "territorial protection" grunts which certain hominids emit when outsides call nonsense on topics from their scientific discipline. Dec 22, 2023 at 16:36

As James K says, nothing is 'close' in a general sense. A number of animals can do things that are somewhat similar to "using language". Parrots are (in principle, not automatically) good at acoustic reproduction of sounds including human speech: they "sound" the most like human speech. Some birds have an open-ended signalling capacity, where a bird has a mating song the semantic content of which is "Me. Me. Me...", but the acoustic content is changed creatively (i.e. quasi-randomly, hard to say why birds do this) in the course of an hour or so of yammering. Honey bees can probably create an infinite set of messages (about food) via iconic variables in their dance. (Domesticated) dogs are relatively good at pairing acoustic inputs with intended messages, and can acquire a large human-language vocabulary if motivated.


Research done by Con Slobodchikoff suggests prairie dogs have a sophisticated communication system that can convey detailed, varied information in their calls.


Beyond identifying the type of predator, prairie-dog calls also specified its size, shape, color and speed; the animals could even combine the structural elements of their calls in novel ways to describe something they had never seen before. No scientist had ever put forward such a thorough guide to the native tongue of a wild species or discovered one so intricate. Prairie-dog communication is so complex, Slobodchikoff says — so expressive and rich in information — that it constitutes nothing less than language.

By waiting, watching and recording, Slobodchikoff soon learned to discriminate between “Hawk!” “Human!” and so on — a talent that he says anyone can develop with practice. And when he mapped out his recordings as sonograms, he could see clear distinctions in wavelength and amplitude among the different calls.

He also discovered consistent variations in how prairie dogs use their alarm calls to evade predators. When a human appeared, the first prairie dog to spot the intruder gave a sequence of barks, which sent a majority of clan members scurrying underground. When a hawk swooped into view, one or a few prairies dogs each gave a single bark and any animal in the flight path raced back to the burrow. (Slobodchikoff suspects that, because of a hawk’s speed, there’s little time for a more complex call.) The presence of a coyote inspired a chorus of alarm calls throughout the colony as prairie dogs ran to the lips of their burrows and waited to see what the canine would do next. When confronted with a domestic dog, however, prairie dogs stood upright wherever they were, squeaking and watching, presumably because tame, leashed dogs were generally, though not always, harmless.

Something in Slobodchikoff’s data troubled him, however. There was too much variation in the acoustic structure of alarm calls, much more than would be expected if their only purpose was to distinguish between types of predator. Slobodchikoff arranged for various dogs — a husky, a golden retriever, a Dalmatian and a cocker spaniel — to wander through a prairie-dog colony one at a time. The recorded alarm calls were still highly variable, even though the intruders all belonged to the same predator class. “That led me to think, What if they are actually describing physical features?” Slobodchikoff remembers. What if, instead of barking out nouns, prairie dogs were forming something closer to descriptive phrases?

To find out, he became a participant in his own experiment. Slobodchikoff and three colleagues paraded through two prairie-dog colonies dressed in either jeans and white lab coats, or jeans and variously colored shirts: blue, gray, orange, green. The prairie dogs produced highly similar alarm calls for each person in the lab coat, except for one especially short researcher. But they chirped in very different ways for most of the different colored shirts. In a related experiment, three slender women differing in height by just a bit meandered through a prairie-dog habitat dressed identically except for the color of their T-shirts. Again the animals varied their calls. And in another study, prairie dogs changed the rate of their chirping to reflect the speed of an approaching human.

If prairie dogs had sounds for color and speed, Slobodchikoff wondered, what else could they articulate? This time, he and his colleagues designed a more elaborate test. First they built plywood silhouettes of a coyote and a skunk, as well as a plywood oval (to confront the prairie dogs with something foreign), and painted the three shapes black. Then they strung a nylon cord between a tree and an observation tower, attached the plywood figures to slotted wheels on the cord and pulled them across the colony like pieces of laundry. Despite their lack of familiarity with these props, the prairie dogs did not respond to the cutouts with a single generalized “unknown threat” call. Rather, their warnings differed depending on the attributes of the object. They unanimously produced one alarm call for the coyote silhouette; a distinct warning for the skunk; and a third, entirely novel call for the oval. And in a follow-up study, prairie dogs consistently barked in distinct ways at small and large cardboard squares strung above the colony. Instead of relying on a fixed repertory of alarm calls, they were modifying their exclamations in the moment to create something new

  • 2
    The article you linked seems to be paywalled
    – minseong
    Dec 21, 2023 at 7:57
  • 1
    blogs.scientificamerican.com/running-ponies/… "Prairie dogs have the most sophisticated vocal language ever decoded. Even better than chimps, dolphins and orcas. This could change with further research into chimp, dolphin and orca language, but right now, prairie dogs are where its at." 9 years ago! Are prairie dogs still the top animal languagers?
    – minseong
    Dec 21, 2023 at 9:45
  • 2
    @DKNguyen This is one of the reasons we expect answers to be self-contained. If you recover access, can you quote the relevant passages in your answer?
    – wizzwizz4
    Dec 21, 2023 at 17:36
  • 3
    @Lambie If that's your view then your definition is so utterly dependent on human criteria that it is completely inappropriate to apply it to non-human things. That's like holding a trial that is structured to always find someone guilty. And wow, just wow. Why do you assume I am in the US? Or that I even know that this is European finding? Because you would be wrong on both counts. You might want to do some self-reflection.
    – DKNguyen
    Dec 21, 2023 at 20:15
  • 4
    @Lambie You say, and yet you are the one who assumed I was in the US, and that I somehow knew it was a European finding and attributed me having a problem with it because of that. It's ironic you talk about personal comments when you are the one who started with them.
    – DKNguyen
    Dec 21, 2023 at 20:20

Likely second next to human language would be genetic code (DNA) of eucaryotic genome. It has the alphabet of 20 "letters" (DNA codes to amino acids) plus 2 more rarely used, also a stop code to mark the end of the coding sequence. It also has various regulatory sequences that do not code themselves, but alter the meaning of the coding regions. Same as in grammar, genetic code is error-tolerant (misspelling a letter most often leaves the word recognizable). I think the complexity approaches to human grammar, and describes how to build the living organism.

Every living cell can read its genetic code, but in plants, animals and fungi it is the most complex. Viruses cannot read they own genetic code but the host (infected cell) does for them.

  • 3
    Interesting out-of-the-box answer. How does this map to human language though? Who is the communicator? DNA seems to be the language itself, but I don't think there is an agent generating the DNA as a means of communication intentionally in this case, so I wouldn't say this answer shows off an "animal" exhibiting language
    – minseong
    Dec 21, 2023 at 9:07
  • 7
    DNA is definitely not language as the word is used on this site.
    – curiousdannii
    Dec 21, 2023 at 12:07
  • 3
    @AudriusMeškauskas DNA is not communication. (At least in natural cells. It could be used by humans to encode a message for communication, but that didn't seem to be what you were suggesting. And even then it would be no more a language than base64 is.)
    – curiousdannii
    Dec 21, 2023 at 14:42
  • 4
    The closest analogy to DNA is computer software code. And while computer programmers call this "programming language", it's markedly different from natural languages. The information it transmits from one organism to a descendant is more like bones found by an archaeologist -- would you say that the ancient animal was communicating with us?
    – Barmar
    Dec 21, 2023 at 16:02
  • 2
    @Barmar Programming is a terrible analogy for DNA. Dec 23, 2023 at 10:45

Linguists do not study animal communication. Whether this is definitive or simply dogmatic is debatable.
Nevertheless, many linguists support a negative, anthropocentric point of view on Biolinguistics.

This isn't quite correct in one view of the Max Planck Institute for the Science of Human History:

We argue that the term “cognition” has often been used by applying an anthropocentric viewpoint rather than a biocentric one. As a result, researchers tend to overrate cognitive skills that are human-like and assume that certain skills cluster together in other animals as they do in our own species.

ctrl+f “language”

  • “The Cooperative Breeding Hypothesis (Burkart et al. 2016; Burkart and van Schaik 2010; Burkart et al. 2009) considers the practice of cooperative breeding to have caused a “cascade” of effects on cognition such as changes in general intelligence, language, prosociality and social tolerance, teaching, and tool-making.

  • “Similarly, the studies of language-trained apes (reviewed in Gillespie-Lynch et al. 2014) give us some insight into what these animals are capable of with a large amount of training, although during evolution, there was no selection pressure to communicate with humans. Thus, these studies do not tell us much about apes, but rather about humans’ specially evolved skills such as language (Morgan et al. 2015; Uomini and Meyer 2013; Uomini 2009; Uomini 2014; Uomini 2017; Uomini and Ruck 2018).

(Bräuer et al. Old and New Approaches to Animal Cognition, J. Intell. 2020, 8(3), 28; https://doi.org/10.3390/jintelligence8030028)

Citations to coauthor Russell D. Gray show much more work being done in this area

A select number of species have demonstrated the recognition of their partner's role, referred to as actively coordinated collaboration, when performing cooperative behaviour. Yet, there is a lack of data on the vocal processes underlying mechanisms of collaboration in nonhuman species

(Meier, Mackenzie Ryan. "Communication, Cooperation, and Coordination: How bottlenose dolphins (Tursiops truncatus) use vocal signals to coordinate a cooperative task.")

Animals vary widely in how much they communicate. A lot of empirical effort has been invested in understanding human language evolution, with primates and song birds as main model species

(Smeele, Simeon Q. "Within and between species approaches to the study of communication and cognition in parrots." (2023).)

Bräuer et al. note that trained apes can follow a pointing gesture, limited to competitive situations rather than cooperative tasks.

It may be assumed that non-human primates are the closest living relatives to human primates. For their communication skills to be further from human's than other animals' are, would require loss

  • What do you mean "would require loss" ?
    – minseong
    Dec 25, 2023 at 5:49
  • I meant that Apes and Humans are related millions of years in the past and the genetic outfit, to the extent that it is required for language, is shared, also with species much more distantly related, but the emergent property of language, to the extent that it derives from general intelligence, has no known identification in the genetic code. So, to the extend that it is maybe genetical, apes are either closely related, which I consider a fact since sign language is proven, or they are intelligent no less than a crow and communication is a case of convergent evolution.
    – vectory
    Dec 25, 2023 at 19:54
  • [cont.] "For their communication skills to be further from human's than other animals' are, would require loss" is saying that, for the genetic outfit to be responsible AND more distant from Human genetics, that would require a loss in the ape lineages. The alternative assumption is in my view not warranted, so I cut it short at that. It would make things much more difficult, inasmuch as it seems that loss is a sufficient explanation. The problem is, of course, that "close" is not well defined in the alternative view, so I pretend (but defend) that this my view is strictly methodical.
    – vectory
    Dec 25, 2023 at 19:59
  • [cont.] (and contentious views like all of the above are straight up bullshit, unmethodological "anthropocentric" mimimi, make my cat talk talk).
    – vectory
    Dec 25, 2023 at 20:02

Not the answer you're looking for? Browse other questions tagged or ask your own question.