Skip to content
NatureHQ

A guided journey · 5 stops

From a sound to a meaning

Start with an animal that answers to its own name, and work outward to what it would actually take to call any of this a language.

The route

  1. Dolphins
  2. Animal communication
  3. Do animals have language?
  4. Animal culture
  5. Animal intelligence

The question underneath almost every popular story about animal communication is whether animals talk. It is a bad question, in the specific sense that it cannot be answered as asked, and the route below is an attempt to replace it with questions that can.

It begins with the single strongest result anywhere in the field: a dolphin, played a synthesised copy of its own signature whistle with every trace of the original voice stripped out, answers. That is a learned, arbitrary label with a receiver who responds to it — one of the properties language is made of, demonstrated cleanly. Starting with the strongest case is deliberate. It is much easier to see what is missing from animal communication once you have seen what is genuinely there.

From there the route widens: to communication as a whole and the standard of evidence it runs on, to the language question taken apart into properties rather than argued as a yes or no, to culture, which is where signals stop being facts about individuals and start being facts about populations, and finally to intelligence, which is the frame everything else keeps getting collapsed into and should not be.

  1. Stop 1 of 5. Marine life

    Begin with the cleanest result in the field

    A dolphin invents a whistle for itself, other animals in its group learn it, and playing a voice-stripped synthetic version back to the owner makes it answer. Every part of that sentence was a separate experimental problem, and the controls — a familiar animal’s whistle, an unfamiliar animal’s whistle — are what make it evidence rather than an anecdote. Starting here gives you a calibrated sense of what a strong result in this field actually looks like, which is the thing most readers have never been shown.

    Read DolphinsA dolphin invents its own whistle as a calf and answers to it forty years later — even when the voice is stripped out.

  2. Stop 2 of 5. Senses and abilities

    Learn the standard the result had to clear

    The dolphin work is convincing because of its controls, not because of its conclusion, and the general principle behind those controls governs the entire subject: information in a signal is a fact about the signal, and communication is a fact about the receiver. This page is where that standard is set out, along with the design — playback — that almost every strong finding in the field rests on. It is also where the difference between a signal and a cue gets drawn, which the next page needs.

    Read Animal communicationA vervet monkey has one call for leopards and another for eagles — and the others react correctly with nothing there.

  3. Stop 3 of 5. Behaviour and cognition

    Take the question apart instead of answering it

    Now the question can be asked properly. Language is not one possession; it is a bundle of properties that happen to co-occur in humans, and each can be looked for separately. Arriving here with the dolphin result and the playback standard already in hand is what makes the property-by-property answer land — you can see which cell of the table the whistle fills, and which cells nothing fills at all.

    Read Do animals have language?Language is not one thing. Take it apart into its properties, and animals have several of them — just never all at once.

  4. Stop 4 of 5. Behaviour and cognition

    Move from individuals to populations

    Everything so far has been about what one animal does to another. Culture is what happens when the same signal is learned rather than inherited and spreads through a population — and it is the property where animals are least equivocally like us. Humpback song sweeping across the Pacific and being replaced wholesale is not a subtle statistical effect; it is a wholesale change in what an entire ocean basin sings, transmitted by learning. This step also reframes the previous one: cultural transmission of form is a linguistic property, and it is the one animals most clearly have.

    Read Animal cultureA great tit that works out a better way to open the box will often abandon it, and do it the way its neighbours do.

  5. Stop 5 of 5. Behaviour and cognition

    Refuse the ranking the whole subject collapses into

    Popular coverage funnels all of this into a single scale with humans at the top, and the route ends by taking that apart. A honey bee encodes a vector to a food source and cannot refer to yesterday; an octopus solves problems with a nervous system arranged nothing like ours. These are not positions on one line, and the temptation to make them into one is the single most reliable source of error in the field. Ending here rather than beginning here is deliberate: the argument is much harder to dismiss once you have seen four pages of specifics.

    Read Animal intelligenceA bee understands zero. A nutcracker remembers thousands of buried seeds. Neither could do the other’s job, and no ranking survives that.

Where this leaves you

You end knowing why "do animals have language?" is answered no and why that no is not deflationary. Several species have learned arbitrary labels. At least one has combinatorial structure. Cultural transmission of signal form is established beyond argument in humpback whales. What nobody has shown in any species is displacement — a signal about something that is not there — and it is genuinely unclear whether that is because it does not occur or because our experiments are worst at detecting exactly that.

The route also leaves you with a working standard you can use on the next headline you meet: a pattern found in a recording is a fact about the recording. It becomes a fact about the animals when somebody plays it back and something changes.

Last reviewed 2026-09-02. All guided journeys