[flueedo]

If you put a bunch of humans (children) together, each with no previous knowledge of language, on an isolated island, they will in all likelihood come up with one. (Something close to this, as a live example, is a documented case I remember having read years ago of an all-new sign-language that emerged spontaneously among a large group neglected deaf-mute children at an institution a few decades ago. I think this was in Nicaragua or some other country in that region) What I mean is, our linguistic ability isn't dependent on culture (though tremendously enriched by it), it's biological. Why wouldn't the same hold for dolphins. I think they're very smart, but I don't believe they ever had something we could decently call a language.

[jacobolus]

http://www.radiolab.org/story/91725-words/

https://en.wikipedia.org/wiki/Nicaraguan_Sign_Language

[RevRal]

Language is a strange and large territory. And, almost by definition, any social animal must be able to communicate with each other; And, in order to communicate, there must be shared formal rules. In essence: syntax.

[ThomasBombadil]

Well, yeah, but those rules will only be unique and useful for probably 5 or 10 dolphins. The rest of dolphin kind won't have any innate concept of the vocabulary they've developed.

Grammar, syntax, vocabulary, implicit context, mood and tone. You'll be starting at square one, with every separate pod, with minimal potential of cross-pollination across pods, based on dolphins that leave one group, and join another, and whatever they bring with them, and manage to learn from new dolphins they meet.

Even the presumable token words, which surely must exist: air, water, fish, dirt. One pod in isolation from the rest might not build these conceptual ideas, or use their noises the same way, and then what of another pod halfway around the world, which they've never met?

Sure, the capacity for language remains, but there's no persistent implementation. Each variant of a dolphin language dies with the dolphin that knows it.