Actually, there is plenty of research done in the area. If my startup gets sold, I'm going to venture there myself, trying to pry it with a combination of information theory and some home-made tools.
The brains of dolphins started growing faster since 35 million years ago (can't find the source, sorry), which means, earlier than our ancestors the great apes. The overall consensus is that despite the fact that they don't use tools (actually, they do, but only rudimentary ones), the dolphins' social lives are as complex as those of humans.
I remember one researcher was saying, "Socially, they have nothing to learn from us". Basically, imagine if the men would not develop tools but instead develop social organisation.
It's been established that:
- bottlenose dolphins have "signature whistles" (or names, if you will) for each other
- they were taught to understand some human language and managed to deduce some grammar rules (when object / subject were swapped)
- entropy test of the bottlenose dolphins' whistles yields results very similar to those of human speech. (In a nutshell, entropy signatures are used to distinguish between different types of information; they are used pretty much universally.)
- they demonstrate some of the traits which are less common among animals (e.g. cruelty to some other species and sometimes each other, or altuism on the opposite, sex for recreation rather than procreation, homosexuality)
It has not been established that the language they use is categorical like human or can refer to abstract entities. Some researchers claim it is the case, but there are no proofs. I am, however, inclined to believe this is the case, because of the entropy test, and since the dolphins demonstrated the ability to learn categorical language, and because they use names.
Worth looking at. It is difficult though because they communicate in ultrasound whistles and clicks.