The more I learn about dolphin communication, the less eager I am to call it “completely non-linguistic” instead of “proto-linguistic”. They’re even able to understand subject and object in SOV, I remember an experiment using signs like:
ball plank put - the dolphin places the ball over the plank
plank ball put - the dolphin tries to place the plank over the ball
And baby talk is a type of register - the speaker is actively changing his speech because of the situation (the hearer, in this case). It’s rather surprising to find it in another animal species than ours.