Unless we are simply going to throw our hands up and say that ANY comparative analysis of animal modes of communication is doomed from the start, then we have to be able to agree on some basic features of communication, at various levels of abstraction, and try to identify them in nature.

I don't think that's anthropomorphizing, any more than it would be anthropomorphizing to look at the skeletal structure of different animals and say "the orientation of the skull and the spinal cord in this species indicates bipedalism, but in this species, where the spinal column enters the skull at a different angle, indicates quadripedalism." You wouldn't say that was bipedal chauvinism, at least not if there was plenty of observational data to support the conclusion that the one group walks upright in the wild, and the other on all fours.

I'm using the word "symbol" here in a narrower sense, as that concept has been used since at least the philosopher C.S. Peirce, to refer not just to a relationship between a sign and some referent. An icon and an index satisfy this definition as well. But a symbol does something else. It relates not just to an object, but to other symbols as well, in a web of logically interconnected concepts that can be combined and recombined endlessly in order to create previously unthought ideas.

My point is not that it's impossible for other animals to do this (although I do think you need cognitive complexity of the kind generally only demonstrated by social mammals, in particular the ability to "model" another creature's mind inside your own, so as to pass your meaning from one mind to the other). What I'm saying is that they rarely, if ever, do it spontaneously, and that IF THEY DID, we would be able to observe that behavior and recognize it. (Check out @colbyhess's comment to this article, where I agreed with him that other intelligent life forms in the universe would probably be intelligible to us, given that certain concepts like mathematics are what philosopher Daniel Dennett called "forced moves" in evolutionary design space.) Large-brained animals capable of symbolic thought will be drawn to certain concepts, and attendant behaviors, by a kind of conceptual gravitation.

I don't discount that nature has found a lot of very interesting ways for animals to experience their world and to communicate (check out Ed Yong's new book "An Immense World"), but there is a difference between a causal connection between sign and object, and some resulting behavior, on the one hand, and behaviors that indicate the presence of combinatorial thought processes utilizing a whole universe of abstract concepts, on the other.

I think it's possible to know when any species is engaged in that kind of behavior. And I even think we see the initial stirrings of it in many social mammals like primates and cetaceans. But I don't see them refashioning their worlds as we have done, because I don't think they have yet acquired the cognitive tools necessary to "paradigm shift" away from the concepts that come pre-built into their intuitive ontologies.

That's not to say that humans are so special. We can't fully escape our intuitive ontologies, either, and wouldn't want to, since our moral reasoning depends on them. And the vast majority of human beings, from the past right down to the present, have not bothered to question the basic physical and social realities they encounter. Most of the ones who do are labelled madmen. A few are called geniuses.

Nevertheless, the quantity and quality of cognitive "play" that humans engage in finds no parallel in the rest of the animal kingdom. Perhaps other species of the genus Homo could boast as much, but we killed them off, so we'll never know.

(Since you brought up the pheromone-based communication system of ants, I have to recommend a fantastic sci-fi trilogy called Children of Time, by Adrian Tchaikovsky. He imagines a genetically enhanced species of intelligent spiders, and what their form of communication would look like. Not surprisingly for animals that have no ears, they use a vibration-based system that relies on relaying messages through their webs, which line every inch of their dwellings. Anyway, they eventually develop computing technology, but it's completely organic, using ants and leveraging their pheromones to create a kind of programming language that causes the ants to organize in ways that store data and execute commands. Note that the ants are not conscious of their individual selves the way the spiders are. It wouldn't even do to say that they are communicating with each other, but are rather that the spiders communicate THROUGH them.)