This is the Italian soprano Luciana Serra singing an aria from Mozart's The Magic Flute. Wonderful, isn't it? Now, the question is: did we evolve to produce this kind of performance?
In another video (sorry, embedding disabled), we can see an example of throat singing from a region in Central Asia called Tuva. This man is able to produce different types of voices using the overtones created in his throat. It looks incredible, but it's possible. In other words: at least in some of us, if not in the vast majority of us, there is a potential for this kind of thing. If you talk to people who do yoga or meditation, or other sorts of physical or mental exercise, they will often tell you that they have discovered something inside them that they didn't know existed. Humans are full of all kinds of potential, including the vocal ones, but our languages only use a very small portion of these possibilities, disregarding the rest as irrelevant. The principle of economy works here: in fact, learning to pronounce the phonetic repertoire of a language requires great effort, and some people are never completely able to master the whole set. The problem gets much worse when we get older and try to learn a second language. By then, we have lost most of the mental flexibility or predisposition that we had as children and find it really hard to produce or imitate the new sounds. The story is well known, and we can find examples everywhere everyday. I know poeple who have learnt Spanish at extraordinary levels of proficiency but who still have problems pronouncing words like piscina or decisión, or find it hard to distinguish between caro and carro.
Children are not born speaking a language. They are born with the mental ability to see the logic of human communication, and during the learning process they have to explore the vocal possibilities offered by their own bodies. But it is only some of these vocal possibilities, in fact a finite set of vowel and consonant sounds (plus suprasegmental elements), that are selected and promoted in each case.
Let's imagine another hominid species with a poorer repertoire of possible vocalizations. Even if their imagianry IPA chart were ten times smaller than ours, they would still have at their disposal a considerable amount of elements to choose from. It's not just what you have but how you exploit the potential that you have. It comes as an obvious conclusion that other hominid species could very well have developed verbal language even if their vocal tract was quite different from ours. The only requisite is that they had the kind of logical thought that leads to human communication.
Some people say language is what makes us human. I think language is just a secondary factor in a much wider scenario: the one created by our minds. Maybe that's why there are so many people doing yoga, or experimenting with sounds or trying to break communicative barriers. They want to break away from the boundaries of finite sets. They want to get a sample of a more global type of human interaction.
Or just to have fun: