In the 18th century, the figure of the Naturkind was the source of much speculation, the most well-known product of which is, perhaps, the tale of Kaspar Hauser. Philosophers supposed that if a child were to grow up without any exposure to human society, the resulting adult would be a living model of what it is a human being is most purely or naturally. One of the most pressing questions such a child would be able to answer, it was thought, was the matter concerning the origins of language: would it be able to speak at all? And, if so, what sort of words would it use?
Somewhat earlier --peaking as a topic, I think, in the mid-17th century-- there was an intense debate as to which language, if any, was the language spoken prior to the fall of the tower of Babel. The most common candidate was Hebrew, but Chinese, Egyptian, and others are also mentioned. Some argued that the Hebrew of which we have traces was not the real Hebrew, and that there is an earlier, 'Adamic' language that, if rediscovered, would disclose divine knowledge to humans.
Interestingly, in Herodotus, there is a sort of combination of these two varieties of speculation. In Book II of the Histories we are told of the Egyptian king Psammetichus, who wishes to determine whether or not his nation is in fact the oldest on earth. He sets up the following experiment:
He gave a shepherd two new-born infants to rear, from ordinary families, and told him to bring them up among his flocks in such a way that no one ever spoke in their hearing; they were to lie in a remote hut by themselves, and he was to bring them she-goats from time to time, give them their fill of milk, and do whatever else needed doing. Psammetichus made these arrangements because he wanted to find out what the children's very first word would be, once they were past the age of meaningless whimpering. And so it happened. One day, when the shepherd had been carrying out this programme for two years, he opened the door and went into the hut, and both children rushed up to him, reaching out their hands, and said 'bekos' (II 2).
It turns out, as Herodotus explains, that 'bekos' is the Phrygian word for 'bread', and that from this it may be concluded that it is the Phrygians, and not the Egyptians, who are the most ancient of all people. By this peculiar reasoning, the oldest language is the language that one would naturally speak if not pushed off course by another, newer language. No explanation is given of language change, but presumably Herodotus, or the person who told him this tale, would suppose that all non-Phrygian words for 'bread' are in fact deformations or corruptions of 'bekos'.
This does not seem to imply any special status among languages for Phrygian, or any indication of a need for a return to the natural way of saying things. In this respect, early modern speculation about the origins of language --both the language acquired by an individual over the course of a lifetime, as well as all the natural languages of the world and the relations between them-- could not have been more different. So far as I know, the search for the first language and the speculation about what a Naturkind would be like were never brought together into a single line of questioning.
This is because, unlike Psammetichus and Herodotus, the early moderns supposed that whatever language a Naturkind might end up using, it would not be the original, primordial language of humanity; when the debate as to what that primordial language was was at its peak, moreover, a century before Rousseau and other speculated about children of nature, it was universally assumed that this was not a 'natural' language, that it would not congeal spontaneously over the course of a childhood suckling at the udder of a she-goat. Instead, it had a divine source, and so all subsequent 'natural' languages, like Latin and French, were corruptions of it not in the way that Egyptian was a corruption of Phrygian, but in a way that implied a loss, or, to put it in more overtly theological terms, a Fall.