Hamann had two separate occasions to debate the issue. First with Herder and the second with Kant. When theorizing about the origins of language, Herder posited that reason proceeded language. Humans, he claimed, were smart animals who did not have the advantages of other species. In order to make up for their deficiencies, they invented language to symbolize what they could already discern with their reason and consciousness. The major difficulty here is that it is hard to see how one's cognition could even function apart from language. For example, how could one decide that they were going to use a series of sounds to represent a "tree" when one does not have language within which one can cogitate about the concepts of "symbol" or "sound" and half a dozen other things that one would need to make that move. Kant argued similarly. He claimed that philosophical language could express pure rationality through its purification from irrational and mythological forms of speech. We see something similar to this in the tradition of 20th-century analytic philosophy. This posits a language above language, which is impossible.
Hamann's answer was that language was of divine origin and that it was the result of God's address. God creates humans by his Word, as he did with the rest of creation. Humans are God's image and therefore can use language. Subsequently, (after Adam and Eve) human beings learn language through their parents, who are for them (in the vein of Luther) masks of God. God is the poet of the world. He is constantly addressing us as his creatures through other creatures.
Reason therefore proceeds from language and functions through language. When reason functions correctly, it is the discerning of God's speech in the structures of creation. Since reason function through language and language through tradition, there are a multitude of manifestations of reason through a multitude of traditions of rationality. The unity of reason will be revealed to us (much like the inner unity of God's works) at the eschaton. For now, the unity of reason can only be glimpsed. Therefore, this does not mean that differ languages and traditions of rationality are hermetically sealed from one another. There is a unity of reason which makes translation between different forms of rationality possible. Even speaking, comments Hamann, is a form of translation. Nevertheless, there is no pure reason, functioning apart from historical forms. Put in Lutheran terms, God and his truth can only be discerned through means.
I think that this is extremely good way of thinking about things and that it helps us clarify some issues relating to faith and reason. First, it shows, in my thinking, the failure of post-Enlightenment theology. If rationality is something traditionary, then there is no universal reason which can tell Christian theology what it can and cannot say. What there is are various forms of rationality operating within certain worldviews and traditions which intersect with one another. Some non-Christian forms of rationality may share overlap with or analogical similarity with Christian truth claims and consequently can be used in a modified form within Christian discourse. A good example of this might be the structural similarities between aspects of the ancient Greek philosophical traditions and Christians accounts of God. Early Christian theologians took these over to help clarify biblical truth claims to a Greek audience. The point is though, that such philosophical claims were not the result of a universal and neutral rationality which Christian theology was bound to respect- rather they were simply another form of tradition bound rationality. In the contemporary scheme, one can see something similar in our acceptance of the rationality of the natural sciences, which are not explicitly Christian, but share the structural similarity to Christian thought in that they posit a rationally understandable creation. A rational God who designs creation can make sense of the claim of the natural sciences that creation exists as such.