At the beginning of his book Danubio, Claudio Magris jokes about the impossibility of finding the true source of the Danube. The closer you get to it, the less definite the answer seems.
Living in the Grisons as I do, I often puzzle over the source of the Rhine and what counts as part of the Rhine. The so-called Anterior Rhine rises in the Surselva highlands, apparently at a lake, and the Posterior Rhine rises further south, somewhere near the mountain called Rheinwaldhorn. These two Rhines have their confluence at Reichenau, and then the Rhine flows past Chur and through a valley northwards to empty into the Eastern end of Lake Constance. It flows out the Western end of the Lake, past Basel and then up into Germany, heading for the North Sea. I have often wondered, as I take my walks along the young Rhine near Chur: who says this is “the Rhine” anyway? And who says the part that flows out of the far end of lake Constance is the same river? It all seems pretty arbitrary. And yet we think of each of the mighty rivers of Europe as springing from a tiny Alpine or similar mountainous source.
“Upstream” and “downstream” are powerful metaphors for both historical and logical derivation. Logically, the more complex and diverse is supposed to derive from some simple principle. Historically, the complex and diverse but related is supposed to “spring from a common source”.
So it is with languages in the Western world, beginning with the exploration of Indo-European in the 19th century. Linguists dreamed of discovering an ancient language from which all the others could be derived: an Ursprache spoken in a supposed Urheimat from which the original Aryans had migrated. First many thought it was Sanscrit. Then the idea imposed itself that the proto-language had been lost, but could be reconstructed scientifically.
As time went on, however, competing ideas muddied the waters, as it were. The “wave theory” proposed that linguistic innovations arose among the derived languages and spread to one another over time. The “substrate theory” proposed that the differences between the derived languages were caused by the influence of non-Indo-European languages spoken in the corresponding regions of Europe. Today the picture does not seem as clear as it once did.
Yet we still long to find the one source upstream which will explain the whole “river system”. Alas, there may not be one, any more than there is one indisputable source of the Rhine. There may be no unifying explanatory principle for all the downstream variety – just more variety upstream. The code of codes, the code that explains all the other codes, may be an illusion.