A main tool of historical linguistics is the set of rules of sound and grammatical transformation governing the language change. One language evolves into another due to cultural or geographic separations of peoples due to migrations or other cultural displacements, such as conquest.
Using the rules of historical linguistics, it appears to be possible to discern patterns of change and to determine which language has shifted into the other.
One such rule is the softening of consonants over time. Thus, for example, the "v" in the Sanskrit "Veda," meaning knowledge, is transformed into the softer English "w" in "wit," "witten," "wisdom" and the German "wissen," which also means knowledge, and derives from the more ancient Sanskrit root. The Sanskrit "deva" is transformed into the softer Latin "deus," Greek "theos," Lithuanian "dewas," Irish
"dia," and Old Prussian "diews." Using such transformation rules, linguists attempt to reconstruct which languages are earlier and which broke off later in the transmutation of language. Historical linguists assume that these rules are constant over time and that they apply to early transformations as well as later ones.
If we assume that the basic rules of language transformations are constant and do not mutate over time, then these conclusions follow. But could there have been
sound shifts in the opposite direction at much earlier times in history? Perhaps different laws applied at the time when Vedic Sanskrit changed from and to other languages. Consider that there are also changes in the reverse direction. For
example, the "g" in the Sanskrit "go," (meaning cow) is transformed into the harder consonant "k," to make the German word "kuh" for cow. The English word "cow,"
pronounced with a hard "k," is a harder, guttural form than the "g" in the Sanskrit "go."