Edward Sapir (18841939). Language: An Introduction to the Study of Speech. 1921.
in many American and East Asiatic languages it must be understood to belong to a certain form-category (say, ring-round, ball-round, long and slender, cylindrical, sheet-like, in mass like sugar) before it can be enumerated (e.g., two ball-class potatoes, three sheet-class carpets) or even said to be or be handled in a definite way (thus, in the Athabaskan languages and in Yana, to carry or throw a pebble is quite another thing than to carry or throw a log, linguistically no less than in terms of muscular experience). Such instances might be multiplied at will. It is almost as though at some period in the past the unconscious mind of the race had made a hasty inventory of experience, committed itself to a premature classification that allowed of no revision, and saddled the inheritors of its language with a science that they no longer quite believed in nor had the strength to overthrow. Dogma, rigidly prescribed by tradition, stiffens into formalism. Linguistic categories make up a system of surviving dogmadogma of the unconscious. They are often but half real as concepts; their life tends ever to languish away into form for forms sake.
There is still a third cause for the rise of this nonsignificant form, or rather of non-significant differences of form. This is the mechanical operation of phonetic processes, which may bring about formal distinctions that have not and never had a corresponding functional distinction. Much of the irregularity and general formal complexity of our declensional and conjugational systems is due to this process. The plural of hat is hats, the plural of self is selves. In the former case we have a true -s symbolizing plurality, in the latter a z-sound coupled with a change in the radical element of the word of f to v. Here we have not a falling together of forms