The scientist is very apt to carry a chip on his shoulder when the word evolution is mentioned. He seems to feel it treason to science if evolution is not regarded as a universal principle, as absolute in its operation as the law of gravitation. Because he believes in progressive development and the survival of the fittest, he mechanically postulates that whatever is is better than all that has passed away. Applied to the institutions of men this principle is abundantly untrue. If it comforts one to classify the differentiation of organ and function from diatom up to man, and the general simplification of structure observable in the historical development of languages as they grow older, under the one label, evolution, he is welcome to do so, but he must meet the difficulties and see the differences. If it is a simplification that the Romance languages have replaced Latin synthetic cases by prepositional phrases, why, after having acquired an analytical future, did they convert it into a synthetic: why has Spanish developed hablarèis "you will speak" from hablar habéis "you have to speak"? Who will assure us that the Latin case-endings did not similarly arise from some sort of attachments of prepositions to their nouns? Why is-(i)bus too heavy for a mere termination? [What is there about -bus that catches the ear of persons who hear Latin? Shakespeare's Costard hits off some of the catching elements of spoken Latin in his honorificabilitudinitatibus, and I can testify to the prominence of-bus in the gibing attempts at Latin word-formation I have heard from mockers.] Who can seriously maintain that-"bus" attached to a Latin stem is inherently any more ambiguous than "by" or "with" prefixed to an English noun? What-bus was to start with, philologists surmise, they do not know. But they do know that Spanish migo is Lat. mecum synthetized and reanalyzed again in con migo which is cum mecum. The psychology of the doubling they understand, but they don't drop the -go from migo; and they accept the fact for the fact, content with the unlettered ancestry of Spanish or Latin or English. Who then, I repeat, shall assure us that the Latin case endings did not originate similarly from some prepositional affix? It is absolutely certain that Latin amabat "he was loving" has been synthetized from an independent word meaning, either "loving" (ptc.) or "for-loving" (infin.) plus -bat "was."
The truth is that all along the line language submits itself to synthesis. We have an interesting exhibition of this in the colloquial American "kinder" and "sorter," for which our language has, I conceive, a real need as a verb modifier. At least I can not express in formal language the very pretty group of associations suggested to my mind by the phrase "He sorter sidled up to her and whispered." Analytic and synthetic are but relative terms and sometimes the synthetic form of expression is the simpler. To me at least, in the French line quoted above, de la Tyrannie, though analytic, presents to me in a blur, as