But when by means, which I cannot conceive, our new grammarians began to extend their ideas, and generalize their words, the ignorance of the inventors must have confined this method to very narrow bounds; and as they had at first too much multiplied the names of individuals for want of being acquainted with the distinctions called genus and species, they afterwards made too few genera and species for want of having considered beings in all their differences; to push the divisions far enough, they must have had more knowledge and experience than we can allow them, and have made more researches and taken more pains, than we can suppose them willing to submit to. Now if, even at this present time, we every day discover new species, which had before escaped all our observations, how many species must have escaped the notice of men, who judged of things merely from their first appearances! As to the primitive classes and the most general notions, it were superfluous to add that these they must have likewise overlooked: how, for example, could they have thought of or understood the words, matter, spirit, substance, mode, figure, motion, since even our philosophers, who for so long a time have been constantly employing these terms, can themselves scarcely understand them, and since the ideas annexed to these words being purely metaphysical, no models of them could be found in nature? I stop at these first advances, and beseech my judges to suspend their lecture a little, in order to consider, what a great way language has still to go, in regard to the invention of physical substantives alone, (though the easiest part of language to invent,) to be able to express all the sentiments of man, to assume an invariable form, to bear being spoken in public and to influence society: I earnestly entreat them to consider how much time and knowledge must have been requisite to find out numbers, abstract words, the aorists, and all the other tenses of verbs, the particles, and syntax, the method of connecting propositions and arguments, of forming all the logic of discourse. For my own part, I am so scared at the difficulties that multiply at every step, and so convinced of the almost demonstrated impossibility of languages owing their birth and establishment to means that were merely human, that I must leave to whoever may please to take it up, the task of discussing this difficult problem. "Which was the most necessary, society already formed to invent languages, or languages already invented to form society?"