Page:The Mediaeval Mind Vol 2.djvu/140

128 a real difference could not fail to assert itself inasmuch as logic was in truth of universal application, while mediaeval grammar never ceased to be the grammar of the Latin language. Nevertheless its terminology was largely drawn from logic.

So dialectic brought both good and ill, proving itself helpful in the regulation of syntax, but banefully affecting grammarians with the conviction that language was the creature of reason, and must conform to principles of logic. One likewise notes with curious interest, that, from their dialectic training apparently, grammarians first found as many species of grammar as languages, and then forsook this idea for the view that, in order to be a science, grammar must be universal, or, as they phrased it, one, and must possess principles not applicable specially to Greek or Latin, but to congruous construction in the abstract; "de constructione congrua secundum quod abstrahit ab omni lingua speciali," are the words of the English thirteenth-century philosopher and grammarian, Robert Kilwardby. A like idea affected Roger Bacon, who composed a Greek grammar, which appears to have been intended as the first part of a work upon the grammars of the learned languages other than Latin. It was adapted to afford a grounding in the elements of Greek: yet it touches matters in a way showing that the writer had thought deeply on the affinities of languages and the common principles of grammar. Of this the following passage is evidence:

"Therefore, because I wish to treat of the properties of Greek grammar, it should be known that there are differences in the Greek language, to be hereafter noted in giving the names of these dialects (idiomata). And I call them idiomata and not linguas, because they are not different languages, but different properties which are peculiarities (idiomata) of the same language. Wishing