Page:Illustrations of the history of medieval thought and learning.djvu/105

Rh school of Roscelin proceeded from the opposite extreme, from experience. Our senses, it was felt, are the only certain warrant for existence, and they only reveal to us the individual. The universals, therefore, the cardinal point of dispute, could only be our own generalisations from observed facts.

Roscelin, who brought the latter view prominently into the field of discussion, was not, however, as is commonly presumed, nor was Berengar, the first nominalist of the middle ages. This position, according to an early chronicler, belongs to John the Sophist, whose identification with John Scotus, long suspected, has been made probable by the acute arguments of Dr von Prantl. So hearty a Platonist as the Scot could not but be a realist iv. in his ontology: but equally little could a logician escape the influence of Aristotle, the philosopher to whom he owed his method; and John's view of language and of the scope and functions of logic is far removed from the arid tradition of Isidore and Alcuin. Dialectic, he admitted, had kinship with grammar and rhetoric, in so far as it dealt with human speech pure and simple. But words and thoughts, and therefore words and things, were definitely if imperfectly correlative. John therefore claimed for dialectic a higher dignity than that of a mere mechanical instrument: it was the searcher out of the common conceptions of the mind, the guide of reason.

It was easy to carry this train of reasoning a stage further, and to argue that the general terms with which logic occupies itself are not its source but its product. The universals, the Scot had agreed, are words; what if they be mere words? Already in his lifetime the suggestion was taken up by Heric of Auxerre, whose pupil, however, saint Remigius of Auxerre, reverted to a declared