CLASSICAL LITERATURE AND LEARNING 373 



There are, however, some other conceptions of a science of Greek 

 literature which if space permitted we might dwell upon at greater 

 length by way of introduction to our main theme, or which from 

 another point of view might even take its place. The best, the only 

 history of Greek literature which is at the same time itself a literary 

 work, is that of Alfred and Maurice Croiset. But despite its fullness 

 of matter and finish of form, it is not the final scientific construction 

 to which Professor Wilamowitz speaking for the new philology, or 

 M. Brunetiere as the representative of the science of literary evolution, 

 look forward. For very different reasons neither would accept as 

 adequate the definition of Matthew Arnold: "I call all teaching 

 scientific," he says, quoting Wolf with approval, "which is system- 

 atically laid out and followed up to its original sources." Now if 

 the sources were accessible, this definition might satisfy Professor 

 Wilamowitz. But the record, like that of geology, is full of faults 

 gaps. And to the twentieth century philologian the science of classical 

 antiquity has come to mean the fascinating art of piecing out the 

 defects of our tradition by conjectural and divinatory combination. 

 Such work is scientific in its nice weighing of evidence and its method- 

 ical use of hypothesis. Where the analogy fails is in the lack of the 

 means possessed by physical science for the control of hypothesis. 

 The consequence is that while classical science slowly advances with 

 wasteful, but, in the sum, not wholly ineffectual toil, the flower of 

 classical culture and the fruits of classical education are choked by 

 a riotous overgrowth of highly specialized pedantry and unverifiable 

 conjecture. In spite of the forty thousand emendations of ^Eschylus. 

 it may be doubted whether the most recent texts of the Agamemnon 

 are any improvement upon those of the eighteenth century. The 

 hair-splitting refinements and the formidable terminology of modern 

 syntax have not impaired the point of De Maistre's observation that 

 "since they have taught us how to study Latin, nobody really learns 

 it." And the dreary literature which has gathered about Homer. 

 Plato, and Cicero, if it establishes nothing else, amply proves that the 

 sane interpretation of great world books depends far more on 

 the total culture which the individual reader brings to their perusal 

 than it does on any collective progress of "science." 



But this is by the way. There can be no question but that in some 

 fields there is real progress in the filling out of the record. This is 

 notably the case in the domain of Attic institutions and Attic law. 

 where combination and conjecture are at once stimulated and con- 

 trolled by the new material supplied by inscriptions. The same may 

 be said of the history of Greek art, which has been completely re- 

 constructed since Winckelmann, and of that history of Greek religion 

 whose future outlines we can dimly discern. How far is it or can 

 it be true of literature? We may hope for anything in what have 



