Quintilian, Institutio Oratoria 1.8:
“To investigate what everyone, including the most worthless people, ever wrote is an example either of excessive misery or empty ostentation, which both detains and destroys minds which would better be left free for other things. Someone who has shaken off all of the dusty fragments unworthy of reading could just as easily spend their labor upon the tales of old women. The commentaries of grammarians are full of these kinds of pitfalls, and even their authors barely know what is in them. It is said of Didymus, than whom no one ever wrote more, that when he was arguing against some tale as though it were wholly fictitious, his very own book was brought against him, since it contained that very tale. This happens especially in fantastic tales, all the way to the point of absurdity and shame, whence comes the excessive license granted to every knave of inventing things such that they safely lie about entire books and authors as they find it convenient, because those things which never existed cannot naturally be discovered. For, in more generally known matters, the forgers are discovered by curious inquisitors. And so, it seems to me that among the grammarian’s virtues will be numbered not to know some things.”
Persequi quidem quid quis umquam vel contemptissimorum hominum dixerit aut nimiae miseriae aut inanis iactantiae est, et detinet atque obruit ingenia melius aliis vacatura. Nam qui omnis etiam indignas lectione scidas excutit, anilibus quoque fabulis accommodare operam potest: atqui pleni sunt eius modi impedimentis grammaticorum commentarii, vix ipsis qui composuerunt satis noti. Nam Didymo, quo nemo plura scripsit, accidisse compertum est ut, cum historiae cuidam tamquam vanae repugnaret, ipsius proferretur liber qui eam continebat. Quod evenit praecipue in fabulosis usque ad deridicula quaedam, quaedam etiam pudenda, unde improbissimo cuique pleraque fingendi licentia est, adeo ut de libris totis et auctoribus, ut succurrit, mentiantur tuto, quia inveniri qui numquam fuere non possunt: nam in notioribus frequentissime deprenduntur a curiosis. Ex quo mihi inter virtutes grammatici habebitur aliqua nescire.