Skeat believes that Chatterton had read very little of Chaucer, probably
only a small portion of the Prologue to the "Canterbury Tales." "If he
had really taken pains," he thinks, "To _read_ and _study_ Chaucer of
Lydgate or any old author earlier than the age of Spenser, the Rowley
poems would have been very different. They would then have borne some
resemblance to the language of the fifteenth century, whereas they are
rather less like the language of that period than of any other. The
spelling of the words is frequently too late, or too bizarre, whilst many
of the words themselves are too archaic or too uncommon."[14] But this
internal evidence, which was so satisfactory to Scott, was so little
convincing to Chatterton's contemporaries that Tyrwhitt felt called upon
to publish in 1782 a "Vindication" of his appendix; and Thomas Warton put
forth in the same year an "Enquiry," in which he reached practically the
same conclusions with Tyrwhitt. And yet Warton had devoted the
twenty-sixth section of the second volume of his "History of English
Poetry" (1778,) to a review of the Rowley poems, on the ground that "as
they are held to be real by many respectable critics, it was his duty to
give them a place in this series": a curious testimony to the uncertainty
of the public mind on the question, and a half admission that the poems
might possibly turn out to be genuine.[15]

Tyrwhitt proved clearly enough that Chatterton wrote the Rowley poems,
but it was reserved for Mr. Skeat to show just _how_ he wrote them. The
_modus operandi_ was about as follows: Chatterton first made, for his
private use, a manuscript glossary, by copying out the words in the
glossary to Speght's edition of Chaucer, and those marked as old in
Bailey's and Kersey's English Dictionaries. Next he wrote his poem in
modern English, and finally rewrote it, substituting the archaic words
for their modern equivalents, and altering the spelling throughout into
an exaggerated imitation of the antique spelling in Speght's Chaucer.