Even when books came into vogue, the collection of a Master was freely subject
to alteration at the hands of his pupil, who, performing as it were the
functions of an editor, selected or omitted passages at pleasure, and even
added (but always with his name) new matter of his own, and sometimes
collections of fresh traditions from other sources. The work, notwithstanding
these alterations, was still known under the Master's name. It is thus that we
find different versions of such compilations, as that of Bokhâri, to vary
both in the number of the traditions and in the subject-matter. It is also
sometimes not easy to trace the original work from which quotations are made.
Tabari; for example, who composed his annals almost entirely of extracts
copied verbatim from previous collections, makes little mention of the Author
from whom he borrows: it is the name of some obscure Sheikh under whom he read
the work which, under the pedantic rules of tradition, figures as his
authority; the name of the real author (Ibn Ishâc, for instance) occurring in
the middle of the long string of vouchers, as a mere link in the transmission.
When he had read a collection under more than one Sheikh, he makes a parade of
his learning by quoting now under the name of one, and now of another. And to
carry the system to the extreme of absurdity, where he had read only part of a
work with a Master, he quotes the part he had not so read under the fiction of
a letter from his Sheikh; letters being admissible as evidence, but not
a manuscript or book!