Page 256 - גנזי קדם ז

26* Jay Rovner

surviving witness ofthis particularmidrashic text, ofthe form ofthe independent
micrographic carpet page, or of both.

In order to gain greater certainty regarding the nature and purpose of the
carpet page, its relation to the midrashic text appearing on its other side must
be clarified. If the carpet page was copied first, it would be understandable
that a blank verso had been left, and that when the carpet page was no longer
needed, someone might have decided to copy the midrashic text there. After
all, parchment was an expensive commodity, and would not be allowed to go
to waste. On the other hand, if the midrash were copied first, then the carpet
page would likely have been intended as a decorative outer cover, something
imitative of a binding.29 It the latter were the case, it is more likely that we
would be looking for reasons other than the liturgical-aspirational-talismanic
ones mentioned above, whether to replace them or to supplement them, in
explication of the putative function of the carpet page.

The matter is not quite so simple, however. The midrashic text is cut short
by one or two lines. On close inspection, moreover, it appears to have been
cropped on all sides. There are virtually no upper or right-hand margins, and,
on the left, the integrity of the endline letters themselves has been affected by
the cutting (see the text supplied in brackets at the ends of several lines in
the transcription, Appendix 1 below). No copyist would use a layout of such
sparse margins. As for the bottom, it is difficult to establish, given the state of
the sheet, whether the missing lines have been cut away, or whether they were
continued on another sheet. If the missing lines were cropped when the bottom
of the page was, then it is further evidence that the midrashic text came first. If
they were continued onto another sheet, then one wonders why the copyist did
not simply use the other side of the page,30 and the most logical answer would

29 It was customary to leave the outside page of a manuscript blank to protect the text from
dirt and debris, and to secure a full opening for the reader’s first experience of the text.
By this time, on the other hand, manuscript codices could either be kept in a container or
attached to boards, and the initial pages (and even leaves) could be decorated, which is why
carpet pages are located there.

30 It is certainly possible that this in fact occurred. Unfortunately, if this other side originally