haps the entireGorgias.^14 If this is correct, we would seem to have the
compiler looking at an actual volume, examining it, and puzzling a bit about
what it might be: perhaps the usual title was lacking, or the papyrus was
fragmentary, forcing the compiler to try to identify the work. Seeing the
name of Callicles, he decided, perhaps, that he was dealing with a dialogue
called theAgainst Callicles. The last lines of the papyrus, in which authors’
names are given together with the phrase ‘‘as much as is found,’’ have
occasioned much discussion. Although no scholarly consensus has yet
emerged, most scholars take the phrase as a Greek equivalent ofomnia
quae extant(‘‘complete extant works’’), and I accept that suggestion here.
15
I take this, then, as the inventory of a real collection of books. The
occasion of such an inventory might be, for example, a purchase or sale (as
in Cic.Fam. 16.20), a gift (such as those made by the ephebes to the
Ptolemaion at Athens, attested inIG 2
2
.1009, 1029, etc.), an inheritance,
or perhaps an owner simply wanting a list of his books in order to
determine what he lacked and needed to buy (as in Cic.Hortensiusfrg.
8 Grilli).
16
In the case of a large collection or a library, a list of this sort
could help a user, whether that user was the owner, a friend of the owner,
or a scholar given permission to use the collection,^17 to determine if a
given book was present and so avoid a lengthy and perhaps frustrating
search among the rolls themselves.
If we apply such criteria to each of the lists in Otranto, we end up with
eight likely inventories (or what survives of inventories) of actual book
collections. I set these out, insofar as possible in chronological order, in
table 10.1.^18
We can make some preliminary observations, none of them very sur-
prising. We would expect most papyri to date primarily from the second
and third centuries, and they do; and the provenance of most of our
- Puglia 1996, 52 4 and 58 9, provides a full discussion of this problem.
- The matter, however, is quite uncertain. Harrauer 1995, 66, arguing that
‹óÆ ıæßóŒåôÆØis not the way a Greek would say ‘‘complete works,’’ took lines 28 30 as
constituting a commission or order (‘‘Desiderataliste,’’ ‘‘Suchliste’’). In favor of his suggestion
is the fact that we find the verbıæßóŒøused in connection with a request for books inP.Oxy.
2192, lines 41 3: ‘‘If you find (Kaí åæßóŒfi Åò) any books I don’t have, have copies made and
send them to me.’’ On the other hand, it is almost impossible to take the first 27 lines of this
list as a list of desiderata, given such items as the odd opisthograph of line 7, which must
surely be an existing volume, and it seems very unlikely that someone who owned so much of
Plato would not already own Homer, Menander, and Euripides. See Puglia 1996, 59 60, for
a full discussion. - Johnson forthcoming points out that some such list is implicit in the request for
books inP.Oxy. 2192, cited in the previous note. - We find such users in the library of Lucullus: Cato and Cicero as friends (Cic.Fin.
3.2.7), Greek scholars granted the run of the place (Plut.Luc. 42.1 2). - Each of the eight contains its own peculiarities, problems, and uncertainties. Space
precludes a full discussion here, although we will consider some of them, and the reader is
referred to Otranto for further discussion and earlier bibliography.
Papyrological Evidence for Book Collections and Libraries 237