For the most part, Roman books didn’t really resemble what you or I probably think of when we think ‘book’. The codex doesn't appear in the material record until the 1st century CE, and even then takes centuries to become the most common literary medium, so the transition is by no means an all-at-once conversion. Until later periods, book format in the ancient world is dominated by the papyrus bookroll. Although we do have some parchment rolls from antiquity, and plenty of papyrus codices, here we’re going to focus on the papyrus roll.

A basic bookroll consisted of a roll of papyrus (we’ll come back to this in a second) with rods called umbilici which were used to easily roll the ends of the papyrus without damaging it. To read a bookroll, then, the reader rolled in one direction and unrolled from the other, moving the sheet sideways as they went. The roll was oriented so that the umbilici and the rolled part were on the two sides (you could hold one end in each hand) when you were reading the writing upright (every pop culture doodle that shows the rolled part on top and bottom is incorrect on this point, as is the fact that most show the roll going in two different directions, so that what is on the inside on one end is turned out on the other, which would be absolutely terrible for preserving writing).

This rolled method makes much more sense if we think about how papyrus was made and sold. Sheets of papyrus weren’t sold individually, they were pasted together into longer rolls, with the kolleseis, or joinings, done so that writing rolled ‘downhill’ from one sheet to the next on the recto side, moving with the fibers. The opposite side was generally left blank, although we do see papyri with writing on the verso when sheets are re-used for other purposes, a bit like today writing something on the back of an envelope or a scrap of paper.

Oddly enough, it might help here (and I realize this comparison is more and more out of date by the day) to think of a VHS tape. Be kind, rewind? When you were done reading a bookroll, you might have had to re-roll over 20 meters of papyrus! (although most rolls are estimated to be more in the range of 3-15 meters, with the caveat that for various reasons - mostly the incomplete nature of our evidence and the challenges of working with fragments - our evidence on this may never be as complete as we'd like it to be). Bookrolls were meant to be stored with the front out, though, rewound back to the beginning of the VHS, as it were, and certain features made this the most desirable way to leave a roll. A blank sheet, called the protokollon, was attached to the beginning of the roll to protect the actual writing inside the roll and prevent fraying, which, obviously, would be on the inside of a roll that hadn’t been rewound. To protect papyrus rolls from damage, ancient sources also describe the use of covers (often as a way of signaling either or both a fancy book or a wealthy reader) or cases for rolls.

I’ve mentioned length of the papyrus roll, but there are also other dimensions to think of and variables to consider. Length of the roll of papyrus affected the eventual diameter of the rolled book, which means that if a smaller book was aesthetically desirable, a shorter papyrus roll had to be used. Book length could be extended by gluing in more sheets of papyrus or shortened by cutting off what wasn’t needed, so this length could be extremely variable. And while there was no completely standardized height, most rolls generally end up in a pretty predictable range determined by the usual method of papyrus production for a given period (Ptolemaic and Roman era rolls tend to have slightly different general roll heights, but the Ptolemaic and Roman eras also had different standards for what they considered the best quality papyrus, so this shouldn’t surprise us too much). This is actually quite helpful, because if we have a Roman era papyrus with, say, 26 cm surviving, and we have the top margin but not the bottom, we can still be sure we have almost the entire column, even if the bottom margin isn’t preserved, because most papyri of that period are roughly 30 cm high.

These rolls don’t survive to us in their entirety. They do survive in fragments and small pieces, in quite large numbers, and there’s a lot we can say about them from that. P.Oxy. 21.2288, for instance, is only about 10 cm high and less than 10 letters across but gives us possible evidence for ancient book repairs as well as a small chunk of Sappho, and that’s just one example of what we can learn even from very small fragments. Papyrus fragments as material objects that date to antiquity give us evidence for certain things that the indirect tradition (texts copied in Late Antique and then Medieval manuscripts that transmit ancient works to the modern world even when we have lost the physical ancient texts) cannot (things like lexical markings, scribal practices, education), just as the indirect tradition, which gives us complete literary works, can give us other kinds of material that would be incredibly difficult to glean from papyri alone.

So although there was no completely standardized bookroll form, and some aspects of formatting and the like varied from period to period, the general form of the bookroll remained stable until it was eventually completely supplanted by the codex. And while we don't have ancient bookrolls preserved in their entirety, we do have surviving fragments that we can use to reconstruct the roll's complete form and some related book practices.

​

Johnson, William A. Bookrolls and Scribes in Oxyrhynchus. University of Toronto Press, 2004.

_______________. “The Ancient Book.” In The Oxford Handbook of Papyrology, edited by Roger S. Bagnall. Oxford: Oxford University Press, 2011.

Reynolds, L.D. and N.G. Wilson. Scribes and Scholars. 3rd edition. Oxford, 1991.