Hi, I'm Peter Adamson and you're listening to the History of Philosophy podcast, brought to you with the support of the Philosophy Department at King's College London and the LMU in Munich, online at historyoffilosophy.net. Today's episode, Made by Hand – Byzantine Manuscripts. Walk into any decent bookshop in the English-speaking world and you'll find Plato's dialogues on the shelves. Maybe even a collection of all his dialogues, which takes up more than one and a half thousand pages, even if you exclude the works that were not really written by Plato but only transmitted under his name. Strictly speaking, of course, Plato didn't write the rest of the book either for the simple reason that he didn't know English. Happily, you can also get Plato in Ancient Greek, for instance in the Oxford Classical Texts series, where it takes up five volumes. But here's the thing. Plato didn't write those Greek texts either, at least not exactly. Most readers, even most professional historians of philosophy, don't give this much thought and proceed as if the printed version was ordered straight from Plato's Academy. In fact though, the modern edition is simply scholars' best guess at what he may originally have written. You'll be relieved to learn that the text of Plato's dialogues is actually relatively secure, but like all other ancient works, they are moving targets and will never be fully established beyond all doubt. The individual words and sentences in that edition have in some cases been a matter of intense philological debate. Nor are modern editors hiding this fact. At the bottom of each page, they've supplied a dense collection of footnotes bristling with Latin abbreviations and bits of Greek, which are alternative versions, called variants, of what Plato may have written. To understand what lies behind those footnotes is to appreciate more fully the astounding fact that we are able to read Plato at all, never mind one and a half thousand pages worth of him, a good two and a half millennia after he lived. It means tracing the long and hazardous journey those writings travelled, surviving more or less intact as empires rose and fell, as Attic Greek fell into disuse as a language of everyday speech, and as philosophical tastes changed. That journey went straight through Byzantium. Without the efforts of Byzantine scholars and scribes, we would not be able to read Plato today. Ancient philosophy, indeed ancient literature as a whole, would barely exist anymore. To see why, let's go back to the beginning of the journey and talk about how writing would have been set down and preserved in Plato's day. There were several options. Many inscriptions in stone survive from antiquity, including at least one philosophical text, a statement of Epicurean doctrine by the Roman era philosopher Diogenes of Oneida. There are also shards of clay pottery with writing on them. Such a shard is called an ostracon, and there was a legal procedure in Athens where political figures could be exiled or ostracized by writing their names on such shards, and some of these survive. We weren't going to write Plato's Republic as a stone inscription or on bits of pottery. For that, the first stage would probably have been inscribing the text onto wax tablets. Hence an ancient report tells us that Plato's final work, the Laws, had just been recorded in wax when he died. Longer term, though, text would be set down on papyrus. This is made from the leaves of a plant that grows especially in Egypt. It came in various grades of quality, with the poorest grade even being used as a packing material. Typically, it would be fashioned into a long scroll which would be gradually unrolled while reading. The text had to be re-rolled after use, much like a cassette tape, if you're old enough to remember those. Unlike a cassette tape, information was usually put only on one side of the papyrus roll. The text would be arranged in columns, written entirely in what we might think of as capital letters, and without punctuation. Learning to read was thus in large part learning to see where one word or sentence stopped and another began. Furthermore, there were none of the accents and diacritical marks you'll see on Greek in modern editions, which were invented later. These don't make reading significantly easier, but do sometimes resolve possible ambiguities, such as the one exploited in a clever remark of Heraclitus, the bow, its name is life, its work is death. The point of this is that the same Greek word can be accented in two different ways, one of which means bow, bios, the other life, bios. By the time we get to Byzantium, all of this will be different, apart from the fact that all documents will still have to be written out by hand, this being the meaning of the word manuscript. The first major change is the introduction of parchment, a word that derives from the name of the Greek city, Pergamon, where it was supposedly invented. Parchment is leather that has been carefully treated to make it as smooth and light in color as possible. If you look closely at some texts written on parchment, you can tell what they are made of, because the surface where the hide was will be lighter in color than the inner side. Parchment came into use already around 200 BC, but did not immediately displace the use of papyrus. Once it did become the dominant material for writing, it clung on tenaciously. Well after the introduction of paper, even into the Renaissance, parchment was still used for particularly elegant or important book production. Among the Byzantines, court documents were occasionally made from parchment that was dyed in imperial purple. As for paper, its introduction was the most important change in the technology of writing between the invention of the alphabet and of the printing press. Paper is typically made from used fibrous material, like linen or hemp. It can readily be made from used rags. The technology came into the Islamic world from China in the 7th century and was taken up in Byzantium in the 9th. Faithful listeners of the podcast will notice that in each case, paper's arrival was followed about a century later by an explosion of intellectual activity. This is not a coincidence. Without paper, we would not have had al-Kindi in the massive Greek-Arabic translation movement in the Islamic world, or Fotius and Pselos in the Byzantine world. Since it was scholars like them who led the effort to study and preserve ancient Greek literature, you can thank the Chinese inventors of paper for the fact that you can still read Plato. The reason paper made such a difference is that it is much cheaper than parchment and much more durable and readily available than papyrus. Actually, papyrus had already been in short supply once its main source, Egypt, fell under the sway of Islam. In the late 7th century, a Muslim ruler there even placed an interdict on the export of papyrus. This obviously encouraged the use of parchment in the Christian world despite its costliness. We can see how valuable parchment was from the fact that, sometimes, the ink would be removed from it to create a more or less clean writing surface on which a new text could be set down. This is called a palimpsest from a Greek word meaning scraped off. Otherwise lost works have been discovered in the undertext of palimpsest manuscripts. A sensational example is a work by Archimedes that was found on parchment written underneath a religious work from Byzantium and was rendered readable through the use of x-rays and other technology. In the philosophy world, there was recently similar excitement at the discovery in a palimpsest of an unknown commentary on Aristotle's categories. There were two other big changes between classical antiquity and the Byzantine period. Unlike Plato, Michael Pselos would not have had to work his way through a scroll while reading. At that time, the standard format was the Codex, which is basically like a modern book with pages made from folded paper or parchment bound between covers. The Codex begins to appear in about the 3rd century AD, especially with legal texts having evolved from the practice of tying together leaves of parchment to make a sort of notebook. Codices begin to outnumber scrolls at about 400 AD, but do not displace them completely until the 7th century or so. A couple of centuries later, the scribes introduce a second innovation. They begin writing in so-called minuscule texts, which you can roughly think of as lowercase letters instead of the old maguscule uppercase letters. But the use of maguscule for more formal texts persists for a good while. Our oldest surviving minuscule text is from the year 835, but still in the 11th century, maguscule is still found in liturgical manuscripts. And finally we have the introduction of paper about a century after scribes started writing everything in the new minuscule script. All of this might lead us to expect that a pagan work like a treatise by Aristotle would have initially existed on papyrus scrolls written in maguscule, one or more of which would be copied in the same script onto a papyrus codex, then into a parchment codex, then into a minuscule text but also in parchment, before finally being copied in the sort of format that usually survives today, a paper codex with minuscule script. While this is the right sequence in technological terms, a given work might not have existed in every form I've just listed. In particular, during the so-called Dark Ages of Byzantium, few pagan works were copied. Once interest in them reawakened, the scribes would have had to use texts in long outmoded formats as their basis. Thus a 9th century paper minuscule copy might be based directly on maguscule parchment from the 6th century. So it's a good thing that parchment is such a durable material. Books were copied and kept in a number of different contexts. The first thing that leaps to mind would be major institutions like the famous Library at Alexandria. It was important not only because of the sheer quantity of literature it held, but also because scholars working there produced the editions that usually lie behind later Byzantine copies. Generally speaking, when modern day philologists try to establish the text of a classical author, like Homer or Plato, they are really trying to get as close as possible to the Alexandrian edition of late antiquity, since it isn't possible to go back further than that. By Xantium, monasteries were less important centers of text production than in the Latin West with a more significant role played by royal and patriarchal libraries. But we should not underestimate the role of smaller schools and private libraries. Institutions such as the ancient libraries at Alexandria and Pergamon, or the collection of books we assume existed at a philosophical institute set up by the Byzantine Caesar Bardas, were threatened by mass destruction in times of political upheaval, whereas a private library might survive. The very preciousness of books also exposed them to threat. We've already seen how parchment was reused in palimpsests, and books might also be sold off to raise funds. A 12th century Archbishop of Byzantium complained about illiterate monks who did this, writing, "...just because you have no trace of culture, must you empty the library of the books that transmit it?" The upshot is that, as Richard Goulet has observed, ancient philosophy is not really preserved down to the present day, but rather transmitted. We have almost no physical texts by philosophers from the classical period, with a few exceptions like that Epicurean inscription in stone, or the private collection of rolls owned by another Epicurean, Philodemus, which was preserved thanks to a volcanic eruption. Texts wore out, were discarded, or lost in fires. Already in late antiquity, scholars were conscious of this. Simplicius copied out quotations from pre-Socratic philosophy when commenting on Aristotle because he knew that readers might not otherwise have access to these texts. Demistius, writing in the 4th century on the occasion of the founding of an imperial library, mentioned that some authors, like the early Stoics, were already in danger of becoming unavailable. Texts were often lost in times of transition for book technology. When codices replaced rolls and minuscule script replaced maguscule, the priorities of the scholars of the time dictated what would be copied into the new format and survive through the change. As a result, the surviving corpus of ancient philosophy is basically what a Neoplatonically inclined Byzantine scholar Leipselos would think worth preserving. So we have thousands of pages of commentaries on Aristotle and the entire output of Plato and Plotinus, but not a single work by those early Stoics who Demistius was already fretting over. The same goes for other fields of ancient literature. We have more than 230 manuscripts of Aristophanes because Byzantine philologists valued him as an exemplar of fine Attic Greek. And the single Greek author for whom the most text survives is Galen because his writings formed the basis for the study of medicine. The quantity of surviving manuscripts for any given author is a fairly reliable indicator of how interested the Byzantines and Renaissance humanists found that author, and also of whether the author was used in teaching contexts. Thus, we have a good 260 manuscripts containing Platonic dialogues, but more than a thousand for Aristotle. His works are also very unequally represented in the manuscripts, with a vast number of copies of his logical treatises, because they were regularly used in the classroom, but comparatively few for works like the Poetics. Even within the logical corpus, the introductory works, categories, and on interpretation greatly outnumber, for instance, the Sophistical Refutations. With the exception of one fragment from that work, all our Aristotelian manuscripts are in various miniscule scripts. Given that miniscule started to be used in the 9th century, this means that our textual evidence for Aristotle begins about one and a half millennia after his death, so it's hardly a surprise that many of his books are lost. We know this because we have ancient lists of his works with our surviving corpus representing only a fraction of what is listed. And his surviving works are not what would have been read by his students at the Lyceum. Almost all Aristotle goes back to an edition of his works produced in the Roman period by the scholar Andronicus. Some treatises by Aristotle, notably the Metaphysics, were only compiled as single works at this editorial stage. But the target of a modern editor of Aristotle is not really even Andronicus's edition. It is rather the lost copies in Maguscule's script that were the basis for the surviving Byzantine manuscripts that are in miniscule. Even if we have a large number of manuscripts for a given work, they will often all go back to one single copy in Maguscule. That copy was transliterated into miniscule in, say, the 9th century, and then discarded. All further copies, which could number in the dozens or hundreds, would thus go back to the initial transcription. If you're asking yourself, who cares, then you haven't thought about the challenge of copying out an entire book by hand. This was of course tiresome work as shown by the prayers that scribes often insert at the end of a copy, giving thanks that their labor is completed. But more to the point, it is effectively impossible to copy out a text of any length without errors. Even if you were copying from a modern, printed text with spaces between letters and punctuation, you'd make mistakes. But imagine copying from a handwritten text like those papyrus rolls or ancient codices with columns of unbroken maguscule. Certain kinds of slip happen quite often. The scribe might skip a line or miss out a phrase because the phrase that follows begins with the same letters. Glancing back and forth between his source copy and his new copy, he jumps from one line or phrase to the next. Or one letter may be misread as another, especially if the letters look similar, as do the Greek letters l and a in maguscule script. Another problem is that manuscripts often have notes in the margin. This was routine practice and anticipated in manuscript production. Manuscripts routinely left wide margins so that they themselves or their successors could make notes. Some manuscripts even leave gaps in the main text, for instance in a historical chronicle so that an uncertain date can be filled in later. Or notes might be made between the lines of the text, for instance by writing a correction above a word or phrase. All of this apparatus might or might not be retained in a later copy. Sometimes the marginal glosses might be copied over as if they were part of the original text. If you're now trying to produce an edition of that original text, you want to eliminate such extraneous material that has crept in while also correcting other errors. That's how you get as close as possible to knowing what the ancient authors actually wrote. This takes us back to those footnotes in a modern-day edition of Plato or Aristotle. The notes record the different versions in the Greek text found in various manuscripts – not all of the variants, but only those the modern editor deems significant. As I said, in the main text above those notes, you're seeing the decision of the editor about which variants to accept, or in some cases Greek that has been hypothesized by the editor as an improvement on what we find in the manuscripts. How does one go about doing this? Let's take as an example a short work by Aristotle called On the Motion of Animals, which was recently re-edited by my colleague here in Munich, Oliver Primavesi. This is a case where all existing copies do indeed go back to a unique manuscript that was written in Maguscule. We know this because there are some mistakes in the Greek found in all copies. This means they must all derive from one single copy that already had those mistakes in it. This gives us an insight into a more general point, one that is key to the task of editing Greek manuscripts. In the first instance, you look not for correct Greek, but for errors. Imagine you have four manuscripts numbered 1 through 4. If manuscripts 1 and 2 share the same mistake, whereas copies 3 and 4 preserve a correct reading, then you know that 1 and 2 must be copies of the same manuscript, or perhaps 2 is a copy of 1, or vice versa, while 3 and 4 are based on some other manuscript that was free of this error. On this principle, it's possible to arrange all the existing manuscripts in a branching diagram that shows which were copied from which. This diagram is called a stemma codicum. Primavesi's new edition of On the Motion of Animals was needed because of a major revision he realized was needed in the stemma. A comparatively late manuscript from the 15th century turned out to have correct readings that were not found in any other copies. Previous editors had focused on earlier textual witnesses that could take them back to a version from the 10th century. This seems to make sense. All else being equal, an older manuscript should get you closer to the original. But as philologists like to say, because philologists enjoy speaking in Latin, recentiores non diteriores, the more recent are not necessarily worse. In this case, a manuscript that is a good 500 years newer than others contain readings that helped Primavesi to get a more accurate text. His edition has 120 changes over earlier ones and that in a text of relatively few pages. Remarkably, many of the improved readings matched the medieval Latin translation by Thomas Aquinas's colleague William of Murabeka, who evidently had access to the same line of transmission exploited by Primavesi. It's worth bearing in mind that with this meticulous labor, the modern textual editor is just doing what ancient medieval and renaissance scholars already did in their own time. They too compared manuscripts to eliminate errors. Thus, late antique and Byzantine commentators on Aristotle sometimes record variants found in other texts. The fact that they were so conscientious actually complicates the business of figuring out which manuscripts are copied from which. The medieval editors might look at additional manuscripts to fix errors in the one they are copying. When modern scholars do this, it's just good scholarship. When older ones do it, it's rudely called contamination between lines of textual transmission. Then too, like modern editors, the Byzantine scribes might correct the text on their own initiative so that it will make more sense. Hence another principle followed by philologists, which of course goes by a Latin name, lectio difficulior, or more difficult reading. This means that if you have two variants, one of which is somehow stranger, though still grammatically possible like a very unusual word, while the other is familiar and straightforward, you should suspect that the easier version could actually be a scribe's conscious or unconscious correction and consider rejecting it on that basis. It should now be clear that we owe a lot to the Byzantine philologists. One of their greatest legacies is the so-called philosophical collection, a group of 17 manuscripts on philosophical and scientific topics that go back to a multi-volume edition produced in Constantinople. It was probably compiled from material gathered by Platonist philosophers in late ancient Alexandria and Athens, maybe for use at the aforementioned institute established by Bardas. The philosophical collection is a treasure trove, which includes among other things Plato's dialogues, works by Neoplatonists like Proclus, Simplicius, and Philoponus, treatises of middle Platonists and by Aristotle and his followers. From the handwriting we can tell that it was copied out by no fewer than eight different scribes who worked closely together. The manuscripts from the collection are now scattered across a number of European libraries, which is typical. For these and other philosophical texts to survive, it was of course necessary not only that the Byzantine scribes copied them out, but that those Byzantine copies themselves survived. Some Greek manuscripts are still to be found in Istanbul. The Ottomans knew Greek manuscripts were valuable and did not just destroy them. But mostly Greek literature exists today because it was spirited away from Constantinople to the Latin West after the capital fell to the Crusaders in 1204, or copied by Western scholars who took advantage of the situation to visit Constantinople and make copies of the manuscripts found there. One fairly immediate result was the sudden rediscovery of Aristotle's works by the Scholastics, whose far-reaching consequences for medieval philosophy we've already explored here on the podcast. We saw that in that context, a surge of interest in classical pagan thought provoked significant opposition, and something similar had already happened in Byzantium. The enthusiasm for Platonism embodied in the manuscripts of the philosophical collection was not shared by everyone, as we'll be seeing in an upcoming episode. But before we get to that, I want to exploit an easily available resource of my own. We've just been discussing the philological efforts of Oliver Primavesi, who dedicates himself to producing improved editions of Aristotle. As it happens, he does that work about 10 seconds walk away from my own office because he's my colleague at the Munich School of Ancient Philosophy. So join me next time to hear more about Byzantine manuscripts from someone who works with the them every day in an interview that will focus on the work he's been doing on Aristotle's metaphysics. That's right here on The History of Philosophy Without Any Gaps.