by David Grant
Robert Graves’ richly embellished dramas of the first Roman emperor to be born outside of Italy, I Claudius and Claudius the God, penned through 1934-35, may in recent times have become more influential than Suetonius ‘dryly indiscriminate’ biography in filling the gap left by Tacitus’ lost account.166 This is another irony, for the full name of the English poet and novelist was Robert Von Ranke Graves; his mother was the great niece of Leopold von Ranke (1795-1886), the German historian who gathered up a wide swathe of source material and documentary facts in his attempt to show wie es eigentlich gewesen, ‘how things actually were’. Von Ranke, who was responsible for the progression of Quellenkritik, source criticism, produced the first ever historical journal in the process, Historisch-Politische Zeitschrift, and he reinforced the value of working with original documentary evidence.167 And if history is full of ironies, then the Muse Clio had a developed sense of the tragic too, for Suetonius, like many of those who preserved Alexander’s tale, failed to mention who his sources were, the primary material he himself may have creatively ‘upturned’ in Rome between his alleged intimacies with Vibia Sabina, wife of the Emperor Hadrian.168
As the Alexandrian scholars had noted, the expansion of the Graeco-Macedonian and Roman Empires homogenised linguistic identity, though at the expense of classical linguistic purity. The term ‘classical’ entered modern English in the 16th century and by 1870 Wilhelm Siegmund Teuffel (1820-1878) had developed his philological classifications, which included the so-called Golden and Silver Ages of Latin: 83 BCE-14 CE and the years 14-117 CE respectively. Like all ‘new’ linguistic ideas, these built upon earlier less systematic models incorporating faulty systems of categorisation based upon political events rather than pure prose style.169 In 1877 Charles Cruttwell refined the groupings by focusing on the progression of Roman literature through the republican and imperial periods in which it evolved from an ‘immaturity of art and language’, to ‘ill-disciplined imitation of Greek poetic models’, and finally to ‘clear and fluent strength’.170
The Silver Age (broadly) witnessed the end of the true Roman Republic so that Diodorus, Trogus, Curtius, Plutarch and Arrian were publishing under successive Roman dictators and emperors. Despite that ‘clear and fluent strength’, Teuffel gave a scathing opinion on the loss of free speech through much of this period, particularly in the Julio-Claudian era (Augustus to Nero, 27 BCE-68 CE); it was a period of ‘continued apprehension’ when natural composition was subordinated to the desire to appear ‘brilliant’, ‘hence it was dressed up with abundant tinsel of epigrams, rhetorical figures and poetical terms’. ‘Mannerisms supplanted style, and bombastic pathos took the place of quiet power.’171 Freedom of expression was apparently not the cure, for Teuffel credited the literature of the ‘happier’ 2nd century that followed as nothing less than a scandalous ‘imitation’.
Able to appreciate the textual markers defined by these eras, modern scholars developed the tools of linguistic ‘archaeology’ to help decode the past and even gene-tag individual historians. Narratologists note that the classical writer had a literary pace, gait, or shuffle. In 1889 Dessau concluded the Scriptores Historiae Augustae were written in the time of Theodosius rather than Diocletian or Constantine (as the manuscripts claimed), noting ‘the uniformity in phraseology and stylistic devices’ which pointed to a single author, not six.172 We can similarly identify where texts noticeably stride ahead or stumble from that uniformity, revealing what may be a historian’s non-seamless switches between sources. When analysing texts dealing with Alexander, for example, Hammond pointed to Plutarch’s change from ‘florid’ structures to ‘restrained and artistic’ prose and especially to clues in his rhythm, identifying ‘iambic and spondaic runs and occasional tribrachs’.173 He also observed Arrian’s switches from narrative tenses when a new source was introduced, whereas Pearson saw ‘Asianic rhythms’ that betrayed the presence of genuine fragments of Cleitarchus.174
Sir Ronald Syme’s linguistic studies concluded Curtius’ style was ‘sub-Livian and pre-Tacitan’, whilst elements within the Metz Epitome place its authorship in the 4th or 5th century CE.175 Only Curtius and the Metz Epitome employed the word testudo (tortoise) to describe the Greek shield formation.176 The term didn’t exist in Alexander’s day or in Greek military history (aside from a mention of something similar in Xenophon’s Hellenika), and so Curtius was delving into his own Latin vocabulary to describe an earlier, but now familiar, shield-locking tactic.177 Using similar observations, linguists determined Carlo Sigonio’s fragment of Cicero’s De Consolatione was a fake, as it employed terminology Cicero could not yet have himself been familiar with. Linguistic progression remains just as visible today; anyone reading the biographies of Alexander by Wheeler or Mahaffy cannot help chewing with difficulty on the prose laid down just a century ago, and Rooke’s 1814 second volume of Arrian’s Anabasis has already been described as ‘archaic’.178
The classical Latin once spoken by the ‘good’ and noble families, the sermo familiaris and sermo nobilis, disintegrated to Vulgar Latin and then into the early Romance languages as the Roman Empire expanded and fragmented. As with Hellenistic Greek koine, sermo vulgaris was first spoken by soldiers, dispossessed townsfolk and slaves across the empire, resulting in marked differences between the spoken and written forms. Letters took on new sounds through palatisation, a process that produced the phonetic splits we read and hear in the romance languages of today. The loss of nasal inflection resulted in the dissimilation of voiceless consonants, and thus Platon became Plato.
The Greek ‘K’, which had become a ‘C’ in Latin though still pronounced as kappa, graduated to a soft ‘S’ before the vowels ‘I’ and ‘E’. Cicero, originally pronounced ‘Kikero’ (we believe), and Caesar, pronounced ‘Kaisar’, took on the pronunciations we are more familiar with in English. In Hellenistic Greece, ‘Y’ and ‘OI’ were often written as a ‘U’, and later the distinction in vowel length was lost as part of a wider monopthongisation. Double consonants were reduced to single and aspirated voiceless stops were changed to fricatives by the 4th century. By the Byzantine period, sounds and their corresponding letters were further simplified when, for example, ‘H’ and ‘EI’ became ‘I’. These processes played their part in textual infidelity as new translations were made.
The early Latin-from-Greek manuscripts, perhaps already incorporating corruptions from these linguistic challenges, went through further transmission processes as Western scripts developed; the latitude for confusion widened still further when Latin texts were translated into the ‘modern’ languages that were not rooted in Latin. When use of the capital and uncial scripts of the 4th and 5th centuries came to an end with what Petrarch termed another ‘Dark Age’, Christian copying of pagan literature halted for almost three centuries, so few manuscripts of this period survived. The so-called Carolingian Renaissance (late 8th to 9th century) saw a resurgence when monks once more started copying the oldest texts available to them into Carolingian miniscule. This lower case script, developed under the patronage of Charlemagne (never fully literate himself), had the advantage of spaces and punctuation, so liberating Latin from the confusing continuous flow of capitals in uncial texts.
New Latin transcriptions made during the 11th, 12th and 13th centuries attempted to restore some sense and order to already corrupted passages, but this, more often than not, further damaged the original prose. The aforementioned defective knowledge of classical Latin and Greek especially, and hence an ignorance of how to separate continuous script, led to nominative cases being confused with genitive and dative cases, and singulars with plurals, especially where the letters a, e, and diphthongs were concerned.179 Add to this the mistaking of proper names with verbs, and the faulty extraction of numbers from uncial text, and we have some idea of the task facing scribes. Mistakes may have been innocent and well-intended but ‘they were rich in the germs of future corruptions’,180 unlike the deliberate tampering of the later 14th and 15th centuries when rel
igion held a heavy hammer over what survived and in what form.
A fragment of the Codex Sinaiticus (John chapter 21, verses 1-25) written in continuous Greek uncial script and dating to the 4th century CE.
As the works that had not been translated into Latin were in less demand in the Middle Ages, the Greek texts of Herodotus and Thucydides were largely bypassed. The case for the resuscitation of Herodotus was further challenged by the criticism heaped upon him in the texts that had been translated. As a result, it is estimated that there are just eight surviving manuscripts of Herodotus’ Histories, whilst the extant manuscripts of Thucydides only date to some 1,300 years post his original. It was not until 1448-1452 that Valla produced a Latin edition of both the ‘awe-inspiring’ Thucydides, and the ‘radiant’ Herodotus (1452 through 1457) who was finally exhumed from his classical slashing with Stephanus’ Apologia pro Herodotus published in 1566.181 It was not an easy task, as Valla explained in his preface dedicated to Pope Nicholas V:
… as everybody admits, Thucydides is steep and rocky, especially in the speeches, in which his books abound. This is clear from what Cicero, whom men of his time called ‘the Greek’, says in his de Oratore: Those speeches have many over-subtle thoughts that can scarcely be understood.182
Valla added: ‘There you see, highest Pope, what Thucydides is like in Greek, and if you decide that in my translation he keeps this same dignity, I shall be oblivious of my labour.’183
But the labours were often divisive. A comparison of two manuscripts containing Livy’s Ad urbe condita libre – the 5th century Codex Puteanus and the 9th century copy known as the Codex Reginensis 762 – further illustrates scriptorium problems.184 The Reginensis transcription was undertaken by eight separate scribes, each given a different section to work on. We even know the copyists’ names as their signatures appear at the end of each quaternion, the manuscript unit of four double leaves. They were also of more than one nationality as evidenced by the repeated nature of the blunders they make and their non-identical ink types. Earlier divergent processes had already been at work on Livy; all manuscripts of the first ten books stem from the single recension commissioned by Quintus Aurelius Symmachus, a Roman consul, in 391 CE. Moreover, the emendation of the editor would have produced this version by selecting what was considered the best of the then extant earlier editions. Epigraphists have detected the hands, once again, of a number of scribes in the recensions that followed, one branch of which is known as the Nichomachean after the named subscribers.185
Fortunately, it is possible to determine if early book sections have been pieced together by different copyists, and whether pages have been lost, by studying the parchment on consecutive sheets, as the hair side of the skin is distinct from the flesh side and the double folding of the quaternion ought to produce matching textures. Von Tischendorf employed the techniques of codicology to highlight such discrepancies when describing the Codex Sinaiticus, believing that four separate scribes worked on the manuscript, with five ‘correctors’ amending the text at different times.186 Recent paleographical studies of the codex do indeed confirm numerous scribes were employed.
The multiple-copyist methodology was, nevertheless, typical when a work encapsulating centuries of history had been either originally released by the author in ‘packets’ (as in the case of Livy’s chapters) or had been packeted in later editions, when they were often separated into decade or pentad divisions that might be a wholly contrived new arrangement. But, understandably, Livy’s 142 books were a challenge to read through without some kind of break; moreover, he had drifted into loquaciousness and repetitions, which, as Horace reminded us, ‘are not well received’.187 Barthold Georg Niebuhr (1776-1831), who finally gave us the Römische Geschichte, an ancient Rome that linked the fate of patricians and plebeians together,188 supposed ‘the declining quality of Livy’s later books was the cause of their loss; they were considered less worthy of copying’.189 In Livy’s defence, the ‘fountain that never trickles’ had himself described the ‘deep water’ and ‘vast depths’ he was wading into from an inundation of recent detail, versus the ‘shallow waters’ of the earlier history of the city.190 The result: only thirty-five books remain in reasonably complete form.
Supervisors or ‘correctors’ were on hand in the scriptorium to immediately edit the quaternions as they were completed, but this was an ink-upon-ink process and the inevitable dittographies, haplographies and parablepses, the scribal errors of repetition and omission, crept in.191 If the corruptions manifested themselves in clear conflicts between what should have been identical manuscripts, we are least alerted, but when multiple texts are divided, we are on softer ground. For example, in Plutarch’s Life of Demosthenes we cannot be sure if it is Theopompus, or Theophrastus, who refused to incriminate the orator, as various manuscripts cite both.192 One scholar tackling the dilemma suggested a ‘meddlesome’ scribe made the error; he is no doubt correct, but which name should we amend?193 Oratorical imitation was part of the classical education syllabus, and so the orators were vulnerable to having the speeches debased. We have six surviving letters supposedly written by Demosthenes; new analyses of an old debate suggest four are genuine and two are fakes, whilst roughly one-third of his extant speeches appear to be less than genuine too.194
Another ongoing debate is the relationship between Photius’ references to the Ten Attic Orators in his Myriobiblon, and the ten Lives preserved in Pseudo-Plutarch. It has become unclear whether Photius used Pseudo-Plutarch directly, or an underlying source used by Pseudo-Plutarch, or whether Photius added material of his own or detail from a compilation that might have included either of the above. Although a study by Schamp targets an early 4th century common source, we don’t know if the two sets of Lives developed from second or even third-generation transcripts.195 But if we pay similar attention to Photius’ own introduction to his collection of epitomes in the form of a letter to his brother, we note the admission: ‘we engaged a secretary and set down all the summaries we could.’196 The term ‘we’ recurs throughout, not ‘I’, leaving us to wonder how much of the epitomised work was Photius’ own, and how much was compressed by the anonymous secretarial hand, or in fact summarised by members of the reading circle he ran.197
Besides scriptorium transmission errors, further loss of chronological precision resulted from the fluidity of the calendar year (in Greece especially), the loss of intercalary months (inserted to realign calendars with moon phases), along with the awkward synchronisation of archon years with consul elections; all troublesome detail for copyists trying to make sense of the order of events.198 Additional slippage arose from the word-denoted and numeral-based counting systems of Greece and Rome, which lacked the numeric dexterity to efficiently deal with large numbers, fractions and percentages. As a result these were awkwardly presented as long sequences of letters and acrophonies, and this must partly explain the disparity in the treasury figures, and troop and casualty numbers we encounter in the battles of Alexander and his successors.
The ‘plus’, ‘minus’ and ‘equals’ symbols were still a millennium-and-a-half away, and the concept of ‘zero’ as a placeholder still eluded mathematicians. It was not until ca. 250 CE (an uncertain and debated date) in Alexandria that Diophantus, ‘the father of algebra’, and his Arithmetica, gave us an abbreviation for (and for powers of) the unknowns and a shorter means of expressing equations. But the use of true numbers in maths with succinct algebraic notation and symbolism – instead of long sentences of letters and words that Diophantus did still employ – made its way west later (in the 8th and 9th centuries CE as a refinement of the Hindu-Arabic system), and only then did it establish a useful and easily transmitted system to deal with and reproduce complex numbers
So, for a host of reasons, manuscripts were, as Digory Whear observed in 1623, ‘interpolated by the hands of smatterers, and most basely handled’.199 Petrarch bemoaned: ‘What would Cicero, or Livy, or the other great men of the past, Pliny above all, think i
f they could return to life and read their own works?’200 Here Petrarch was referring to the sad condition of the ‘treasure house’ of Pliny’s Natural History that he had purchased in Mantua in 1350; some seven intact manuscripts survived, though none of them are earlier than 850 CE (older examples were either incomplete or survived as palimpsests).
But not everyone has shared his enthusiasm for poring over the past. In a posthumously written historical review of 1847, the English archaeologist, Sir William Gell, commented on the tedious job of copying newly found manuscripts and the Herculaneum Scrolls: ‘If Omar, according to the tale, burned the library of Alexandria, we have doubts whether he ought not to be honoured as a benefactor of our race.’ Gell followed with: ‘The palimpsests so laboriously deciphered, have given us scarcely anything that is either of interest or value.’201 And for all his efforts in educating scholars to adopt the opposite view, when Petrarch died, as his Will made no reference to his library of manuscripts, the collection was subsequently seized by the Lords of Padua and scattered across Europe.202