Winged Words: Stenographers at Work

In a letter to a new stenographer, Basil of Caesarea admonished him to take care to form his characters well and to punctuate the text correctly.¹ Like other fathers of the church, Basil availed himself of the service of stenographers when dictating letters and treatises and for recording sermons.² While in the Roman world stenographers had been employed since the Republic, they were not equally popular in the Greek world until late antiquity.

The literary evidence from a previous period provides some information, as itinerant sophists sometimes employed stenographers to record extemporaneous speeches. In the Life of Apollonius of Tyana (1.18), Apollonius departs from Antioch with two attendants from his father’s house. One was a stenographer and the other a calligrapher, a scribe who was supposed to turn the stenographer’s notes into a fair copy. Philostratus also mentions stenographers in connection with the munificent Herodes Atticus, who gave the sophist Alexander rich gifts: “ten pack-animals, ten horses, ten cup-bearers, ten shorthand writers, twenty talents of gold, a great quantity of silver, and two lisping children from the deme Collytus, since he was told that Alexander liked to hear children’s voices.”³ However, the veracity of this striking passage is doubtful, because sources for Philostratus’s sophists are mostly oral.⁴ Lucian also alludes to shorthand writers recording a speech “so that nothing would be lost,” but the attribution to him of the work containing this statement is dubious.⁵

Shorthand is a system of accelerated writing used to transcribe the spoken word, a technique that has been used across past centuries and in different cultures.⁶ In contrast to abbreviations, in which some or even most of the letters of a word are omitted,⁷ shorthand uses signs to reproduce the whole text, that is, strokes of the pen and symbols instead of letters. The system was invented during the last decades of the Roman Republic or the early years of the empire.⁸ Modern terminology used for such fast writing includes Schnellschrift, Tachygraphie, and Kurzschrift in German, and tachygraphy, shorthand, and stenography in English. In the second century stenographers’ services were in sporadic demand. Galen was thought to frequently use stenographers in his practice,⁹ but more recent research found that he did not utilize the technique often, perhaps because it did not give him any personal advantage.¹⁰ Yet friends and acquaintances unable to attend some of his medical demonstrations would send their own tachygraphers to take notes, and sometimes after giving a demonstration he would need to repeat his text for them.¹¹ On one occasion, after Galen had addressed an issue, his friend Theuthras asked him to repeat the lecture to a slave whom he would send to copy it down.¹² Another friend, the consul Boethus, sent him some stenographers to whom he could dictate the text of a lecture.¹³ In these cases, unusual situations forced the doctor to repeat what he had just said, but only in the last circumstance did he dictate merely a summary to the tachygraphers.

It was in late antiquity that stenographers became numerous and indispensable in producing records of church conferences and in preserving homilies.¹⁴ In fifth-century Arles they even acquired their own saint and a martyr, Saint Genesius.¹⁵ The increasing popularity of stenographers in late antiquity is best explained by the common habits of writing and dictation.¹⁶ In late antiquity, when dictation was usually preferred to personal writing, it is referenced in the surviving literature as dictare, or “to dictate.” Yet dictare could also mean “to say aloud and to dictate something to someone” or even “to compose something, while saying it aloud at the same time,”¹⁷ reinforcing the power of scribes and stenographers.¹⁸ The papyri also indicate that in late antiquity, professional hands were preferred, as letters became more elaborate and full of rhetorical expressions.¹⁹ We may imagine that at that time, the cacophony of voices dictating texts and letters, engaging in debates, and preaching in church became staggering.

A text taken down by a stenographer would then need to be transcribed into Greek or Latin for Christian figures like Basil of Caesarea and Jerome who often used stenographers.²⁰ In his work On the Trinity, Saint Augustine appears to use the verb conscribere to refer to this process when mentioning that he gave a sermon in church that was then written down,²¹ while in Ep. 223, Basil uses the term metagrapsai. Though it is easy to assume that a different scribe would “translate” the text and rewrite it in elegant handwriting, the ability to read stenographic signs was actually part of a specialized education with which most people were not familiar.²² The sources point to a double process: the text was first turned into Greek or Latin characters by someone with stenographic training, and then a specialized scribe might copy it into formal handwriting. At a previous stage the stenographer may have produced a copy to later be emended by the author or constitute the Gesta of a conference. Specifically, Jerome has described the process using the appropriate Latin terminology: a stenographer would take down (excipere) the text using a pen (stilus) on waxed tablets. He would then “translate” it, producing schedulae. The author would emend it, reread it, and finally hand it over to a librarius, who would copy it formally. A stenographer who was supposed to read from his text during a conference would have to do so from the “translation” of his work rather than directly from tablets, a procedure that was protested during the Conference of Carthage.²³

Did educated people and scholars know tachygraphical signs? This is an important question concerning the likelihood that Arrian used stenography himself. However, the evidence from the papyri is basically nonexistent. Only a text of Plato contains some signs that were regarded as tachygraphical but are unknown and perhaps were invented by the scholar who added substantive Greek notes to the text in a fast, small, and practiced hand. This papyrus from the second century, P.Oxy. XV 1808, contains an excerpt from Book 8 of the Republic, which includes the famous passage (546 b–c) on the “Nuptial” Number.²⁴ It is uncertain whether the second writer was transcribing from another book roll or taking notes during a philosophical lecture, but the speed and apparent struggle of his note-taking renders the latter scenario more likely.²⁵ The writer also employed an abundance of abbreviations of the kind typical of commentaries. No other papyri annotated by scholars or students exhibit this feature.²⁶

Education and Social Position of Stenographers

Competence in shorthand depended on the flawless memorization of simple signs according to a syllabary, followed by whole words and phrases and inflectional endings.²⁷ School training culminated in learning the so-called Commentary, whose texts provided a list of over eight hundred stenographic signs that referred to Greek words.²⁸ Learning the whole system required years of training and commitment.²⁹ According to this technique, words were rendered through signs that represented each syllable, with the result that less memorization became necessary.³⁰

Concerning pedagogical issues, little is known about how stenographers learned their skills and training, or what the monetary compensation of scribes was for both documentary and calligraphic texts. Two papyri from the second century, P.Oxy. IV 724 and P.Oxy. XLI 2988, mention learning stenography; both are contracts of apprenticeship.³¹ The first, the more complete of the two, refers to a boy apprenticed for two years to a shorthand writer (σημειογράφος). The father paid the teacher a sum of 120 drachmas in three installments; the final one would be due at the time “when the boy would be able to write from every kind of everyday language and to read (back the symbols) faultlessly.”³² It seems, therefore, that by the second century stenographers were active in the Roman East. It is difficult to know how extensive the teaching in these cases was, including whether it truly covered the entire system of tachygraphy.³³

Naphtali Lewis remarked that the sum of 120 drachmas given to the teacher was noteworthy, because in the second century that could serve to buy a small house.³⁴ Other scholars before Lewis tried to interpret these papyri by comparing the figures with Diocletian’s Edict of Maximum Prices, issued in the year 303. And yet, according to the Edict, a teacher of stenography in the time of Diocletian would have earned much less than his counterpart in the second century.³⁵ It is necessary, though, to consider the evidence anew and in greater detail.³⁶ It would have been impossible for a teacher to make a living on this amount of money. It is instructive to compare the earnings of other professionals listed in the Edict. The elementary teacher earned 50 denarii; the salary of a grammarian was 200 denarii, and that of the teacher of rhetoric 250.³⁷ It appears, therefore, that stenographers in the fourth century earned more than elementary teachers but considerably less than both teachers of higher education and second-century teachers of stenography. This is somewhat puzzling, because stenography was more popular and more necessary in the fourth century. The only solution to this dilemma would be that in late antiquity a teacher would have more than one student and what had previously been taught through individual instruction was instead imparted in a class setting.

Two further texts (not on papyrus) refer to young people learning shorthand. One is part of the Hermeneumata, the bilingual school handbooks in Greek and Latin that were preserved in medieval manuscripts and most likely derived from third-century Gaul. One vignette presented in the text says, “I go out to the school of the arithmetic/shorthand/Greek/Latin/rhetoric teacher,” a phrase that cannot suggest that stenography was part of the curriculum that led to rhetoric.³⁸And yet this text is significant because it refers to a situation or even a school where stenography was taught.³⁹ An important question is whether stenographers (or professional scribes) obtained some literary education as part of the system of liberal studies. If they did, the circumstances must have varied drastically depending on the level of their skills and power.⁴⁰ The stenographic Commentary involved some familiarity with literature, but it is unclear how extensive, and evidence from the papyri is uncertain and ambiguous.⁴¹ Many shorthand writers achieved high positions under Emperor Constantius II, and in the later fourth century, after Julian’s attempt to constrain their power, they gained enviable status and large incomes when found to be necessary by emperors, as H. C. Teitler has shown.⁴² In these roles their tasks went far beyond taking down dictation, and some even enjoyed an emperor’s confidence and became trusted advisers.

At the end of this excursus and in light of the following chapters, it is essential to vindicate the credibility and precision of stenographers, who were often accused of imprecision and omissions in their work. I have recently surveyed the available evidence, concluding that they did not inflate their work; in fact, they recorded the proceedings of Christological conferences rather accurately.⁴³ Stenographers were indispensable in reproducing those documents, and their texts were then “translated” to produce the official versions of the conference, the Gesta. These were tinged by extreme polemic, and their language was fundamentally oral. It seems, however, that stenographers were rather innocent and often simply caught in the crossfire. Diminished and insulted, they appear to have done their work as well as they could.

Likewise, my analysis has shown that transcripts of sermons, which were imperfect pieces of rhetoric, can be trusted to have recorded what was delivered verbally. The evidence is admittedly incomplete, but I argue that stenographers appear unjustly accused of transforming texts and of partiality. Most of the extant sermons do not seem to have been edited, but rather represent the versions that stenographers wrote down during delivery (including preachers’ informal admonitions and requests for silence). When two texts of a sermon appear in the manuscripts, one shorter and one more extensive but differing from the other in only a few points, it is possible to identify the latter as the version taken down by a stenographer. These stenographers seem to have completed their transcriptions somewhat unthinkingly, without abbreviating certain parts or taking personal initiative. Stenographers did not edit preachers’ texts but instead took them down mechanically, reproducing all the numerous repetitions.

In part III we will discover that Didymus’s commentaries on Psalms and Ecclesiastes, though usually considered taken down by stenographers, were actually recorded by students. My analysis of the lectures reveals that someone shortened the students’ questions and included mistakes of every kind that marred Didymus’s lectures, artifacts that are atypical of stenographers.

1. He is of course building on a common Homeric metaphor.

2. Ep. 333.1. Terms used for stenographers are exceptores and actuarii. In Greek, notarioi and exceptores occur in the papyri sporadically in the second and third centuries, though the evidence increases a great deal in late antiquity. Cf. Teitler 1985, 29–31. On the evidence of papyri, inscriptions, and the literary sources, see Boge 1974, 73–102.

3. Philostratus, VS 574. Alexander was traveling and the tachygraphers had to record his performances. Cf. Heath 2004, 259–65.

4. Swain 1991, 163.

5. Lucian or Ps. Lucian, Demosthenes 44. Scholars have criticized Libanius for apparent contradictions in his attitude versus stenography, but as usual one has to take into account that his statements were different in his orations and in epistolary communications, as the conventions of the epistolary genre prevented him from revealing any highly negative feelings. See Teitler 1985, 198; A. Jones 1973, 572–73; Cribiore 2013, 124–26 passim.

6. Cf. Samuel Pepys, who composed his seventeenth-century diaries in many books in one of the standard forms of stenography of the time. People labored to decipher and transcribe the diaries until the key to the stenography system used was apparently found in Pepys’s library.

7. Cf. chapter 6 regarding Epictetus and Arrian.

8. On the Tironian notes, see Ganz 1990.

9. Mansfeld 1994, 24n36.

10. Boudon-Millot 2007, xcii–xciii.

11. Galen, De libris 1.12.

12. Galen, On Bloodletting against the Erasistrateans in Rome 1 (Kühn xi, 194, 16–195).

13. Galen, De praecognitione 5.20–21.

14. The excellent comprehensive works of Teitler are informative and irreplaceable but mostly technical. See Teitler 1985; 1990, 3–15; 2007, 518–45.

15. Genesius refused to write down the names of some Christians. Cavallin 1945; Ronchey 2000.

16. See the rhetor Julius Victor, Ars rhetorica 27. Cf. Pliny, Ep. 9.36 and Cicero, Ad. Fam. 16.10. Dio, however, in Or. 18.18 advises a young man to dictate. In the fourth century Eunapius praised handwriting, but he was isolated (VS 502, 551).

17. Arns (1953, 37–40), on the practice of Jerome, rightly considers scribo as a reference to a text being written by a secretary. Jerome listed the whole possible gamut of literary activities: hoc ipsum quod loquor, quod dicto, quod scribo, quod emendo, quod relego (“what I say, what I dictate, what I write, what I emend, and what I reread”). Gal. 3, 6, 10; PL 26. 433.

18. Augustine preferred to dictate even his marginal notes; Retract. II.58. See Cameron 2011, 490. Possidius reports that he wrote only one brief, sixteen-page text in his own hand; the rest of his vast production was written by scribes.

19. See Bagnall and Cribiore 2006, 15–19.

20. For Basil see Epp. 134, 135, 223. For Jerome, see Preface to the Comm. In Gal. PL 26.309a. See also Pref. Comm. In Ep. Ad Gal 3, an amusing passage where Jerome appears to be victimized by a servant. One wonders about the reasons behind this pitiful self-portrait of Jerome. One is that Jerome was influenced by Quint. 10.319–20.

21. On the Trinity 15.48.

22. Arns 1953, 62–64.

23. Gesta GCC 2.35; the original could have some defects. Franchina 2010, 1014.

24. McNamee 2001; McNamee and Jacovides 2003 (Plato with a new edition of the notes in the margin). Cf. also McNamee 2007, 20, and more recently, Seiler 2013.

25. One should also consider the option that the notes might be preliminary material written by someone who was giving a lecture on the Republic.

26. The presence of signs in two other papyri is very uncertain. An oratorical text perhaps of Andocides (PSI inv. 2013) shows two symbols that have been variously interpreted, with the latest editor doubting that they are tachygraphical (Conti 2013). Another papyrus from the seventh century contains some lines of a Greek text difficult to identify along with some stenography. There may not have been a relation between the Greek lines and the tachygraphical signs.

27. McNamee (2001) explains the whole system with clarity.

28. Milne and Mansfield 1934, 7–8; Menci 1992.

29. In the medieval period, the simpler method of syllabic tachygraphy was devised. See Costamagna 1990, 83–94 and note 30.

30. Only seventy-six tachygraphical papyri survive, and all of them are difficult to interpret. See Menci 2019. Most of the evidence is documentary, concerning, for example, the subscriptions of notaries, which are often followed by tachygraphical signs, and notes taken during oral presentations in legal proceedings. See, e.g., Diethart and Worp 1986.

31. Other contracts of apprenticeship to learn various skills were found in Egypt. Two Coptic ostraca of the seventh to eighth centuries bear a declaration of a priest who is promising some money to a man to teach his son to read and write. See Boud’hors 2013.

32. It has been recognized that the translation of the ed.pr. “when the boy writes fluently in every respect” was faulty. The expression logou pezou should be translated as “prose or ordinary language.” In P.Oxy. XLI 2988. 9–10 the word and the almost identical expression “when the apprentice can take down and read from any kind of prose (or everyday language)” confirms the meaning.

33. These boys were probably slaves; see Cribiore 2001b, 51. Lewis (2003) remarked that one should avoid overenthusiastically imagining that, like Cicero, the owners of these slaves were anticipating another Tiro. Lewis pointed to the fact that the aim of these contracts was exclusively mercenary, so that these boys were trained in order to produce income for the owner.

34. If we break down the figure, the teacher would have earned five drachmas per month per pupil, a good salary. Considering that wheat typically cost eight drachmas per artaba, the teacher would have been earning 5/8 of the price of an artaba per month.

35. Notarius, Edict 7, 68. He would have earned 75 denarii (that is 300 drachmas) per month. As wheat then cost 1,333 drachmas per artaba, this later teacher would have been earning only 22.5 percent of the price of an artaba per pupil per month.

36. See the limited remarks in Arns 1953, 60 and Schlumberger 1972–74, 223. See also Boge 1974, 76–78. Of course, the exact status of the prices in the Edict is much debated.

37. Scribes earned money according to the quality and competence of their writing: a hundred lines of calligraphic writings cost twenty-five denarii; of less informal literary writing, twenty denarii; and of documentary writing, ten denarii. Edictum Diocletiani de pretiis rerum venalium, col. vii 39–41. In these cases, the prices depended on the amount of time employed to produce highly formal texts.

38. See C18, that is, Hermeneumata Celtis 18. Dickey 2015, 2:170.

39. Stenography was not taught in villages. See Ep. 157 of Gregory of Nazianzus: his nephews had to leave their town and go to the city of Tyana to learn the art.

40. A letter from Augustine shows that some stenographers only had a smattering of the art (Ep. 44.2 of the year 398).

41. Torallas Tovar and Worp 2006, 29–35. Though the stenographic signs are missing, there is no doubt where the list originated. Most of the entries belong to the stock of basic words, with the addition of some poetical and technical terms (for example, medical terminology) and words from lexicographical sources such as Hesychius.

42. Teitler 1985, 34–37; 68–72 passim.

43. Cribiore 2021.