Fairwell to Smooth Breathings! and Patreon Update (Patreon)

Dearest Patrons,

Your generosity is what allows me to do the work I am most passionate about, and I am deeply grateful for this. I apologize that some of the regular series have been updated less frequently. This is due to the fact that I am preparing for you a rather grand project that is months in the making. The audio series are going to resume with more regularity quite soon. And even better, I think, will be the exclusive Patreon release of this big project (which will remain unsaid for the sake of not spoiling the surprise) in the coming weeks. 

I intend to release this exciting project to you, I hope, before the end of the month. In the meantime, I wanted to share with you an extract to the introduction to that project, which will explain why, as you may have seen in my most recent ScorpioMartianus video, I have decided not to write the smooth breathing any longer in Ancient Greek.

I would like to add that my little essay here is not a plea for others to stop writing the smooth breathing; everyone should feel free to follow their own heart and reason in this matter. It is mere an απολογίᾱ, of sorts: an explanation for how I come to my decision.

Thank you again, and feel free to express your opinion on the matter.

FAIRWELL TO SMOOTH BREATHINGS

Standard Ancient Greek orthography contains the diacritical marks known as rough and smooth breathings. These did not exist in Attic orthography until 403 BC when Euclid the Archon reformed the way Greek was spelled in Athens. Prior to that year, Classical Greek was written something like this: εχω “I have,” hεξω “I will have.” The reform led to the result we have today: ἔχω “I have” with smooth breathing, and ἕξω “I will have” with rough breathing. Thus the rough breathing corresponds to English and Latin /h/, while the smooth breathing does not carry phonetic value.

In consulting with colleagues Logan Kilpelä and Carla Hurt, who have long advocated the abolition of the smooth breathing at least in pedagogical texts (as did the authors of Ancient Greek Alive Paula Saffire and Catherine Freis), I have persuaded myself to remove the smooth breathings from future materials. I also spoke about this with Stefano Vittori, who takes the opposite position and advocates for their retention. His anthithesis to Logan’s and Carla’s thesis has brought me to the following synthesis.

The common understanding is that the smooth breathing exists to help mark the beginnings of words in an orthography that seldom employed spaces or interpuncts. Since so many Greek words start with a vowel, the utility of both smooth and rough breathings in reading is immediately clear. Take this example text:

Ἰδοὺ αἱ πύλαι ἀνοίγονται καὶ τὸ ἱππικὸν ἡμῶν εἰς τοὺς πολεμίους εἰσβάλλουσιν. Οἱ βάρβαροι περίφοβοι, οἳ τὴν τοῦ ἱππικοῦ εἰσβολὴν ἀνέχειν οὐ δύνανται, τὰ ὅπλα πρὸς τὴν γῆν βάλλουσιν καὶ εἰς τᾱ̀ς μεγάλᾱς ὕ̄λᾱς φεύγουσιν.

Without spaces or punctuation:

ἰδοὺαἱπύλαιἀνοίγονταικαὶτὸἱππικὸνἡμῶνεἰςτοὺςπολεμίουςεἰσβάλλουσινοἱβάρβαροιπερίφοβοιοἳτὴντοῦἱππικοῦεἰσβολὴνἀνέχεινοὐδύνανταιτὰὅπλαπρὸςτὴνγῆνβάλλουσινκαὶεἰςτὰςμεγάλαςὕλαςφεύγουσιν

The same without rough or smooth breathings:

ιδοὺαιπύλαιανοίγονταικαὶτὸιππικὸνημῶνειςτοὺςπολεμίουςεισβάλλουσινοιβάρβαροιπερίφοβοιοὶτὴντοῦιππικοῦεισβολὴνανέχεινουδύνανταιτὰόπλαπρὸςτὴνγῆνβάλλουσινκαὶειςτὰςμεγάλαςύλαςφεύγουσιν

It is apparent to me that the breathing marks are very helpful in recognizing word boundaries. (I believe this explains also why the ubiquitous letter sigma has a final variant: ς.)

However, Stefano’s argument is that the presence of a smooth breathing is not merely to mark the beginning of a word, but is also an orthographic signal from the scribe to the reader that clearly states: “I didn’t forget the rough breathing on this word: it really does begin without /h/.” He notes that such diacritics are used in Syriac for similar purposes of deconfliction.

I was glad to hear Stefano’s view, since it helps to explain why the smooth breathing would come to be a part of so many ancient manuscripts, and why the tradition passed on through the Ancient, Mediaeval, and Modern Greek languages all the way to the 20th century. It also seems quite reasonable to retain the exact spelling found in an ancient manuscript or inscription for purposes of textual analysis; for example, if a Latin inscription has iuli caesaris uictoria, it makes sense to present the text exactly so when transmitting the initial transcription, without distinguishing majuscule from miniscule letters, without converting the consonantal ‘i’ and ‘u’ to ‘j’ and ‘v’, without adding macrons, and without standardizing the genitive ending. But to normalize the text for the purpose of clarity, I would change it to Jūliī Caesaris victōria since the pedagogical value of these orthographic features is quite high. Equally, if an ancient manuscript or inscription has the smooth breathing, a loyal presentation of the mark is obligatory for precise analysis.

But for pedagogical texts, and, as more people learn to write in Ancient Greek, for general communication, the smooth breathing only presents a non-trivial obstacle: fluent reading requires a certain degree of speed, nearly at the speed of speech, and using any reasonable font size in modern books or on digital screens, even those with good vision will occasionally misread ἑ for ἐ — this is not a bug, but a feature of true fluent reading itself; our visual system when reading works by jumping to focus on the center of each word (or of word clusters in high reading fluency), and the mind determines meaning and pronunciation not letter by letter, but more like a whole graphic. Thus, given the optical smallness and similar appearance of ἐ and ἑ, the chance for confusion is quite high.

For those who do not pronounce Ancient Greek with /h/, one can admit it does not cause confusion in recitation since both diacritics are simply ignored by those speakers; however, the vast majority of users of Ancient Greek do make a phonetic distinction. Therefore, the presence of a very small mark with no phonetic value that looks nearly identical to another with very important phonetic value, especially when next to accents, such as ἄ vs. ἅ, ἆ vs. ἇ, slows the learning process, and is an obstacle to reading even for experts. This has been observed by myself, Carla Hurt, and Logan Kilpelä over years of teaching, and, I would imagine, by teachers of Ancient Greek for many generations.

The only pedagogical reason for the retention of the smooth breathing would be to prepare students for all other texts that bear it. This is the main reason I can foresee to justify their inclusion early in the curriculum.

Thus the practical argument has two sides. Side one: it is practical to teach Ancient Greek with smooth breathings since the student will eventually encounter texts where they occur, and they ability to adapt is dubious. Side two: it is practical to teach Ancient Greek without smooth breathings since this improves learning speed, and adapting to read Greek text with smooth breathings is easy.

Experimentally gathered evidence would be required to create some kind of statistical analysis after semesters of trying both approaches. For now, both the observed problem and the positive result seen by me, Logan, and Carla are merely anecdotal.

Stefano’s argument, which is that the smooth breathing’s purpose is for the scribe to tell us “this vowel indeed has no aspiration,” reminds me of the fact that the grave accent was once written on all vowels that did not carry the acute or the circumflex; for example, instead of ἡ Ῥώμη ἡ τῶν Ῥωμαίων πόλις ἐν τῇ Ἰταλίᾳ ἐστίν, this would be written ἣ Ρὥμὴ ἣ τῶν Ῥὼμὰίὼν πόλὶς ἒν τῇ Ἲτὰλίᾲ ἒστίν. This convention was eventually dropped, since the grave accent used thusly was redundant for its own absence. Equally, one may write in Latin with both macrons and breves: Rōmă, ŭrbs Rōmānōrŭm, ĭn Ĭtălĭā ĕst, but these carry essentially no useful information; moreover, much like the smooth breathing, they reduce the legibility of the text, since we have to ignore quite a lot of ink on the page that is not conveying anything particularly pertinent to the sentence, and in fluent reading the breves can be misinterpreted as macrons. When I presented this counter-argument to Stefano, he responded saying the breves actually make the text more precise, because then we know that the scribe truly intends the vowels to be short. There is value in this, and I’m glad Stefano brought this to my attention. However, I opine the value is much lower than the confusion of additional diacritics.

Instead, if we simply trust the scribe, as we have trusted Ørberg himself who taught us the long and short vowels in every word through only marking the long ones, we understand the exact same information in the text. Indeed, Ørberg used the breve only in situations where the vowel is normally long, such as illĭus, which tells us the incredibly useful fact that the standard spelling and pronunciation is illīus. Equally, I imagine that a written Ancient Greek standard eschewing smooth breathings would only retain them for psilotic dialects of Greek (ones that don’t have the sound /h/) like Aeolic. Thus Sappho’s poem, normally written like this:

Δέδῡκε μὲν ᾱ̓ σελάννᾱ

καὶ Πληΐαδες, μέσαι δὲ

νύκτες, παρὰ δ’ ἔρχετ’ ὤρᾱ,

ἔγω δὲ μόνᾱ καθεύδω.

I would instead transcribe like this:

Δέδῡκε μὲν ᾱ̓ σελάννᾱ

καὶ Πληΐαδες, μέσαι δὲ

νύκτες, παρὰ δ’ έρχετ’ ὤρᾱ,

έγω δὲ μόνᾱ καθεύδω.

Hence words which have aspiration in Attic-Koine, like ὥρᾱ, are written as ὤρᾱ, emphasizing the contrast, while words that don’t normally carry aspiration like έγω (though note the change of accented syllable in Aeolic) have no mark. This seems to be to be a considerably more efficient and effective employment of the smooth breathing, and parallels illĭus for illīus.

By my lights, given my positive experiences hitherto in removing the smooth breathing in pedagogical texts, I can only justify the smooth breathing for its aesthetic value. Here again is the above sample text, but the smooth breathings removed:

Ιδοὺ αἱ πύλαι ανοίγονται καὶ τὸ ἱππικὸν ἡμῶν εις τοὺς πολεμίους εισβάλλουσιν. Οἱ βάρβαροι περίφοβοι, οἳ τὴν τοῦ ἱππικοῦ εισβολὴν ανέχειν ου δύνανται, τὰ ὅπλα πρὸς τὴν γῆν βάλλουσιν καὶ εις τᾱ̀ς μεγάλᾱς ὕ̄λᾱς φεύγουσιν.

When I first saw the text without the smooth breathing, it immediately looked odd to me. But since its removal causes zero confusion, I found that I had completely adapted to its absence in less than a day, and that, immediately after the initial change, I no longer missed it all. Having flourishing strokes growing dendritically from letters is an extremely lovely orament, as are the curves of cursive handwriting; but there is a reason why cursive style text is seldom employed in printed materials, even well after technology has afforded us the ability to convert any text into a cursive style: legibility is key.

In the end, Stefano’s cogently presented argument in favor of the smooth breathing has convinced me all the more that their removal is as sensible as the removal of non-final grave accents. Similarly, the removal of breathing marks on geminated ρ has become standard in the past century, where in past centuries it was universally common: geminated ρ used to always carry a smooth then rough breathing, for example ἄῤῥην “masculine,” which we would transcribe as arrhēn, in order to remind us that the geminated trilled /r/ of Ancient Greek is voiceless (see this video of mine on the pronunciation of ρ for more). But geminated trilled /r/ is always voiceless in Ancient Greek, thus it has been left out of most newer texts in the past century due to the fact that it is not necessary.

If we put in the same category non-final grave accents, breathing marks over geminated ρ, and initial smooth breathings, then we can appreciate how the standard has slowly changed for the better, and ending the smooth breathing’s ubiquity is an appropriate step in the same direction, consistent with the orthographic tradition of making small updates for the sake of efficiency and pedagogy. Take the word for “farewell,” which would more archaically be written ἔῤῥὼσὸ, and then became ἔῤῥωσο, passing to ἔρρωσο, and now, finally, to έρρωσο. I think it’s time we said farewell to smooth breathings.