The Blog of Ariphron is a language-learning blog. I expect it to eventually branch into other languages that interest me (Middle High German? Old Vernacular Chinese?), but for now it will focus on Ancient Greek and my attempts to learn it by making audiobooks.
This idiosyncratic approach to language learning came to me not because I think it is the most efficient way to a reading competency, but because for me, with the way I think, it seemed to be a necessity.
I first started learning Ancient Greek in a summer course when I was a university student. The textbook was Hansen & Quinn (Greek: An Intensive Course); all the homework involved translating the artificial example sentences in the book; and while there was some oral drill practice in class and the instructor tried to bring us through the first part of Plato’s Apology, the bulk of class time was spent going over the homework exercises. I learned several things from this experience:
First, I don’t want to spend a lot of time translating in the early stages of learning a language – especially when “translating” means writing out awkward, stilted, and wordy English formulas presented as equivalent to elegant and concise uses of Greek particles and verb forms.
Second, the Erasmian pronunciation scheme really bothered me. (Erasmian means following the general principles of a suggestion once made by Erasmus. In practice, an Erasmian system is one in which each sound of Greek is described as “like” a sound in your native language, and then – for English speakers – you go about pronouncing Greek words with English sounds and English accents. There are almost always a few cases where different sounds of ancient Greek, spelled differently, are pronounced the same because your language doesn’t have the right variety of sounds.) That pronunciation simply put everything out of balance. There seemed to be a beauty to the accentual system and the interplay of rough breathing with plosive consonants, and a logic to irregular inflections, all just beyond my grasp, and I felt that if I just found the right pronunciation scheme, everything would become intuitive. My instructor recommended that I read Vox Graeca by W. S. Allen, and I did read it that summer. What it suggested was a big improvement over what I heard in class, but after some exploration I decided that it would be too hard for me to practice, especially when it came to the tone accent.
Third, it seemed very wrong to practice all these advanced grammatical patterns when we had practically no basic vocabulary. The textbook was designed to go through all the fundamental grammar of the language in a minimum of time. When you learn a new grammatical concept, you get a few new words to illustrate it, and the exercises are constructed using those words to explore the possibilities of the grammatical construction. Everything is totally grammar-oriented, and it is often a stretch to say that the example sentences “make sense.”
Not long after the conclusion of that course, I gave up on learning Ancient Greek, taught myself some French, brought my German up a level, and some years later started learning Chinese.
In early 2013 I decided to start learning Greek again. With Chinese tones, I had a lot of experience practicing arbitrary intonation contours until they were rapid and reliable, and listening for the contrasts that they express, and I felt equipped to practice a system for realizing the Greek tone accents until I could execute it consistently and tell if it needed improvement. I started with two-word phrases, working through vocabulary lists out of Hansen & Quinn and declining or conjugating each word in turn. About that time I read Devine & Stephens, The Prosody of Greek Speech, and tried to put its ideas into practice, and I found the basic textbook/readers Athenaze and A Greek Boy at Home, which introduce a lot of basic vocabulary in context before getting into advanced grammar. These became the focus of my study for the rest of the year and into 2014.
When I started making recordings of myself reading passages in Ancient Greek, the only audience I had in mind was myself, later. While I was speaking, I knew if something felt right, but I couldn’t really tell if it sounded good. Recording gave me the distance to judge my pronunciation with some objectivity. But after about a year I decided to make some of my recordings available online. Partly this was because I wanted some feedback on them, although I was not optimistic that anyone would be qualified to give me detailed constructive advice on my own pronunciation scheme, but the main reason for putting them online was because I felt that they might be useful to other learners of Ancient Greek.
Audiobooks can be useful even if they are not pronounced perfectly, and even if they are badly enough pronounced as to be unintelligible. It’s nice if an audiobook is suitable to listen to in the car or in the gym, but not necessary. The most likely activity for a modern student and Ancient Greek audio is listening while following along the printed text. This has many potential advantages over simply reading silently with book in hand, and is almost equally valuable no matter how the words are pronounced. First of all, if you have not yet mastered the writing system, the audio can correct your reading whenever you misread a letter or a diacritical mark. Second, it provides pacing. When reading silently, you can always decide to speed up or slow down your pace; you can stop to look up a word or reread a passage. This multiplicity of choices – none of them available when listening to a real interlocutor in conversation – can be paralyzing. Following along makes reading into an activity more like listening, and therefore more valuable for language learning, although there certainly are students who can get the same benefit without using a recording, as they have the discipline to maintain a good pace without external stimulus. I believe that listening is the fundamental language skill, which means that the ability to process and make sense of a stream of words (perhaps better analyzed down to a stream of morphemes, or word-constituents), delivered once, and at a pace determined externally, is the core competency for control of a language. The exact modality by which the stream is communicated to the listener/reader and decoded is a secondary consideration.
For people who aspire to practice conversational Ancient Greek, it should be invaluable to make and listen to audio recordings, and with this application it matters a great deal how you pronounce the language. If you use a system that distinguishes many fewer sounds than the habitual manner of the ancient speakers who are supposed to serve as your model, then for clarity you will have to avoid many idioms that they found natural, because one expression will just happen to sound like something else with a completely different meaning. Many grammatical features will also become problematic, because, for instance, a dual noun may sound like dative singular, or a verb in the subjunctive will sound like the indicative. This doesn’t mean that speakers who use different pronunciation schemes will have trouble communicating. On the contrary, provided listeners have a modest familiarity with the pronunciations used by speakers, misunderstandings will be tolerably rare.
When I read something in a foreign language for which audio is available, I generally follow along with the audio, sped up or slowed down to the rate at which I can catch all the words/morphemes, but just barely. The tool I use for this purpose is Sonic Visualizer, available for free from Queen Mary, University of London. It has speedups and slowdowns in reasonable increments: “Playback speed: +40%” means sped up by a factor of 1.4, and “Playback speed: -40%” means slowed down by a factor of 1.4. (This may be an incorrect use of percents, but it’s clear enough.) The speed-adjusted audio sounds much better to my ears than other free options I have tried, such as Apple software (old versions of QuickTime Player or modern versions of iOS Podcasts) or VLC. Usually I divide the reading into segments of five minutes or so, listen to each segment two or three times before going on with the next, and only then decide if there are words I need to look up.