The Ukrainian alphabet: Looks a lot like Russian!
Please bear with me just a little longer, we are almost at the point where we can “decide” whether or not Ukrainian is a full-fledged language on its own, or just a dialect of Russian. But first we had to work our way through some business involving alphabets and phonology. Ukrainian uses the same Cyrillic alphabet as Russian, well, admittedly, with the addition of a couple of other letters.
But we learned that alphabets do not make a language. I gave the example of Serbo-Croatian. “Serbian” and “Croatian” are exactly the same spoken language, but each uses a different alphabet. Lordy, I could start writing these here English words in Cyrillic letters, but it wouldn’t make them Russian. Moral of the story: Don’t look at the alphabet, it’s not a factor in this case. But do look at other things, such as the actual catalog of phonemes of a given language, regardless of how they are spelled on paper. And we have a little bit more work to do in this regard, as we tackle the issue of those pesky [allophones]. After we finish with Phonology, then I will leave you in peace, Dear Readers. For our purposes, it is not necessary to go through all the layers of grammar and morphology, or whether a language is Agglutinative or gluten-free. We already know that Ukrainian is an inflected language, just like Russian!
Otto boning up on his debating skills.
So, let’s get to work again. As I curtly informed Otto, “You are entitled to your own opinions, but not to your own facts.”
“Otto”, as I mentioned before, is one of those Russian “kvass patriots” who denies that “Ukrainian” is either a language or an ethnic group. Otto also believes that anybody who does not condemn the writing down of Central Asian languages in Turkish letters, is a Trotskyist and a Jew, not to mention a cuck-socking maggot. I am not sure if Otto believes that Aristotle was born in Belgium, but Otto definitely “believes” that that no European language contains palatal plosive type consonants. Like that German consonant in the word “Zeit”.
Leaving Otto to his intellectual misery, let us turn to a much more pleasant company, namely our native guide Ishi and his brilliant pale-faced companion, the American Anthropologist. We’ll call him Kemosabi, which is Indian for “Wrong Brother”. Where we left off, Ishi was questioning the old doyenne of the tribe, while Kemosabi was busily scribbing down the resulting utterances using pencil and paper and the international phonetics alphabet. Job #1, using minimal pairs of words, is to catalog the meaningful sounds, or phonemes of the language. And to weed out the allophones. Rule #1 of a good alphabet: No stinking allophone deserves its own letter.
What Is An Allophone?
The best way to explain is with an example. As the old woman chatters away in her tongue, Kemosabi keep hearing her emitting this buzzing [ZZZZ] sound. Like there was a bee in her bonnet. He thinks, maybe this language includes the /Z/ phoneme. Like the first sound in English “zoo”. But he isn’t sure yet, until he can find a minimal pair of words that either prove or disprove his hypothesis.
“My barkats are bothering my chikinoz.”
Now, as it so happens (and crafty Ishi already knew the answer, but he wanted Kemosabi to figure it out for himself), this particular made-up language uses the same plural-marker as does English. Just by sheer coincidence. For example, their word for “hen” is “chikino“. When the old lady is talking about her hens, she keeps saying “chikino-z“. Meanwhile, their word for dog is “barkat“. When she she talks about her (plural) dogs she says “barkats“. (I forgot to mention, that in this language, every animal word ends in either the consonant /t/ or the vowel /o/.)
Eventually it dawns on Kemosabi that the allophones [s] and [z] are simply variants of each other. They are one and the same plural marker, with [s] being pronounced after a consonant, and [z] being pronounced after a vowel. This is a normal thing that happens in many languages. Here is how it happens at the lower (physical) layer of acoustic phonetics: The sound known as [s] is what phoneticians call a “Fricative sibilant dental alveolar” consonant. The sound known as [z] is exactly the same. Both sounds are pronounced exactly the same in the human mouth. Native or bilingual English speakers: Try this experiment: Say out loud the /S/ sound of the word “Sue”, stretching it out as much as possible, hissing like a snake. Now do the same with the /Z/ sound of “zoo”. Notice that your mouth, teeth, tongue, etc. are in exactly the same position for both. The sole difference between these two sounds is that your vocal chords vibrate when you say /Z/. If you don’t believe me, then place your hand on your own throat, like you were going to strangle yourself, and try the experiment again.
This distinction is called “voiceless” vs “voiced”. There are numerous consonant pairs where one member is voiced, and its twin is voiceless. Oh, and it goes without saying that ALL vowels engage the voicebox, hence, All Vowels are Voiced. It’s only consonants that get to make that choice: They can go quietly, or they can go noisily.
Moral of the story: It totally makes sense for a supposed “pair” of consonants to actually be one and the same consonantal phoneme, pronounced exactly the same way, except for that one tiny distinction of voiced/voiceless, depending on the context. For this language, Kemosabi has determined, with some hints from Ishi, that [s] and [z] are one and the same phoneme. Again, it totally makes sense, from a physical point of view, that the voicebox, fully engaged in the pronunciation of that /o/ vowel in “chikino”, continues to vibrate, like a lost echo, as the plural marker is pasted on, turning the [s] into a [z]. Again, this happens in many languages, it’s just a physical fact, and Kemosabi knows that too, having studied so many other languages. There is an interesting process going on here, where Kemosabi , in compiling his catalog of phonemes, has to peek upward (at the morphology and vocabulary) as well as downward (at acoustics), while working at the level of pure phonology. It’s kind of a holistic process. And he comes to the determination that Ishi’s language has no [z] sound per se, [z] is just a variant — an allophone — of [s].
But wait! ejaculates the discerning reader. What about the word “zoo”? Well, “zoo” is an English word, and we’re not talking about English, dummy, we’re talking about Ishi’s language. I forgot to mention that Ishi and Kemosabi already determined that there is not one single other word in this language that employs the [z] sound. It only occurs in that one context, as a plural marker after a vowel. As for the [s] sound, why, there are tons of other words that use this sound, for example, their word for “river” is /sa-wim/. Proving once and for all, that the the real phoneme here is /s/. And so, Kemosabi decides that their alphabet, which he is creating for them, will include a symbol for /S/ but not for /Z/. When he creates their new dictionary and writes down the plural word “chikinos”, he will spell it with an /S/, even though the lady pronounces that final sound as a [z], with her wrinkly old throat vibrating like a musical instrument.
How To Train Your Dragon
I will close today’s installment by putting in another plug for voice-recognition software, such as Dragon Speak and others. As I noted before, I don’t own any stock in these products, nor do I get paid to advertise them. I don’t even own one myself! (I just type words the old-fashioned way.)
Nine out of ten doctors prefer to dictate their medical notes!
My primary point here being to “prove” to Otto and the others that Linguistics is actually a Science. I may be preaching to the choir here, but you’d be surprised how many people out there in the world don’t believe that Linguistics is a real science. Yet these products, boasting something like 98% accuracy in transposing speech to text are the proof that it is! Were these great products not based on scientific principles of phonology, then they could not possibly wrack up such an impressive result.
I’m sure that we have all been in the situation where we are floundering through some telephone menu based on “voice recognition”. The machine on the other end cannot understand a word that we are saying. I remember a particularly frustrating experience trying to change a flight out of Nashville, Tennessee. Where people talk funny to begin with. The robot on the other end: “Please say your flight number.” I speak the words slowly and distinctly. “I’m sorry, I didn’t get that,” says the bot. After several unsuccessful tries and almost an hour later, I am screaming into the phone: “Can you please just put me through to a human??”
Thus proving that “voice recognition” software is a crock. Except when it is based on scientific principles of phonology. As is Dragon. And here is how it works: You train the software to recognize your own personal speech patterns, regional accent, whether or not you thpeak with a lisp, etc. By speaking into a microphone a text which was picked specifically because it contains all the phonemes of the language (in this case, say, English), in the various acoustic contexts.
“And so I say to you, my fellow Americans: Ich bin ein Berliner!”
For example, here is a link to the training manual for Dragon. If you jump forward to page 3 you see that the vendor gives you a choice of text to use when training your dragon to the sound of your own voice. You pick one of these texts and read for about 30 minutes. One of your choices is President Kennedy’s Inaugural Address. Please resist the temptation to fool around and attempt a Bostonian accent. This will only confuse your Dragon.
As you blather on, Dragon absorbs all your little quirks. (This is what is lacking in, say, Airline Voice Recognition phone menus, which don’t know your voice from Charlie Manson’s.) For example, if you really do hail from Boston and always pronounce the word “car” as “cah”, then Dragon makes a note of the fact that you tend to drop your final /r/’s.
End result: Dragon has tied your speech patterns to the catalog of phonemes of the English language, taking into account their various allophonic contexts. Then it gets more complicated, because the software has to maintain a huge database/dictionary of vocabulary and know how words are spelled, etc. But we have covered enough for now, and I think we are ready to move on to the issue of the Ukrainian language.
[to be continued]