Where we left off yesterday, in our survey of the history of Scientific Linguistics: We were talking about the very fruitful American school of Anthropological Linguistics, whose most famous figure was Edward Sapir. Sapir was born in what was then Prussia, now a part of Poland, but emigrated to the U.S. as a child, and became a student of another great scholar, Franz Boas, whom some call “The Father of American Anthropology”.
This marriage of Anthropology and Linguistics made sense in the American context, since these scientists had a wealth of native cultures and languages to study. Right out there on the Indian reservations! There were literally hundreds of new languages just waiting to be written down. So many languages, so little time… Put yourself in a mind experiment: What would you do if you were placed among people who appear to speak gibberish, and you were asked to quickly learn their language and develop an alphabet for them?
Well, the first thing you need to do is hire a Native Guide. Somebody who speaks both their language and your language. This person will translate back and forth for you. Without this bilingual assist, alas, the job simply can’t be done. Let me explain: Since my Athabaskan is a tad rusty these days, I am going to be making up some egregious examples from a fake language, but I am not making up the method; this is really how it was done, silly as it all might sound.
So, you and your native guide, let’s call him Ishi, approach the oldest member of this tribe, she is a nice old lady of some 125 years, she and Ishi are only a handful of people who still know this particular language. After they die, it’s gone, so you have to act quickly and get it all written down. You and Ishi are offering this tribe a complete package: You’ll come up with a way of writing down their language (unlike Saints Cyrill and Methodius, you don’t actually need to invent an alphabet for this, you’ll just use the Latin alphabet, possibly with some diacritical marks) and compile a vocabulary and grammar. Most of Linguistics is about Words, but in order to get to the Words, you first have to start at a lower physical level, at the level of Sounds.
Step #1 is to compile a catalog of the phonemes of this language. In other words, the vowels and consonants. Languages vary, a typical language might have, say, 30 consonants and 10 vowels. The record for a “consonant-heavy” language is some extinct Caucasian language called Ubykh, which logged a whopping 84 consonants! It is said that Ubykh only had 2 vowels. Well, with so many consonants, it only needed two vowels! On the other side of the ledger, there is a language called Sedang, related to Vietnamese, which is said to have 55 vowels! But those are extreme examples. Castilian Spanish, for example, is a much more “normal” language, with only 5 vowels and 20 consonants.
Phonetics vs Phonemics
This distinction is a source of endless confusion and is, in fact, what sparked my internet feud with “Otto”. Humanoid mouth-parts comprise a complicated apparatus consisting of tongue, teeth, lips, palate, larynx, lungs and even noses. All together, this apparatus is capable of emitting a huge number of different sound waves on the acoustics spectrum. But not every sound emitted by a human mouth is meaningful speech. A burp is not meaningful speech, no matter what your six-year-old tries to tell you. If a Castilian is talking to you in real words, then, logically speaking, each one of those words must be comprised of the aforementioned set of 25 phonemes that have been catalogued for this language.
It is said that when the ancient Greeks first encountered people speaking a completely unknown tongue, it sounded to them like these guys were just babbling “BA-BA-BA-BA!” From this, the Greeks called them “BARBARIANS”, which is the origin of that word. And no, it would not be possible for any human language to be so binary as to have only one consonant and one vowel. If that were the case, then the language could only possess a very small handful of words, such as “BABA”, “ABBA”, “ABAB”, “BABAB”, etc. Unless they went in for really long words with many syllables, for example”BA-BA-BA-BA-AB-AB-BA-AB”, which is their word for “Stinky Greek”.
Returning to our friend Ishi, it is time for him to sit down with that little old lady and start writing down her language. First, recall the 7 layers of Scientific Linguists, starting with the most physical layer and moving on upwards to the most abstract:
- Acoustic phonetics – the study of raw sounds and sound waves
- Phonetics – how uttered sounds are produced in the human vocal apparatus
- Phonemics – the study of meaningful sounds, like those 25 Castilian phonemes
- Morphophonemics – the study of meaningful sounds in combination
- Morphology – the study of words and their various parts
- Grammar – the study of words in phrases and sentences
- Semantics – the study of language as a component of Symbology
Remember that your and Ishi’s goal is to compile a list of words and phrases of Language X. Which starts with a catalog of phonemes. And includes the task of separating out the real phonemes from their acoustic variants (which are called allophones). But here you are forced to cheat a bit, you have to peek ahead at the higher level – morphology – because you need information from that level in order to catalog the phonemes. It’s sort of a chicken and egg thing. The end goal here is to derive a set of what Linguists call “minimal pairs”. This would be a pair of words with two completely different meanings, but the utterance of them differs by only one sound. You get bonus points if can find minimal pairs that rhyme.
Some examples in English: cot, dot, got, hot, jot, lot, not, pot, rot, tot.
From these examples alone, we can safety catalogue at least 11 phonemes in the English language, namely K, D, G, H, J, L, N, P, R, T, plus the vowel O. We’re not done yet, but it’s a good start. We know these sounds are phonemes (=meaningful sounds) because each of these sounds helps to distinguish one word from another. These sounds don’t carry meaning in and of themselves, but in the context of a word, they help to deliver information (as computer scientists say) and semantic meaning.
So, basically, you and Ishi, you need to get going and rack up some words. So Ishi starts by respectfully addressing the old woman: “Grandmother, can you say something for me in the old tongue?”
The old woman points at a nearby hen scratching at the ground and utters [made up language]: “Klaatu barada nikto chikino.”
You scribble this down and turn to Ishi: “What did she say?”
She say: ‘The chicken is scratching at the ground.”
[to be continued]