| View previous topic :: View next topic |
Ishmael
In: Toronto
|
|
 |
|
I have long been a skeptic of Mick's contention that an infinite number of sounds exist in spoken language. But a few nights ago, I became a believer.
A friend of mine has learned to speak Thai. He was describing to me how the same word means three vastly different things when pronounced in ways that -- I swear to god -- sounded absolutely identical to my ear.
How would one go about alphabetizing such a language?
What effect would 500 years of alphabetization -- using a foreign character set derived from another spoken language no less -- have upon the spoken form of a language like Thai?
|
|
 |
|
 |
|
Mick Harper
Site Admin
In: London
|
|
|
|
| But isn't spoken language also an artifice? |
The evidence (even from Horizon this week) is that human language is organic. I do not deny that 'artifice' might be involved along the way "What shall we call that?" but basically there is no overriding intelligence.
| If there were a natural language, which would mean a natural correspondence between thing and name and a natural syntax (not to mention an objective differentiation of the things), there would be only one language. But no: vocabulary and the specific elements of grammar are arbitrary. |
This is definitely wrong. Birdsong tells us that, yes, there are underlying commonalities (as with all human languages) but variation begins from Day One. It appears to me that the observed language variation on the earth today (that is sans the linguists' twitterings) are completely 'organic'. It is only when writing (and perhaps linguists) enter the fray that artifice starts to edge out natural development.
| Now when you say that languages cannot be alphabetised because there are a great number of sounds in any language, do you mean dialectal variation? |
No. I mean every natural language, including all dialects have a great many sounds. Try spelling out a bit of Mummerset speech, and you will give up at the beginning of "Aargh, Jim, lad."
| If you did, I would understand your meaning to be: alphabetisation is always based on a single, usually urban dialect and ignores the others, therefore the whole language is not alphabetised, but only a fragment of it. |
That is correct. But it only applies to post 800 AD when it became generally realised that even a single, urban dialect could be alphabetised. It is probably true however that that single, urban dialect had itself undergone a good deal of gentrification, aping the full alphabetised written languages about it.
| But apparently you don't, because your example for the notion that every language contains 'an infinity of sounds' is that the initial sound of actual and apple are different even in the mouth of the same speaker, which makes dialect distinctions irrelevant. |
No, you're missing my point. I am saying that we only hear (and to an extent, say) the two a-sounds alike because we have been trained for several hundred years by the fact that a is used for both. Originally they were (let's say) upple (as in rustic stage daft-English) and ectual (as in stage upper-class English). The accents/ dialects of England are themselves all alphabetised bastards by now but they do retain many of the various original sounds that organic English used to hold.
| I say that even if that difference exists, it's insignificant because it's only perceivable by polymath geniuses, and so any given language uses, say, thirty or forty categories of your multitudinous sounds, which we dullards perceive as only thirty or forty sounds. So where's the obstacle to alphabetising? |
Yes, you are right. Now. Although even the International Phonetic Alphabet says you need 120 sounds or so to render English (check that, Mabel). The only obstacle to alphabetising is that we have to get these 120 sounds down to twenty-six letters (which we do by various techniques, mainly di-graphing and only then via specific education). But I am claiming that in c 1100 AD there were let's say a thousand sounds in English -- too many for even the most extreme alphabet soup. So in them days only education was available to know which set of Roman letters corresponded to which particualr word. Remember, the poor sods did not even know they could use the good old ideogram method!
| But a few nights ago, I became a believer. |
Yes, you would have had the same experience with a Chinese-speaker or indeed anybody with a language that is ideogrammatic (I assume Thai is either ideogrammartic or alphabetised very recently). Alphabetic-speakers learning these languages always complain about these things because they have spent hundreds of years ironing them out!
|
|
| |
|
|
|
Mick Harper
Site Admin
In: London
|
|
|
|
|
In the Horizon programme, my contention was illustrated by somebody, for a laboratory experiment, having to come up with hundreds of made-up 'English' words. Every single one was completely alphabetical! They did not even make use of di-graphs, just the twenty-six letters. Naturally nobody noticed this. Our desire to be alphabetical means that even ch and ea sounds are becoming slightly unnatural.
|
|
| |
|
|
|
|
|
|
|
|
| A friend of mine has learned to speak Thai. He was describing to me how the same word means three vastly different things when pronounced in ways that -- I swear to god -- sounded absolutely identical to my ear. |
Thai is a tone language and you're tone deaf, or at least not familiarised enough with the tones to distinguish them. It's also possible that your friend is not very good at rendering the tone differences.
| How would one go about alphabetizing such a language? ...What effect would 500 years of alphabetization -- using a foreign character set derived from another spoken language no less -- have upon the spoken form of a language like Thai? |
Wiki says: "The Thai alphabet is derived from Old Khmer script...which is a southern Brahmic style of writing called Vatteluttu. ...According to tradition it was created in 1283 by King Ramkhamhaeng the Great." But what difference does it make if the alphabet is foreign or autochthonous?
| I assume Thai is either ideogrammatic or alphabetised very recently |
Vid. supra and weep.
|
|
| |
|
|
|
|
|
|
|
|
| The evidence is that human language is organic. |
| Birdsong tells us that, yes, there are underlying commonalities (as with all human languages) but variation begins from Day One. |
| It appears to me that the observed language variation on the earth today (that is sans the linguists' twitterings) are completely 'organic'. |
| It is only when writing (and perhaps linguists) enter the fray that artifice starts to edge out natural development. |
| Try spelling out a bit of Mummerset speech, and you will give up at the beginning of "Aargh, Jim, lad." |
An alphabet is simply a convention, but a complex one, which never pretends to represent in a one to one fashion the sounds of speech. For example, the letter e in written Italian, which we are told is highly 'phonetic', can represent a closed or an open vowel, and in unaccented syllables it must be closed. In accented syllables it can be either. In the pin-yin romanisation of Mandarin, u can be u (as in rune) or ü, e can be e (as in bet) or ê: the trick is that the second variant only occurs after certain consonants, and the first never occurs after them. And English uses five letters to represent 12 or 14 vowel sounds.
So however many sounds a language has - clearly distinguished, not infinitely subtle variations which don't affect meaning - a literate speaker won't be forced to eliminate sounds, but will pronounce written words as he has learned them. If I come across a word ending in -ism, I pronounce -izum even though the vowel between the two consonants is not written, its absence doesn't hinder me saying it.
|
|
| |
|
|
|
|
|
|
|
|
| the International Phonetic Alphabet says you need 120 sounds or so to render English (check that, Mabel) |
Well, the IPA chart on the Internet contains 75 consonant symbols in two groups, 60 pulmonic and 15 non-pulmonic, 28 vowel symbols plus 31 diacritical marks: that's 134 symbols for all the sounds so far detected in the world's languages, both lettered and unlettered. This of course includes many symbols for sounds that don't occur in any English dialect, including various clicks, voiced implosives, uvular fricatives and other earthly delights from the linguistic garden.
But I'm going to leave this before we get into a heated argument over how many phonemes can dance on the head of a pin. I just want to ask one question before I bugger off to another burning issue:
Have you, or has anyone else - say, an anthropologist or a missionary out studying the cultures of unlettered peoples or saving them from their heathen ways - ever actually heard these hundreds of different sounds in a single language?
|
|
| |
|
|
|
|
|
|
|
|
English, you say, has half a million words, as collected in the OED, and its European neighbours, something like 50,000 or 100,000 stretching things, which seems mysterious.
This, I think, is only a question of dictionary-making. The OED records a large number of archaic and obsolete words, and was besides the first of its genre.
Between 1951 and 1964 the Dictionnaire alphabétique et analogique de la langue française was published in six volumes by Robert. My Google searches haven't come up with the number of words or entries, but a one-volume abridgement (1967) lists 80,000 words.
The Grande dizionario della lingua italiana under the editorship of Salvatore Battaglia, 21 volumes, contains 183,594 "pure" words plus an unspecified number of variants.
The Grosses Wörterbuch der deutschen Sprache in an unspecified number of volumes contains 200,000 different headwords.
English still comes out fatter, but the OED was a forty-year-+ project, with another forty for the 2nd edition, with innumerable contributors. The French and Italian dictionaries are the work of one or two authors, plus their assistants of course, but not projects on the scale of the OED.
|
|
| |
|
|
|
Mick Harper
Site Admin
In: London
|
|
|
|
| But unlike you I don't place a negative value on artifice. |
I fear we are at cross-purposes. I am merely saying that a natural language arises in the absence of a single controlling intellect. This subtle but important difference came up once before with the origin of roads: one lot of roads, organic ones, are made quite naturally by people going the same way all the time and tending to make paths which others will use even if not particularly direct for their purposes. On the other hand trunk routes cannot arise this way and have to made by 'a single controlling intellect' (in some guise or other).
But notice modern languages might be greatly affected by a 'single controlling intellect', eg French under the Academic Francaise, German after the Lutheran Bible etc. However, primitive languages ie ones before literacy enters the picture must develop chaotically. But of course this does not necessarily mean 'with more complexity'.
| I can't know exactly how Long John will form the sounds you've represented by those letters. A Mummerset speaker will. |
This is the point I keep on making. The letters won't tell you unless you already know the language. Then you can readily guess.
| An alphabet is simply a convention, but a complex one, which never pretends to represent in a one to one fashion the sounds of speech |
Wrong! Every language with an accompanying alphabet in Classical times had a one-to-one letter to sound relationship. You know enough Latin to know this. How do you explain such a weird language? Are you honestly saying that a language can spring up perfectly naturally with just twenty-odd vowel and consonant sounds? At the very least (and it is a model worth thinking about) it requires an elite making a fetish of 'clipped' language and those conventions spreading through the general population.
But, as I also keep telling you, later the Irish discovered that, so long as you knew the language already and so long as you were prepared to learn various spelling conventions and loads of special cases, any language could be alphabetised.
| But I'm going to leave this before we get into a heated argument over how many phonemes can dance on the head of a pin |
This is not our argument but phoneticians'. Their specialism would scarcely exist if they had to admit that there is no scientific way of limiting the number of phonemes. Just look at a voice print and this will tell you that even now after centuries of standardisation around RP, there is an infinite number of pronunciations, each one unique to every person.
| Have you, or has anyone else - say, an anthropologist or a missionary out studying the cultures of unlettered peoples or saving them from their heathen ways - ever actually heard these hundreds of different sounds in a single language? |
I'd like the answer to that myself, though all anthropologists complain of the 'complexity' of primitive languages. However, they use the IPA same as everybody else does, otherwise they too could hardly get off the ground with their native dictionaries. One day I'll tell you the joke about the Sioux, their dictionary-compilers and my linguistic opponents.
Thanks for the indefatigable research into word numbers--very useful. My figures (forget where they came from originally) are clearly inconsistent with the THOBR argument. Your numbers are much better--ie English is oldest but everybody else has been going full bore for donkeys'.
|
|
| |
|
|
|
|
|
|
|
|
Okay, I get what I didn't get before in what you're saying. Just one thing: Robert Graves has a lot to say about early alphabets in The White Goddess: one salient point is that various of those he mentions, including the Latin, were composed of exactly 15 consonants and 5 vowels for arcane reasons having to do with tree names, associations of these with the months and various other symbolic things. This means that some of the letters had to represent more than one sound.
To take a few examples, Latin e and o each represent an open and a closed sound (scholars call them short and long, but in Italian it's open and closed, so I assume it was the same in Latin), and v was v and u. Whereas in classical Greek there were two e's and two o's, and of course there were digraphs, ou for u, eu for (I presume) -- and so forth. There must have been more doubling up according to necessity and the degree the alphabet represented the sounds of spoken language: I accept your notion that the Latin was unusually artificial, but its origin must have been in speech.
|
|
| |
|
|
|
Mick Harper
Site Admin
In: London
|
|
|
|
Ishmael stumbled on the sacred nature of alphabets by another route (without entirely proving his case). Robert Graves (from memory) was mostly dealing with northern alphabets, and the evidence from t'north seems to indicate that alphabets were only sacred there ie were used in, say, divination and tombstones but not for ordinary communication.
This might be because alphabets are a Druidic invention ie used by a memory-based intelligentsia for religious purposes, and were then nicked by 'Phoenician' (as they weren't then) tin-traders who turned them into something useful (as traders are wont to do). From there the idea of alphabets spread everywhere that the Druids did not control ie in the Mediterranean and then eastwards as far as India and northwards through Germany but not Britain and Gaul.
| I accept your notion that the Latin was unusually artificial, but its origin must have been in speech. |
This seems perverse and requires an argument.
|
|
| |
|
|
|
|
|
|
|
|
| Ishmael stumbled on the sacred nature of alphabets by another route (without entirely proving his case). |
Praps he should persist. I think he would also do well to elaborate the theory he outlines in the Ancient Islamic Empire thread, although one or two things need revision.
| Robert Graves (from memory) was mostly dealing with northern alphabets, and the evidence from t'north seems to indicate that alphabets were only sacred there ie were used in, say, divination and tombstones but not for ordinary communication. |
Having no other information about the use of Druidic alphabets, I'll take your word for it. But Graves also speaks of early Greek and Latin alphabets having 5 vowels and 15 consonants, numbers he ascribes to religious reasons. His understanding is that the same religion was practised throughout Europe and had close affinities with Asian religion.
| I accept your notion that the Latin [alphabet] was unusually artificial, but its origin must have been in speech. |
This seems perverse and requires an argument. |
I mean that if as you say the intention was practicality, even if the written language was simplified it would hardly have been practical had it required the learning of pronunciation, grammar and syntax far removed from the speech of the people who used the writing. Nor does a kind of Italic Esperanto seem practical.
|
|
| |
|
|
|
|
|
|
|
|
| Mick Harper wrote: | | This might be because alphabets are a Druidic invention ie used by a memory-based intelligentsia for religious purposes, and were then nicked by 'Phoenician' (as they weren't then) tin-traders who turned them into something useful (as traders are wont to do). From there the idea of alphabets spread everywhere that the Druids did not control ie in the Mediterranean and then eastwards as far as India and northwards through Germany but not Britain and Gaul. |
There was, it seems to me, no single point of origin for the "alphabetisation" process. The process, which is an attempt to encode sound into written signs rather than encoding concepts in signs (i.e. idiograms), has originated quite separately in different forms in many places.
The Japanese, the Koreans and the Thai invented syllabaries. The Egyptians and the Maya created glyphs which starting as pictograms subsequently converted into sound symbols. And even cuneiform came to be used in an "alphabetic" sort of way. Some folks seem convinced that the Dravidians had writing that did not owe anything to the later Aryans and so on. Even the Chinese invented an alphabet which they later suppressed/abandoned.
The invention of alphabetised writing is somewhat analagous to the invention of the wheel. It happened almost everywhere that had settled societies.
|
|
 |
|
|
|
Ishmael
In: Toronto
|
|
|
|
| berniegreen wrote: | | There was, it seems to me, no single point of origin for the "alphabetisation" process. |
There is always a single point of origin.
|
|
| |
|
|
|
|
|
|
|
|
| Ishmael wrote: | | berniegreen wrote: | | There was, it seems to me, no single point of origin for the "alphabetisation" process. |
There is always a single point of origin. |
No there is not.
And it is conspicuously the case in the field of concepts and ideas.
One classic example is enough to prove my point: the Theory of Evolution through Natural Selection was conceived quite separately by two men almost contemporaneously (i.e. Wallace and Darwin). Although the popular version of history may make it seem that "Darwinism" had a single point of origin, the truth is that there were two absolutely separate and independent points of origin. That they were subsequently merged, by courtesy of the good graces of Wallace is irrelevant to this issue.
|
|
| |
|
|
|
|
|
|
|
|
| Ishmael wrote: | | A friend of mine has learned to speak Thai. He was describing to me how the same word means three vastly different things when pronounced in ways that -- I swear to god -- sounded absolutely identical to my ear. |
Don't feel bad about this, Ishmael. You are merely suffering from one of the side effects of neuroplasticity. That is to say that, through habituation, we reinforce the recognition and use of the sounds of our own language and drop sounds that we don't use which are "irrelevant" to us.
For example, a Japanese child of, say, 4 years can hear and recognise the difference between "l" and "r" (as used in English speech). A Japanese adult cannot. An adult English speaker finds it very difficult to hear and correctly pronounce the consonantal sound represented by the Hiragana characters which are called in English "RA", "RI", "RU", "RE" and "RO", but which are normally pronounced at a midway point between "r", "l" and "d". But Australian primary school kids learning Japanese for the first time have no problems at all.
How would one go about alphabetizing such a language?
What effect would 500 years of alphabetization -- using a foreign character set derived from another spoken language no less -- have upon the spoken form of a language like Thai? |
But the Thais don't have a foreign character set. The Thais invented their own alphabet --- http://en.wikipedia.org/wiki/Thai_alphabet .
Your problem here is that you are operating with the false paradigm of "a single point of origin".
|
|
| |
|
|
|
|
The Library Index 
FAQ 
Search
Register
Profile 
Log in to check your private messages 
Log in 
