Semivocales

I am very sorry to bother all of you – especially Jorj and Simon – with this jeremiad. I just cannot help thinking all kind of odd things. Please, read and smile, don’t be angry! I am sorry also that I don’t write this in Elefen. English makes it easier to distinguish the meta language (Eng.) and the object language (elefen).

I’ve been telling my friends that LFN has the excellent property that the spelling and the pronunciation keep well together. No such ambiguous rules like for example in Interlingua. I have recently translated into Finnish the 1st chapter of the grammar. It opened my eyes: The consonants are all right easy and clear in LFN, but the vowel system is a bit complicated, especially for a person who is not used to Roman languages but still expects an easy language to study. I am maybe too late to give any new suggestions, but in any case I must write this for myself. In fact my suggestions don’t concern a big set of words.

I found two things problematic: 1) The rules of stressing 2) The double role of ‘I’ and ‘u’.

The first one I cannot help nor give an advice. For me whose own language keeps always the stress on the first syllable of the word, every rule (other than that of Finnish) is a bit difficult. Yes, Esperanto and Ido have easier rules for stressing than LFN, but the rules of LFN give a more beautiful sound of speech. It is possible – all right – to learn the stressing rules, but it needs plenty of motivation! In fact I found the stress rules quite coherent until I read the last rule telling that in case of AE, AO, EA, EO, OE, and UI the stress goes after the last consonant. Oh no! Is this really necessary?! I could very well stress also here along the basic rule. But I think we need this extra rule for making the language sound more “natural” and I do not protest.

The problem #2 is worse. In designing their vocabulary, most auxlangs try use words in such a form that they could be easily recognized. This criterium is not possible to fulfill without losses in coherency. I like the most of the compromises that LFN has made. Many words – like “accept” (>”aseta”) - have lost some of their familiarity, but gained simplicity and beauty. But: Avoiding to use ‘y’ and ‘w’ we have lost a lot – and completely in vain!

First time when I saw the word “ciui”, I was very much embarrassed. I could hardly have recognized it without reading the explanation. Still, every time I see that word, I feel uneasy. For Heaven’s sake, why not to write “ciwi”?! Other examples: what are “ianci”, “ien” and “uisce”?! Aha: they try to be the same as “yanci”, “yen” and “wisce”! Furthermore: The word “ierarcia” comes from the Latin word “hierarcia” by dropping the ‘h’. That is why I think that “ierarcia” should be pronounced [ierarkia], not [jerarkia]. Evidently the word “uigur” should not be pronounced [wigur]! I am not good in zoology, but I don’t think that the word “iena” (English: “hyena”) should be pronounced [jena].

If I had power over LFN, I should say: Let’s not use ‘i’ and ‘u’ as semivowels in the beginning of a word or a syllable. Use ‘y’ and ‘w’ instead! Everything will look more clear and international, and we have a tool to distinguish between [i] and [j], resp. [u] and [w] where needed.

The double role doesn’t bother in the end of diphthongs. Thus, in many cases, such that “paia” we can think ‘i’ as the second vowel in a diphthong “ai” rather than the first vowel in “ia”. In other words I prefer to think [paj-a] to [pa-ja]. But for example in the word “Iaue” (a name of God) we cannot (for etymological reasons) think [jaw-e] but [ja-we], therefore we should write “Yawe” instead of “Iaue”. The same with “ciui”: “iu” is not a diphthong, so let’s write “ciwi”!

If – a big IF - we want to make changes for these reasons, which words would change, which words would not?

i) When a semivowel begins a word, it should be marked by ‘y’ or ‘w”. There is a rather small set of such words.

ii) The clusters of three vowels with ‘I’ or ‘u’ in the middle would be untouched in the most cases. Only if the cluster could not be interpreted as a combination “diphthong+vowel” we would denote the middle vowel ‘y’ or ‘w’. These occasions are very rare.

Here is a list of most words that would change:

ia – ya (An extra advantage: “ya” doesn’t look like “la”.) iac – yac Iaman – Yaman ianci – yanci iard – yard iate – yate Iaue – Yawe ien - yen iodle – yodle ioga – yoga ioio – yoyo iudi – yudi uafel – wafel ualabi – walabi Ualonia – Walonia uaua – wawa ueb – web ueste – weste uiguam – wiguam uisce – wisce uolfram – wolfram Taiuan – Taiwan mazaua – mazawa alaui – alawi ciui – ciwi

For example following words would not change because they don’t begin – I think - with a semivowel:

iena, ier, ierarcia, ieroglifo, ion, ionosfera, uigur

harri ()

• Both your points are closely related: both revolve around semivowels and diphthongs.
• We say “aséntua” rather than “asentúa”, because “ua” is as a diphthong, a single syllable. In fact, the list of final vowel pairs that behave in this way is simply the list of rising diphthongs, phonetically [ja] [je] [jo] [ju] [wa] [we] [wo]. The rule in Elefen is that only I and U form diphthongs, and so “estráe” has three syllables and is stressed no differently from “estrádi”.
• “-ui” is the only exception: it could just as well be “pródui”, but we happened to prefer the sound of “prodúi”. The location of the stress is not particularly important in Elefen; the rules are actually just guidelines so that people can aim at the same thing. The dictionary says that “asi” is pronounced “así”, but it really doesn’t matter if you say “ási” instead.
• Regarding the spelling, why should [ja] be spelled “ya” in “ia” and “iac” but not in “belia” and “paia”? [ia] and [ja] sound extremely similar (as do [ua] and [wa], etc), and the only difference between them in Elefen is their effect on the location of the stress, which is not even very important. This spelling is, I believe, one of the oldest features of Elefen, and it’s certainly too late to change it now (especially as I’ve spent several months preparing the printed dictionary using it :) Most of the words you list are very obscure; I would say don’t worry about them. If “ioga” is less recognisable than “yoga”, so be it. You could say the same about “caxmir” and many other words. It is what it is, and we are where we are. Simon

Thank you for the comments! I understand very well that my suggestions come too late, but I want to explain some point: Semivowels are - that is how I feel - pretty much stronger in the beginning of a word or a syllable than in the end. For example in the word “auto” ‘u’ hardly differs from a common ‘u’. Finnish is rich in vowels, yet we don’t have the strong ‘w’. It is a bit difficult for us. We must make a little effort to learn it when studying the English “west”, “well” and “whisky”. So I feel that the too ‘u’:s are two separate phonemes. We have plenty of words that begin with an “ui” or “uo”, but there ‘u’ is not the same as ‘w’. The ‘j’ is very common both in the beginning of the word or somewhere in it. But our ‘j’ is clearly stronger than ‘i’ in the end of diphthongs. For example pronouncing the Finnish word “laita” my tongue keeps free at ‘i’, but in “laji” my tongue hits the upper gum at ‘j’. So I don’t feel very natural to denote these phonemes by the same letter! What I tell here about Finnish, is - I believe - true for most European languages. Test what feel with your tongue and lips when saying “May” and “yet” or “row” and “willow”! harri ()

A little note yet: You write: why should [ja] be spelled “ya” in “ia” and “iac” but not in “belia” and “paia”? I don’t think [bel-ja] but [bel-i-a]. Not [pa-ja] but [paj-a]. I wanted to see ‘y’ only in the beginning of a syllable, where it is strong. harri ()

• The suffix is “-ia”, so not really any different in pronunciation from the word “ia”. So I think [bel-ja]. Rules for syllabification would be more complicated to master than the consistent spelling with I and U. How do you feel about the U in “segue” and “suave”? Simon

According to the grammar, ‘i’ is pronounced as ‘j’ only 1) in the beginning of the word, 2) as the middle vowel in a vowel trio, and 3) as the end vowel in a diphthong. So I cannot think [bel-ja], but - as you say - the sound is very similar. I think “segue” like [se-gwe] and “suave” [su-a-ve], stresses on “se” and “a”. harri ()

• Actually the grammar also says that I is [j] when LI or NI are intervocalic, as in “folia” (or “belia”) and “anio”. I say both “segue” and “suave” as two syllables: [‘seg-we, ‘swa-ve]. Is the [w] in [se-gwe] easier for you to pronounce than the one in [we-ste]? Simon
• It sounds to me as though you perceive prevocalic “i-” at the start of a syllable as a fricative [ʝ], whereas I perceive it as an approximant [j]. Simon

Yes, I forgot the LI- NI- rules! There are in fact so many rules that the language looks difficult to a novice. (I see no motivation for this rule.) I lived a quite time with ido without really noticing the GU- rule there. I pronounced happily [líng-u-o] and were sorry that I must force my brain to [ling-wo]. No, I have no problems in pronouncing [we-ste] or [se-gwe], not anymore, but they are exotic for us Finns. The way a common Finn would use is [u-es-te] and [ling-u-o]. As said we have (almost) one-to-one congruence in spelling and pronunciation. (Elefen is far from it.) I think you are right. Maybe the Finnish ‘j’ is that you call fricative, and the Finnish ‘i’ as the second in a diphthong an approximant. I don’t know very well these terms. I called them strong ‘j’ and weak ‘j’. You must know that I like elefen very much and can stand the thing, that everything is not exactly how I wished. However: if I happen to write an essay where I mention Jahve or kiwi, I will spell them Yawe and ciwi, even if a thunderstorm would hit me. (haha) harri ()

• Incidentally, I detect very little difference between my pronunciation of “i” in “maia” and “ma ia”; both are [ma-ja] with an approximant, i.e. essentially a very brief instance of the vowel [i]. And in my pronunciation of “mais” and “auto”, I can briefly hear the second vowel with its usual sound. This is different from my pronunciation of English, where “mice” doesn’t sound quite the same as “mais”, and “ow!” doesn’t sound quite the same as “au!” (to a large extent, this is probably because the second element of these English diphthongs is lax: [ɪ, ʊ] rather than [i, u]). When I say “ia” at the start of a word, the [j] sound has no friction or hiss to it. This is what I mean when I say it’s an approximant rather than a fricative. “Approximant” is basically what a vowel counts as in consonant terminology! Simon
  • (I should have said: “approximant” is basically what a close vowel counts as in consonant terminology. Obviously [a, e, o] are not approximants (but [i, u, y, ɯ] are). Simon
• The motivation for the LI and NI rule is that the Romance languages tend to have special palatal versions of L and N (phonetic symbols ʎ and ɲ. The real question, in my mind, is not why “belia” is pronounced [belja] and “anio” [anjo], but rather why the rules don’t say that “famia” is [famja] and “adio” is [adjo]. Ultimately, though the difference in sound between [i, j, ʝ] is very small, and will typically be dwarfed in practice by other features of a given speaker’s accent. Not worth worrying about! Simon

Alo, Harri! Your comments are very interesting. Part of what you suggest is “taken care of” by the fact that pronunciation in elefen is very forgiving. The rules are there as guides, not as commandments. One may say “álo” o “aló”, and no one will misunderstand. The same applies to most of the examples you list. However, in order to guide people a little better, I think that (if Simon agrees) we might add :P (“preferred pronunciations”) to the dictionary for the words iëna, iër, iödo, iön, iönosfera, Iönia, iönica, iöta, and uïgur. I am less certain about ierarcia, ieroglifo, ieratica. You are quite right that elefen might have been a bit clearer if we had permitted w and y, but we didn’t, and it is a bit late in the game for that kind of change. But future elefenists might decide to change things a bit - just not till Simon and me are in our graves! jorj

Another: Cuait /ku-ait/ jorj

• “Uigur” is already marked as “uïgur”, but I’ll add the others. Is “ier” really two syllables, though? It comes from the French “hier”, which is pronounced [jɛʀ]. For the same reason, I share your uncertainty about “ierarcia”, “ieroglifo” and “ieratica”. Harri, you’ll be amused to know that I’ve used the letters “j” and “w” for clarity in a few of the pronunciation indicators already in the dictionary (e.g. “chihuahua” [txiwawa] and “Utah” [yuta]). Simon
  • pardona: me ia era con “ier”. esce lo es posible ce la pronunsia ia cambia? me ia studia franses sirca 50 anios en la pasada :o) jorj
    • Cisa “hier” ia cambia alga sua pronunsia, ma me trova [jɛʀ] en un disionario franses de 1983. Simon
    • Wiktionnaire dise ce on pote pronunsia “hier” como \jɛʁ\ o \i.jɛʁ\. Angel
      • a! merci e grasias, angel! lo es sempre felis cuando on descovre ce on no es completa senil. :o) jorj

Thank you, Simon and Jorj, once more for the interesting and illuminating comments. I hope the dictionary project goes well. I’ll go on with mine, i.e. the translating the grammar into Finnish. harri ()

Marvelous! jorj