[YG Conlang Archives] > [ceqli group] > messages [Date Index] [Thread Index] >
On 5/13/05, Rex May - Baloo <rmay@hidden.email> wrote: > Hm. Just playing around with permissible initial clusters... > > As you said, Eo doesn't make a list anywhere that I know of, but there are > what seem to be rules. Kind of a 'general average european' system. There are implicit rules *now*, which can be deduced by a posteori analysis of the words actually used in Esperanto. The fact that there were no explicit rules up front made it easy for people to borrow root words that were hard for others to pronounce. > Making a list of initial consonant clusters is only half the story, because By initial consonant clusters I mean clusters of consonants (in the general phonological sense, not your special ceqli sense) that can begin a word. So if you have a morpheme "skraq", then /skr/ is the initial cluster, not /sk/. > while you can have sk-, it can be followed by any vowel or semivowel, r or > l, and then it gets iffy. I'd consider skq unpronounceable, but not > necessaryily skm or skn. Curious, that, since for clusters in the opposite order most people seem to find /Nk/ easier to pronounce than /nk/ or /mk/. I would disrecommend any initial or final cluster of three or more consonants in an IAL, and particularly these /skm/ and /snk/ possibilities. > One point is that, like Eo, Ceqli borrows almost > all words, so they are going to alreay fit somebody's phonotactics:). Yes, but I thought the point was to look for the subset of words in the source languages that fit ceqli's phonotactics. You can't do that if the phonotactics of ceqli are constantly mutating by influence of whatever languages you borrowed words from recently. > I > decided that some Russian roots were just too consonant rich for borrowing. > zdrav, chlen, etc. I don't see how initial /zdr/ is any harder to pronounce than /skr/. I would readily use either in a personal artlang, but neither in an IAL. > Anyhow, so far I'm just following my own instincts as to what things are > pronounceable. Again, it's sort of academic, because Ceqli isn't likely to > want to borrow a root from any language that has any words (ok, except > russian) that I'd consider beyond the pale. > > I consider Loglan's phonotactics to be flawed because it permits > monstrosities like 'sii' (syee) which are certain to devove into 'shee' in > my opinion. I don't know; if Loglan is ever used by a speaker community as a primary oral language and its phonology mutates away from the written standard, that kind of cluster is just as likely IMO to mutate into a palatal fricative (like German ich laut) as a postalveolar fricative (English "sh") Or the palatal approximant might disappear before /i/, leaving the alveolar fricative unaffected. But you're right that it doesn't seem like a very stable cluster. > I have my doubts about my own "xyen" (dog) and whether it's > different enough from "xen" Again, if ceqli is used by speakers whose phonology evolves away from the written standard, I would make no bets about which way /Sjen/ is going to evolve; /Sen/ and /Cen/ seem equally likely, and there are other possibilities. Would you like me to take a look at the ceqli word list and come up with a draft phonotactic rule set, which you can then shoot holes in and maybe use as a basis for your own rules? I'm not sure when I'll have time to do it, but it sounds interesting. --- Jim Henry http://www.pobox.com/~jimhenry