Modest Proposals for spelling reform

[King John]

For some time now, I’ve been working on a set of proposals – Modest Proposals, one might say – for reforming the spelling of our beloved language. Many thanks are due to several Ladíntschen who let me bounce ideas off their often greater knowledge and insights, in particular Sir Tomás Gariçéir, S:reu Cresti Siervicül, and Barôn Hooligan. I have no reason to think these gentlemen will agree with everything here, and whatever mistakes I’ve made are mine and not theirs, but these Proposals would be much worse had I not had their input.

As we discuss these Proposals – all of us, whatever our politics or citizenship, who love the Talossan language –, I would ask that my thoughts and arguments be treated merely as those of a private person. When you see mistakes here, point them out! If I argue, and you’re not convinced – argue back!

My object in these Proposals is threefold – to get rid of some exceptionally ugly letter combinations; more importantly, to make the Talossan language easier to type and read and learn by reducing the huge incidence of odd, unnecessary, and often meaningless diacritical marks; and most importantly, to make it possible, for the first time in the history of the Glheþ, to look at a written Talossan word and be able to tell, unambiguously, how it should be pronounced ¬– including on what syllable the stress should fall.

These Proposals, if adopted, would affect the spoken language – pronunciation and vocabulary – very little. That is on purpose. I have plenty of thoughts to offer on both subjects, but for now, let’s just concentrate on getting the writing system fixed, and not on changing the language itself.



A. ORTHOGRAPHIC REFORM. There are a lot of very quirky things in Talossan orthography that have been suggested, at one time or another, as things to change. (The digraph tg to represent th-as-in-thin seems to be one of the least popular.) Sch for sh-as-in-fish, th and dh for th-as-in-that, mh and bh for v, vh for w, fh for h, the letter ç – these have all been suggested for correction or elimination. But most of those corrections and suppressions are excluded from this list. My main goals here have been to correct the vowel-system, which is both the worst-broke and the most annoying thing about Talossan orthography, and to get rid of some least used and most awkward multi-character sequences. NOTE that proposals A2, A3, and A4 taken together would reduce the number of vowels in Talossan from 13 to a much more realistic 8, one written vowel remaining for each vocalic phoneme. (By comparison, Spanish and Italian have 5, French 8 or 9 depending on dialect, Latin 7 or 8 depending on what linguist you listen to, and so on.)

My fear is that a wholesale reform of the consonants would do one bad and one very bad thing – it would obscure certain historical remnants of the development of the language that I’d like to see use keep, and (much worse) it would make Talossan look a lot different than it does now. We want to make the language easier to use; we don’t, I think, want to make it to look like a different language.


A1. Get rid of the horrible silent -ë on the infinitives of irregular verbs (and the noun pëvarë). A spelling feature that serves only to “remind” one that the verb is irregular is simply absurd; no native speaker would even think of a verb’s being irregular. “Irregular verb” is a concept used by linguists and teachers and inventors of languages, not by real speakers and writers.

A2. Use ö for ø. They represent exactly the same sound, and ø is found in only two words.

A3. Use ou for oû. There is no reason whatsoever to keep ou and oû separate.

A4. Use a single letter for each set of vocalic allophones.

A4a. Use a for all instances of a, â, and å. An unaccented a, in many languages (including Talossan), can be pronounced as a schwa. No need to write it with a different letter â. There are no minimal pairs among these letters except ár (our) and år (year) – both of which non-Romance words should (I think) be deprecated, in favour of the far better noastra and anneu.

A4b. Use e for all instances of e and ë. Same reasoning as with a and â. Careful examination of the vocabulary leads me to believe, more and more firmly the more I study the question, that ë is not a phoneme in Talossan. It appears most frequently before liquids in words simply transliterated from English (ërël, piutër, xhërm) or in verbs that end -ëschar. Sometimes it appears in places where it’s hard to believe anyone pronounces it as a schwa – for instance, in pësc or sërxhënt. In no case does a minimal pair exist between e and ë, except se (the reflexive pronoun) and së (the possessive marker); which is the sort of exception that proves the rule, since së might very well be analyzed as a particle (or case ending) attached to the preceding noun.

A4b1. EXCEPTION. But abstract nouns that end in -ië, of which there are several, change to -ia instead of -ie.

A4c. Use i for all instances of i and î. (Does anyone actually even know what sound î was supposed to represent?) The only minimal pair in the Treisoûr between i and î is pint (point) and pînt (pinto bean), which I find less than conclusive, especially as the Treisoûr also includes several double-entries of the same word being spelled with i or î indifferently. NOTE that this change will require changing the spelling of words that now use the combination cî (pronounced ki) to use chi instead.

A4d. Use u for all instances of u and û. The letter û, while apparently fairly common, actually isn’t. It appears in the (unnecessary) oû combination, in the prefix ûn-, in a group of words beginning dûc- where it’s already pronounced like u, and in only about 15 other words, most of them taken straight from English (rûm, fûn, etc.). In a real language, especially in a Romance language, this un-Romance sound would probably have been completely assimiliated to u or a. There are no minimal pairs.

A5. Replace the tetragraph s-ch (prounounced as in fish church) with schcia, schce, schci, schcio, or schciu, or (at the end of a word – and yes, there are two of these in the Treisoûr) schtsch. S-ch is a grossly unnatural spelling, hard to remember, and is ambiguous (unless we want to make a ridiculous rule that you can’t hyphenate a word ending with s together with another word starting with ch).

A6. Replace the tetragraph s'ch with sq (before e or i) or (otherwise) with sc. The rule on pronouncing q would now say that q is pronounced k after s or before u, and otherwise ky. (There are only five words now that are spelled with sqe or sqi, and none of those should really have that y sound after the q.) Spanish uses sque or squi to represent the sounds ske or ski; Talossan pronounces qu as kw, so rather than sqeu and squi, we can use sqe and sqi to gracefully represent the same sounds.

A7. Replace çh with c'h. Exceptions: in parpaçhar, praisaçhar, and isnaçhoçhâ (the first two from Latin g, the third from Greek gamma, replace çh with g. It’s a very rare sound, occurring in fewer than 20 words, and would (in a natural language) be quickly assimilated to a normal native sound, like our c’h or g.

A8. Recognize j (in all the pronunciation-guides) as a full-blooded Talossan letter, a variant spelling of i when i is a consonant, pronounced as consonantal y. J is actually fairly common in Talossan.

A9. Note that k, w, and y are rare letters occurring most often in words borrowed from foreign languages, but drop the silliness about their being pronounced as in the language of origin. (K and w are pronounced as in English; y as an English consonantal y or a long e.) What native speaker knows or cares what language a word came from, or how it was pronounced? Borrowed words are pronounced using the phonetic set of the borrowing language.

A10. är is unnatural to pronounce except before a vowel. So ...

A10a. Modify all uses of ärC (where C is some consonant) to arC. That’s how anyone would pronounce it anyway, so we might as well spell it that way.

A10b. Modify all words ending in är to end in aer. (Notably är, air, would become aer. This is a conservative change, since most words from Latin aer are already spelled with aer in Talossan.)

A11. When tg represents the sounds of t and g, rather than þ – a very rare occurrence –, make it t’g instead. So avint’guard, rönt’gen instead of the current (and ambiguous) avîntguárd, röntgen.

A12. Talossan has exactly three words in which rh is pronounced r rather than sh – gavarhál (snipe), rhesus (rhesus), and perheliôn (perihelion). Change these to gavarál, resus, and pereliun (after the pattern of apeliun = aphelion), and we’ll be able to say that rh is always pronouced sh. (Thanks to Lord Hooligan and S:reu Siervicül for seeing how to simplify my original proposal.)

A13. Replace the awkward trigraph gñh with nh, which makes exactly the same “ny as in canyon” sound. Or perhaps, if this is regarded as too far-reaching a proposal, we should simply declare nh to be an authorized (and even favoured?) alternate spelling for gñh, rather as tg is for þ or ss for ß.

A14. Replace the difficult trigraph glh with lh, which makes exactly the same “lli as in million” sound. Or again, if this seems too big a change, we can make lh an authorized (and favoured?) alternate spelling for glh.



B. MARKING STRESS. Possibly the worst drawback to the current Talossan writing system is the inability in many words to mark stress. (And another is the unnecessary marking of regular stresses in lots of words, based on the fact that the default position of stress in Talossan was misunderstood early on in the formation of grammars and vocabulary lists.) The following changes are designed to simplify and greatly reduce the use of diacriticals in Talossan, and also to make the stress-marking perfectly reflect the actual pronunciation of the language.


B1. First, this is the (new) "Default" Rule of Stress: Any word of more than one syllable is stressed on the vowel or diphthong before the last consonant in the word; except that the consonants in the endings -as, -en, -ent, -eux, -ic, -ici, -ilor, -laiset, and -mint don't count. (These are the “always-unstressed” suffixes.)

B2. If the actual stress doesn't follow the Default Rule, it has to be marked.

B2a. If ä or ö or ü has to be stressed, mark it by changing the diacritical to a circumflex (â or ô or û).

B2b. The other five vowels (a, e, i, o, and u) have their stress marked, if they need it, with an acute accent (‘). We should note that some writers and typesetters prefer the grave accent (`) for marking stress, especially at the end of a word. This is perfectly correct; the choice is a matter of personal style, or even handwriting.

B2c. SLIGHT EXCEPTION. There are four two-syllable words – aici, bici, mici, and pici (horses, nibs, crumbs, and tops) – which can’t follow the Default Rule because they’re too short to have a vowel before the last consonant before the ending -ici. So any two-syllable word ending in -ici (or -ilor or -laiset – but there are currently none of these) is stressed on the first syllable, which remains unmarked.

B2d. ANOTHER SLIGHT EXCEPTION. A third-person singular future verb (for instance, o lirarha, he will read) can (optionally) be stressed on its last syllable. Except for the irregular future será, this optional last-syllable stress is never marked.

B2e. EVEN SLIGHTER EXCEPTION. Even though acest can optionally be stressed on the first syllable, it is always written without a stress mark.

B2e. SORT-OF-SOMETIMES-EXCEPTION. When words are hyphenated together, each of the joined words is stressed and spelled without regard to the other(s). So, for instance, when pronouns and verbs are joined in inverted order with a hyphen (as in parta-t-o or possadra-t-a or pensetz-voi), don’t write parta-t-ó, to indicate the stress on the pronoun, as if the combination were a multisyllabic word.

B3. If the actual stress does follow the Default Rule, it should not be marked. This is true even if some other form of the same word does mark it; for instance, apricó (apricot) in the plural is apricons, without the mark.

B3a. EXCEPTION. Even though there’s no need to mark the stress on a one-syllable word, a stress mark is still used on some one-syllable words to avoid confusion with other words pronounced and otherwise spelled the same. There’s no need to continue to mark és or è or frù or tú, because there are no words es or e or fru or tu. But a = she and á (or à) = to; la = the, lá = there; etc. (Even though there is no word mar, már will keep its stress marker so as not to be pronounced mash – see point B3b.)

B3b. ANOTHER EXCEPTION. We need to be able to distinguish between the pronunciation of a verb infinitive ending in -ar (which is pronounced ash), and a non-verb ending in a -ar (pronounced ar). Talossan already marks almost all these non-verbs with a stress marker -ár; I propose that we finish the job, and establish the rule that any non-verb ending in a stressed -ar must have a stress marked (-ár) in order to have its last letter pronounced r. (And if it’s really stressed on some other syllable, both stress marks will be written, one to show the stress and one to show that the r is pronounced r. There are only five such words -- ceáiçár, Máxhár, ísobár, Stáiár, and xhágár.) For example, Noemvár or barbár (barbarian), which without the stress-mark would be pronounced Noemvash or barbash. The change is minimal, affecting only 18 words; the only common one is the preposition par, “by means of”, which I would suggest we simply add to the list of exceptionally-pronounced words rather than spelling it pár. NOTE that a few non-verbs are formed from verbs in such a way that they should be pronounced with the sh sound – sa’star (maybe) and pevar (power), for instance. These take no stress marker. BUT we then need an additional rule that the plural of such a noun, formed with the -en suffix, will also require that the r be changed to rh to maintain the sh sound. So the plural of pevar is pevarhen, not pevaren.)

B4. If a word has more than one of the “don’t-count” endings, stacked up, don’t count the last one, but do count the others. For instance, ünic is accented on the first syllable, which doesn’t need to be marked because the -ic is a don’t-count ending; but ûnicmint has to have the first-syllable stress marked.

B5. Stop marking ïa thus to indicate it’s not the diphthong ia. If the i is stressed, make it ía; if unstressed, it needn't be distinguished from the diphthong, so write it ia (or ja). I believe there will no longer be any words in which ï is required for correct pronunciation; arc’haïsmeu will be written archaísmeu, mhïus as mhíus, etc. If there turn out to be any where í won’t do the trick, we can go on using ï.

B6. Go on using éu to distinguish it from the diphthong eu.

B7. Do not mark more than one stress in a word, except when you have to write ía or éu to distinguish the double-vowel from the diphthong, or when you write -ár to suppress the pronunciation of r as sh (see B3b), and the real stress lies elsewhere in the word, in which case it has to be marked. As an example of this last case, consider mhodernistic (no stress mark needed) and pséudumhodernístic (where the é forces marking of the real stress on the penultimate).

Note that a few words which used to be spelled without an accent-mark now need one: for instance, cióvec (man), política (policy), aicí (here), éu (I).


As an illustration of the kind of impact implied by the adoption of these proposals, consider the following passage from Sir Tomás Gariçéir’s Introduction to the Treisoûr, first given in the old spelling, and then in the new:

Másmînt ça mîmtenença viensâ ziferençù simpil más mült împîrtînt da toct i öðreux glheþen del mundeu: c’è ünicmînt és solamînt Talossán. C’è el soleu glheþ del mundeu entiéir qi încorpora la cúlturâ és el pînt da vhischtâ ünici da Talossa. Acest glheþ isch Talossán în ün vej în qët aucün altreu glheþ non pût estarë txamáis. És els xhînts qi se ladintschiçent, qi aprendent, parlent, és scrivent el Talossán, os zeviennent vrätsmînt Talossáes în ün vej în qët aucüns altreux xhînts non povent estarë txamáis. Ça sembla, që dels Talossáes començent finálmînt à realiçar acest. Os ziscovrent, që el glheþ Talossán lor zona ’n experiençù ünicmînt Talossán dal vidâ, és ünâ aválità ünicmînt Talossán d’en exprimar. La Comità për l’Útzil del Glheþ tigñhova ’n vräts exploçaziun dels mîmbreux dîn acest år, és për el pirmalaiset fäts dîn ár þistôriâ, ünâ grupâ dels Talossáes se viensiçeva és cuntravaleva súper el glheþ, për solvar dals fragâs dal speliçaziun és dal gramáicâ, és për cunmetrar és preparar acest zictziunár.

Masmint ça mimtenença viensa ziferençú simpil mas mült impirtint da toct i öðreux lheþen del mundeu: c’e ûnicmint es solamint Talossan. C’e el soleu lheþ del mundeu entieir qi incorpora la cúltura es el pint da vhischta ünici da Talossa. Acest lheþ isch Talossan in ün vej in qet aucün altreu lheþ non put estar txamais. Es els xhints qi se ladintschiçent, qi aprendent, parlent, es scrivent el Talossan, os zeviennent vrätsmint Talossaes in ün vej in qet aucüns altreux xhints non povent estar txamais. Ça sembla, qe dels Talossaes començent finalmint á realiçar acest. Os ziscovrent, qe el lheþ Talossan lor zona ’n experiençú ûnicmint Talossan dal vida, es üna avalitá ûnicmint Talossan d’en exprimar. La Comitá per l’Útzil del Lheþ tinhova ’n vräts exploçaziun dels mimbreux din acest ar, es per el pirmalaiset fäts din ár þistoria, üna grupa dels Talossaes se viensiçeva es cuntravaleva súper el lheþ, per solvar dals fragas dal speliçaziun es dal gramaica, es per cunmetrar es preparar acest zictziunár.

Still quirky, still Talossan, but much easier to type.



C. EXCEPTIONS

To complete the task of predictable pronunciation, we really need a complete list of words that are not pronounced the way the pronunciation tables would lead you to think. Here are the ones I know about:

acestilor (those), pronounced as if written “acéstscheler”
ir (to go), pronounced as if written “íar”
lhor (them), pronounced as if written “thor” or “ðor”
lo (it), pronounced as if written “lu”
o (he), pronounced as if written “u”
os (they), pronounced as if written “usch”
sieu (his, her), pronounced as if written “schu”
sieux (plural of sieu), pronounced as if written “schusch”
so (impersonal objective pronoun), pronounced as if written “scho”
thoct (an older form of toct, all, still used in certain phrases), pronounced as if written “hoct”
tir (to have), pronounced as if written “tíar”

Words spelled with a (formerly â) in which the a is pronounced as if it were i (formerly î):
portrateu
quand
quant
revand
Romania (first a only)
Romaniesc
sanc (and any word with the prefix sanc-)
sigñhificand
trans

Words spelled with g but pronounced as if it were xh:
ageu, -x
legeu, -x
regeu, -x
regipäts, -ilor
vice-regeu, -x

Query: Just how is gothique supposed to be pronounced? Surely not goðikwe?
Query: Shouldn’t “Lesotho” be written “Lesuto”?
Query: Should “Lutheran” be written “Luteran” or “Lutgeran”? Or is it really pronounced with a ð sound?
Query: Is the v in vräts silent? The Scúrzniâ Gramáticâ (section 17.33) transliterates ¿non c'è vräts? as “[nonCeræc]”. Is this a mistake, an exceptional pronunciation, or an unknown rule at work?


Respectfully submitted,

– John Woolley

[Sir C. M. Siervicül]

Wail wi ar taking ûbäut linggwîstîk riforms, Ai riûlaizd thät Tûlos'n Inglîsh cood yuz û lat ûv w'rk, tu.

I'm somewhat-more-than-half kidding. My serious point is that we should take Talossan seriously as a language, and tread very very lightly in making "reforms." If we would hesitate to "fix" the many aspects of Talossan English that are incredibly inconsistent and confusing, we should be similary hesitant to "fix" the confusing aspects of our other national language. I agree in whole or part with many of the proposals discussed here, but want to encourage everyone to take a critical and conservative view of any attempt to change a living language - Talossan or any other.

I'm going to limit myself to replying to section A of the original list of proposals, for the moment. That will make for more than adequately lengthy post.

john said:

A1. Get rid of the horrible silent -ë on the infinitives of irregular verbs (and the noun pëvarë). A spelling feature that serves only to “remind” one that the verb is irregular is simply absurd; no native speaker would even think of a verb’s being irregular. “Irregular verb” is a concept used by linguists and teachers and inventors of languages, not by real speakers and writers.

I would like to hear the thoughts of others on this, but I don't see a problem with this proposal.

A2. Use ö for ø. They represent exactly the same sound, and ø is found in only two words.

Only two words, but very historic ones (particularly støtanneu). I agree in principle with this change, but let's not make it an absolute rule. Rather, we should simply recognize that both letters represent the same sound, and that ö is far more common. I would use ö by default in the dictionary (including for stötanneu and löngeu), and describe ø in the pronunciation guide as a "variant form of ö, historically used most commonly in the words støtanneu and løngeu."

A3. Use ou for oû. There is no reason whatsoever to keep ou and oû separate.

I have mixed feelings about this proposal. The only reason to keep ou and oû separate is historic, but that's a big one. The -oûr ending is common and a fairly distinctive feature of written Talossan.

A4a. Use a for all instances of a, â, and å. An unaccented a, in many languages (including Talossan), can be pronounced as a schwa. No need to write it with a different letter â. There are no minimal pairs among these letters except ár (our) and år (year) – both of which non-Romance words should (I think) be deprecated, in favour of the far better noastra and anneu.

I had a very hard time with this, but I think I agree. Final â is also very distinctive in Talossan, but in this case I think the benefits may outweigh the loss.

A4b. Use e for all instances of e and ë. Same reasoning as with a and â. Careful examination of the vocabulary leads me to believe, more and more firmly the more I study the question, that ë is not a phoneme in Talossan. It appears most frequently before liquids in words simply transliterated from English (ërël, piutër, xhërm) or in verbs that end -ëschar. Sometimes it appears in places where it’s hard to believe anyone pronounces it as a schwa – for instance, in pësc or sërxhënt. In no case does a minimal pair exist between e and ë, except se (the reflexive pronoun) and së (the possessive marker); which is the sort of exception that proves the rule, since së might very well be analyzed as a particle (or case ending) attached to the preceding noun.

I'm unsure about this. Dropping ë would change the pronunciation of a LOT of words, and also take a LOT of accents out of the written language. That's a big change, and I'm not sure it's really necessary. Maybe you could analyze së as a particle attached to the preceding noun (and by the same token you could analyze se as a particle attached to the following verb, like in Spanish), but they're still different morphemes, and the difference still seems phonemic to me.

A4c. Use i for all instances of i and î. (Does anyone actually even know what sound î was supposed to represent?) The only minimal pair in the Treisoûr between i and î is pint (point) and pînt (pinto bean), which I find less than conclusive, especially as the Treisoûr also includes several double-entries of the same word being spelled with i or î indifferently. NOTE that this change will require changing the spelling of words that now use the combination cî (pronounced ki) to use chi instead.

I would have to agree that the sound of î as Ben originally imagined it is dead in spoken Talossan. I would argue, however, for maintaining î in some of the cases where it's most commonly used, like the prefix în- and the suffixes -mînt and -înd. The former two are never stressed, and the last is always stressed, so the goals of predictable pronunciation and stress would not be affected, and the sheer number of words involved (literally thousands) justifies the exceptions. There would still be some degree of inconsistency, but I guess you could say I don't care enough about making the language thoroughly consistent.

A4d. Use u for all instances of u and û. The letter û, while apparently fairly common, actually isn’t. It appears in the (unnecessary) oû combination, in the prefix ûn-, in a group of words beginning dûc- where it’s already pronounced like u, and in only about 15 other words, most of them taken straight from English (rûm, fûn, etc.). In a real language, especially in a Romance language, this un-Romance sound would probably have been completely assimiliated to u or a. There are no minimal pairs.

My comments about A3 and A4c are applicable here. My counterproposal is correspondingly similar - let's keep ûn- and -oûr and change û to u in the few oddball words. Again, the exceptions shouldn't be too much of a burden to remember, don't threaten the goal of predictable pronunciation and stress, and are justified by the number of words affected (over 800)

A5. Replace the tetragraph s-ch (prounounced as in fish church) with schcia, schce, schci, schcio, or schciu, or (at the end of a word – and yes, there are two of these in the Treisoûr) schtsch. S-ch is a grossly unnatural spelling, hard to remember, and is ambiguous (unless we want to make a ridiculous rule that you can’t hyphenate a word ending with s together with another word starting with ch).

Reluctantly, I agree that this proposal is a good one. In earlier discussions, I had suggested shc instead of schc, but sch is far more common than sh, and shc would be easily confused with sch.

A6. Replace the tetragraph s'ch with sq (before e or i) or (otherwise) with sc. The rule on pronouncing q would now say that q is pronounced k after s or before u, and otherwise ky. (There are only five words now that are spelled with sqe or sqi, and none of those should really have that y sound after the q.) Spanish uses sque or squi to represent the sounds ske or ski; Talossan pronounces qu as kw, so rather than sqeu and squi, we can use sqe and sqi to gracefully represent the same sounds.

I don't like this. It requires changing the spelling of dozens of words and further complicating the pronunciation rules for the letter Q. I also happen to like s'ch. Changing it to sq in many cases will obscure the etymology of words that currently have s'ch, like es’cheçâ (chess), which presently resembles many of its cognates, but no cognate that I can find has a Q in it.

A7. Replace çh with c'h. Exceptions: in parpaçhar, praisaçhar, and isnaçhoçhâ (the first two from Latin g, the third from Greek gamma, replace çh with g. It’s a very rare sound, occurring in fewer than 20 words, and would (in a natural language) be quickly assimilated to a normal native sound, like our c’h or g.

I oppose this proposal, too. I'm quite fond of this sound, actually, and think it's tenable. What's the rationale for the pronunciation of Afghanistan becoming Afc'hanistán? All the cognates that I can find either preserve the gh or simplify it to g. Our mythical Berber forebears would have no problem with this sound, and neither should we. This sound or a sound extremely close to it (voiced velar fricative or voiced velar approximant) actually exists in related languages, like Gascon (fricative) and Spanish (approximant). It does exist in too few words at present. Alternative solution - let's expand the number of words using it! Here's a start: aarçh! - a generic piratical interjection.

A8. Recognize j (in all the pronunciation-guides) as a full-blooded Talossan letter, a variant spelling of i when i is a consonant, pronounced as consonantal y. J is actually fairly common in Talossan.

Agreed.

A9. Note that k, w, and y are rare letters occurring most often in words borrowed from foreign languages, but drop the silliness about their being pronounced as in the language of origin. (K and w are pronounced as in English; y as an English consonantal y or a long e.) What native speaker knows or cares what language a word came from, or how it was pronounced? Borrowed words are pronounced using the phonetic set of the borrowing language.

Agreed, with some wiggle room (doesn't have to be explicit in our pronunciation guide) for individual speakers and recent borrowings. Lots of English speakers have a few phonemes that are only used in borrowed words or names.

A10. är is unnatural to pronounce except before a vowel. So ...

A10a. Modify all uses of ärC (where C is some consonant) to arC. That’s how anyone would pronounce it anyway, so we might as well spell it that way.

A10b. Modify all words ending in är to end in aer. (Notably är, air, would become aer. This is a conservative change, since most words from Latin aer are already spelled with aer in Talossan.)

A more conservative approach would be to permit an alternative pronunciation of ä when it's followed by a consonant cluster that makes it difficult to pronounce as æ. German does something kinda sorta similar with ä. If we're not eliminating ä altogether, this wouldn't hurt anything. The orthography of the language ought to be more conservative than the pronunciation.

A11. When tg represents the sounds of t and g, rather than þ – a very rare occurrence –, make it t’g instead. So avint’guard, rönt’gen instead of the current (and ambiguous) avîntguárd, röntgen.

This makes perfect sense. And if tg representing two separate sounds can become t'g, s'ch representing two separate sounds can stay s'ch (A6).

A12. Talossan has exactly one word in which arh is pronounced ar rather than ash – gavarhál, snipe. Change the rh to r (gavarál), and then we’ll be able to say that arh is always pronouched ash.

This has been changed since I started typing up my reply, but this change doesn't really seem necessary to me. "Rh is sh only in verb infixes" seems like an easy enough exception to remember. Why change the spelling of a word to change the exception into a general rule that doesn't apply to any more words than the exception did?

If limiting the pronunciation to only the verb infix is irrational, let's just make the pronunciation consistently "sh" and be done with it. It makes a fair amount of linguistic sense. The h serves to devoice the r in the verb infix. Devoiced r and sh are similar enough that the sounds merged. And the same mechanism would work for the handful of other rh words. The rh in rhesus was devoiced in the original Greek. And the r in perheliôn could have similarly devoiced under the influence of the following h. Gavarhál, being a rarely-used word, assimilated to the pronunciation of the thousands of other rh forms. In summary (and assuming that the status quo needs to be changed at all) we should view "rh" as originally representing a voiceless alveolar-to-retroflex trill or approximant, which eventually merged with the voiceless postalveolar fricative (sh). Heck, we could even allow speakers to optionally maintain a distinction between the sounds, which would allow the verb pronunciation to make more sense for some people.

A13. Replace the awkward trigraph gñh with nh, which makes exactly the same “ny as in canyon” sound. Or perhaps, if this is regarded as too far-reaching a proposal, we should simply declare nh to be an authorized (and even favoured?) alternate spelling for gñh, rather as tg is for þ or ss for ß.

For me, awkward spellings are more of a feature of Talossan than a bug, and should only be messed with to resolve contradictions and inconsistencies. I don't think that applies here, so I don't think this needs to be changed.

A14. Replace the difficult trigraph glh with lh, which makes exactly the same “lli as in million” sound. Or again, if this seems too big a change, we can make lh an authorized (and favoured?) alternate spelling for glh.

S-ch is to be replaced by schtsch (a heptagraph), and yet a humble trigraph is labelled "difficult"? For shame, your Majesty! My point above applies equally to this one.

This will probably be a Shibboleth for me: anything that changes the spelling of the word "glhetg," I oppose. "If you don't type 'glhetg', you're not using the glhetg!"

[txaglh]

hey, good to see, that in Kingdom there is also a will to do something with the horrible over-diacriticised, historical Talossan spelling. i will carefully read it and compare to the reform we do plan. more comments later.

definitely, this topic is a reason, why we should have Istitút del Glheth Pan-Talossán, IMHO.

[Hooligan]

Unlike the first two posters in this thread, I hope to make what I have to say on this subject short. However, just like them, I bet I'll fail. But wish me luck; I'll go as quick as I can.

A1. Remove the (silent) -ë from the infinitive form of irregular verbs. I agree it shouldn't be there. To me it is obviously some remnant of the fact that all verbs, not just those that conjugate irregularly, must have had something after the r. Given the fact that the infinitive is pronounced as English "-osh", I bet that ë was once an "h" (see A12). Ben was simply transcribing from the works of a Taloçait whose handwriting was sloppy. He obviously transcribed the regular verbs first, and, unable to tell what the final -H was in the scrawl he had before him, he left it out. He transcribed the irregular verbs later, from the works of a Berber scribe with only slightly better handwriting than the first, and the verb-ending H's looked like ë's to Ben. (You really can't blame him; it seems to have been an all too common trait in Berber scribes to get careless with the pen at the ends of their words.)

My opinion: I agree, remove the -ë. However, as an option (though one that I alone may actually think should be preferred), if we truly want spelling to reflect pronunciation, and aid the new student of the language, restore -- as an option -- the long-lost Berber H, on all verbs, regular and irregular. If we say "-osh", we should at least allow the word to be spelled "-arh".

A2. Use ö for ø. I agree. The two are the same letter. So let either be used whenever the other is used.

My opinion: Prefer ö over ø, but, as an option, allow the use of either one at any time, just as tg can be used at any time in place of þ.

A3. Use ou for oû. I had occasion to speak with Justice Siervicül about his post, and although he agrees with the vast majority of the proposals here, he feels that providing some small and easy-to-understand exceptions to them would allow Talossan to suffer fewer changes to orthography and yet reap the full benefits of all the reforms being suggested. In this case, I agree with him.

My opinion: Per A4d (which I will discuss below), say that "û" can be replaced in all current spellings with "u". This (as suggested in proposal B2a) allows "û" to represent the stressed pronunciation of "ü". HOWEVER, make an exception to the stress rule, specifying that when "û" appears in two specific "ancient" affixes (namely the prefix "ûn-" and the suffix "-oûr"), it does not represent a stressed pronunciation, but is purely decorative. Note that the use of "û" in these affixes would not be mandatory, but would be allowed as an option. The use of "û" in the affixes may die over time in preference to "u", but this small, easy-to-understand exception to proposal B2a will allow Talossan to preserve more of its appearance, without impacting pronunciation.

A4a. Use a for all instances of a, â, and å. I agree. Even though the Cosâ would become the Cosa, I agree with both posters that the â adds nothing.

My opinion: Yes, change all letters "â" and "å" to the unaccented "a". As for the minimal pair mentioned by the first poster, rather than mandate "noastra," leave "ár" (yes, with the accent, per the proposed rule B3a) for English "our" and use the unaccented "ar" for year (or the perfectly valid "anneu" form of that word, if you are of a more Romantic bent). I say to King John: Nice try, but we will not give up ár "ár".

A4b. Use e for all instances of "e" and "ë".

My opinion: I lean toward this one, but I reserve final judgement, and it seems that a middle-ground, especially if "ë" is not phonemic, would be to allow, as an option, "ë" to be written "e".

A4c. Use i for all instances of i and î. Here again, I believe that a very comfortable compromise exists, and what you're about to read is going to seem extremely similar to my response to A3, above.

My opinion: Yes, say that "î" can be replaced in all current spellings with "i". This (as suggested in proposal B2a) allows "î" to represent the stressed pronunciation of "ï". HOWEVER, make an exception to the stress rule, specifying that when "î" appears in three specific "ancient" affixes (namely the prefix "în-" and the suffixes "-înd" and "-mînt"), it does not represent a stressed pronunciation, but is purely decorative. Note that the use of "î" in these affixes would not be mandatory, but would be allowed as an option. The use of "î" in the affixes may die over time in preference to "i", but this small, easy-to-understand exception to proposal B2a will allow Talossan to preserve more of its appearance, without impacting pronunciation.

A4d. Use u for all instances of u and û. I touched on this one a bit in my discussion of proposal A3.

My opinion: With the exceptions of "decorative" circumflexes allowed as an option on the prefix "ûn-" and the suffix "-oûr", yes, every "û" becomes "u".

A5. Replace s-ch with schc, or schci (when preceding a "u", "a", or "o"), or (at the end of a the two words that end in s-ch) schtsch. [Yes, I restated this proposal, making it simpler.] This impacts only a few dozen words, and not only does this make them they prettier to look at, but also easier to read. I except from this praise those that would be modified to end with "schtsch", which is fairly ugly to my eye.

My opinion: This is a good change. Using a hyphen within a word to indicate a specific pronunciation is not a normal (or proper) thing to do. Additionally, word processors are happy to break lines on typed hyphens, without realizing that the hyphen needs to do double-duty and be changed to a double-hyphen. My only reservation about this proposal is that I wish there were a better alternative to the proposed word ending -schtsch, which adds three letters to short words. I know it currently only affects two words, but one of these is the word for "passion," which deserves to be beautiful to look at. Might I propose "schce" as the new orthography for a word-ending s-ch? This would allow the Talossan word for "passion" (which is currently "stras-ch") to become "straschce" instead of the rather more unweildy "straschtsch".

A6. Replace the tetragraph s'ch with sq (before e or i) or (otherwise) with sc. Although it does seem a bit unnecessary to have three or four ways to spell the "SK" sound, since "s" makes the "s" sound, and "ch" makes the "K" sound, and the apostrophe is precisely how a Talossan digraph or trigraph is separated to indicate separate pronunciation of the parts on either side, I do not see how we can deny that "s'ch" can be used in spelling to indicate the "SK" sound. The word for "fish" seems to me to provide a good argument for the preservation of s'ch. "Fish" is "pësc." A "fisherman" is a "pës'chéir." I do not think we can make him a "pesqéir." The spelling of all words built from "pësc" should be built from the fact that it ends with a "c". The man is a "pesc+éir", but since the C must be pronounced as a "K", it needs to become "ch" since an "e" is coming... but then, since this would give us "sch" (which would be pronounced "sh"), we must divide that with the apostrophe. By simple natural Talossan, "pësc" must become "pës'chéir."

My opinion: I do like the look of "s'ch". It also is easy to read and pronounce. It's "S" then "K". It's obviously different from "sch", and I am a fan of the use of the apostrophe to indicate the full pronunciation of the elements of what would otherwise be a bigraph or trigraph. I think that it is a feature that cannot be removed. I also agree with Cresti that "chess" looks better as "es'cheça". I wondered if "the squirrel's acorn" would look over-punctuated (it is "el castelán s'is'chir"), but it actually looks cool, and I can read it just fine. I am in the "there's no need to abolish s'ch" camp.

A7. Replace çh with c'h. The linguistic argument, that a feature used in fewer than 20 words could not be supported, is strong. It is a very cool sound, though, and one that anyone can make, an often does, every day.

My opinion: Cresti convinced me on this one. I believe we should keep "çh", and resolve to use it more. The Berbers of North Africa would be appalled if they saw that a language born from their soil had no way to express that piratical exclamation that Cresti proposes: "arçh!" However, I do agree that the three words listed as exceptions, especially isnaçhoçhâ, should be respelled with "g". Yes, I know, this leaves only sixteen "çh" words, but think what they of the Barbary Coast would think of us, without a way to spell a gutteral oath!

A8. Recognize "J" as a full-blooded Talossan letter. This is only common sense.

I completely agree with this one, and even take some credit for bringing up the silliness of the current state of "J". We can't name people "Del Cjar" and "Cjara" and claim that "J" is a foreign letter. The number of truly varied Talossan words that contain a "J" (especially in the construction "CJ") is well over a hundred. It's a Talossan letter.

A9. Admit K, W, and Y to the alphabet too. Once again, there is very little argument to be made against this one.

We cannot use the word "Clark" and "Atatürk" as often as we do and still consider them foreign words. We cannot have citizens named Matáiwos and claim that W is a foreign letter. To say that K and W only appear in words that were imported from other languages, wholesale and without any change whatsoever, is simply lying. From what language did we borrow "drêwitg"? Do we really want the river that provides the border of our nation (el Manáweg) to continue to contain a letter that is not even recognized as a Talossan letter? The case for "Y", however, is not as clear. From looking at the two dozen words containing "Y", I do not see any reason why they shouldn't all be respelled at best. I propose and argue strongly that "Y" be abolished altogether, and not even allowed the "except in imported words" exception, and that the words that use it be respelled.

A10. Drop the umlaut from "är(C)" and re-spell word-ending "är" words as "aer." Cresti makes a strong argument that the unpronounceable umlaut simply means it won't be pronounced, so it does no harm. However, I think this proposal is sound and with merit.

My opinion: I believe that in the evolution of a language, marks that indicate something unpronounceable would be the first thing that people would stop writing. Letters fall silent and people will continue to write them until Kingdom come, but diacritical marks such as this one fade away. People will stop pausing in their handwriting to draw two dots over a's that truly don't need them, and I think we should too. And especially since everything else "air"-related is already spelled in Talossan using "aer" rather than "är" (such as aerodromeu [aerodrome], aerodünámic [aerodynamic], and aeronautici [aeronautics], to name just a few), the word for "air" really should be "aer" anyway.

A11. Use the apostrophe to separate the digraph tg when necessary. All I really have to say here is, "good catch."

My opinion: This one makes too much sense, and it can be nothing but an accidental oversight that "t'g" is not currently being used in those few words mentioned in the proposal.

A12. Acknowledge that "rh" is always pronounced "sh". You are welcome, Your Majesty, for my involvement in pointing this out.

My opinion: Not surprisingly, I'm for it. It makes complete sense. I even believe that it provides the linguistic basis for the pronunciation of our infinitive verbs -- the "H" has been dropped over time (but remains in the future tense endings), but the pronunciation remains. My only objection is that I insist that the poor snipe continue to be spelled gavarhal, and given his full proper, certainly original Berber, pronunciation -- "gaw-vosh-AWL". The unfortunate bird has lived long enough with people mispronouncing his name as "gaw-vuh-RAWL." Let the snipe be as Talossan as the spelling of his name obviously shows him to be. Let his "RH" remain. As for rhesus and perhelion, I suppose I am for their proposed new orthography, since pronouncing rhesus as "shay-soos" is probably not what we want, though that pronunciation would make for a rather humorous Christian/Darwinian bumper sticker: "Va Salvadour isch Rhesus C'hriost."

A13. Allow "nh" to be written in place of "gnh". Looking back at what I have written, I seem to be in favour of proposals that allow the old, but make the new a legitimate alternative.

Multiple ways to spell the same word is a feature of many languages, including (perhaps more than any other) English, and is something that I like. American English has dropped the u from colour, but an American knows that both "color" and "colour" are exactly, 100% the same word. On this basis, I would favour this proposal. In fact, I was about to give it my full blessing. But then I looked at the treisoûr, and realized that as written, the proposal is unsound. Adopting "nh" as a digraph for "gnh" would mean that many words would be need to be respelled (for example, conheçéir [connousseur] would need to become con'heçéir). Worse, the fact that an "h" is already mandated after an "n" in certain words for assistance in pronunciation (for example, "national" is "naziunal" but "nationwide" is "toctnhaziunál"), means that it looks to me like there may not be room for "nh" after all. However, there is ample room for "ñh", and perhaps the proposal meant to make this the digraph that is suggested. I have looked and "ñh" appears nowhere in our language except in the trigraph "gñh", and therefore I propose that this proposal be altered to allow ñh (not nh) as an optional alternative to gñh.

A14. Allow "lh" to be written in place of "glh". Most of what I have to say here I have already said above, in A13.

Although I know that our glhetg is a glhetg, on the "You say tomato, I say tomahto, you write colour, I write color" basis, I do come out in favor of this proposal. I checked, and only two words in our language use "LH" outside of the "GLH" trigraph. In both cases (one of them is filharmonic) the introduction of an apostrophe to separate "LH" is natural and almost expected. In fact, a familiarity with Talossan makes filharmonic look unnatural without an apostrophe. Someone who knows Talossan sees "LH" and immediately thinks "as in miLLIOn". Seriously. Look at filharmonic. Admit it, if it's Talossan, when you look at it, you want to pronounce it "fill-yar-monic". There's no "G" but the "LH" is enough for your eye. It should be "fil'harmonic" for a Talossan speaker to see it the way it needs to be read. This convinces me that "LH" is already a digraph to our eyes, and that the "G" is optional. I support this proposal. Colour/Color. Glhetg/Lhetg.

Like Cresti, I too will leave the B and C set of proposals (on stress and exceptions) for later comment. However, I will say here that I have reviewed the stress proposals carefully, looking at their effect on the treisoûr, and my comments on the proposals will be extremely positive. As you've already seen from some of what I said above, I listed B3a as one of the many proposals I agree with, especially as it allows me to save the word "ár" from the dreaded "noastra monster".

Hooligan (yes, I failed miserably in my plan for brevity; sorry)

[King John]

I'm going to hold off replying, for the most part, until my interlocutors comment on my "Stress" proposals, since spelling and stress-marking are two aspects of the same problem, and can't really be dealt with in isolation.

But:

siervicul said:

Wail wi ar taking ûbäut linggwîstîk riforms, Ai riûlaizd thät Tûlos'n Inglîsh cood yuz û lat ûv w'rk, tu.

Now that's just unfair! The changes I've proposed, even if every one of them is adopted, leave Talossan looking only a little changed. Picking a passage of Talossan at random, consider this:

Old spelling:

Eu afirm, solenâmînt, që eu, [nôminâ], cün fidálità, rompliarhéu l'ôifisch da Sieu Maxhestà së Seneschál del Regipäts Talossán, és zefençarhéu, àl miglhôr da v'aválità, la sigürità del Estat Talossán. Så viva el Regeu!

New spelling:

Éu afirm, solénamint, qe éu, [nómina], cün fidalitá, rompliarhéu l'ôifisch da Sieu Maxhestá se Seneschal del Regipäts Talossan, es zefençarhéu, al milhor da v'avalitá, la sigüritá del Estat Talossan. Sa viva el Regeu!

Hardly the sort of wholesale revamping that Cresti parodies!

— John

[Sir C. M. Siervicül]

I said I was kidding. English is less phonetic than Talossan to begin with, so it had a lot farther to go to catch up. Seriously, though, under your proposal shouldn't "ôifisch" take an acute accent instead, to avoid confusion with the stressed version of o-umlaut?

[Hooligan]

I feel the need, now that the light of day has arrived, to clarify my long post just a bit, by saying (in case it wasn't clear to anyone) that I am a big fan of almost every one of the proposals before us. The only ones, in fact, that I believe I would say that I (currently) do not support are A6 and A7.

My comments on the other items were all intended (in the same spirit as the proposals A13 and A14) to allow for the option that some well-established spellings be acknowledged as things that can be left to die of their own accord; for example, such "decorative" circumflexes as would live on as deprecated exceptions would do so mostly in older text, and still would be -- just as the original proposal suggests -- banished from dictionaries, declared irregular, and discouraged in new writing.

In fact, to make my positions completely clear, here they are:

• A1. Yes.
  
• A2. Yes.
• A3. Yes (suggesting deprecation for specific affixes).
• A4a. Yes (and preserve àr, just as là).
  
• A4b. Still deciding, but leaning Yes.
  
• A4c. Yes (suggesting deprecation for specific affixes).
  
• A4d. Yes (suggesting deprecation for specific affixes).
  
• A5. Yes (wishing only that -schtsch could be improved for "passion")
  
• A6. No.
  
• A7. No.
  
• A8. Yes.
  
• A9. Yes (for K and W, but completely drop the letter Y...please).
  
• A10. Yes.
  
• A11. Yes.
  
• A12. Yes.
  
• A13. Yes.
  
• A14. Yes.

Finally, because it seems less clear to me on re-reading it with the sun shining, my counter-proposal on A1 was at least half tongue-in-cheek. I imagine, as much as it makes sense to me, I would stand alone if I truly proposed allowing (as an option only) the infinitive of all verbs to be spelled with "RH" (as pronunciation says they properly should).

Hooligan

[King John]

siervicul said:

I said I was kidding. English is less phonetic than Talossan to begin with, so it had a lot farther to go to catch up. Seriously, though, under your proposal shouldn't "ôifisch" take an acute accent instead, to avoid confusion with the stressed version of o-umlaut?

Yes, exactly right: óifisch.

[Sir C. M. Siervicül]

Alright, here's my response to the rest of the Modest Proposals.

john said:

The following changes are designed to simplify and greatly reduce the use of diacriticals in Talossan, and also to make the stress-marking perfectly reflect the actual pronunciation of the language.

I support the objective of making stress-marking reflect the actual pronunciation of the language, as well as simplifying and regularizing the actual stress rules. I like the flavor of Talossan's orthography, though - words bristling with diacritical marks like the stacks, turrets, and barrels on a dreadnought. So to "greatly reduce the use of diacriticals" is not a big selling point for me.

B1. First, this is the (new) "Default" Rule of Stress: Any word of more than one syllable is stressed on the vowel or diphthong before the last consonant in the word; except that the consonants in the endings -as, -en, -ent, -eux, -ic, -ici, -ilor, -laiset, and -mint don't count. (These are the “always-unstressed” suffixes.)

This is a huge improvement and would make stress far more natural and realistic (for a Romance language). I have a couple of minor suggestions for refinement of this rule. First, if a word contains a vowel with an umlaut, that vowel is stressed. That would avoid having to change words like üpsilon. Under the current rules, the first syllable is stressed. Under the new rule, the last syllable would be stressed by default, so you'd have to change the accent to û to keep stress on the first syllable. I don't see any problem with recognizing that accented vowels ought to be stressed by default. If there's more than one accented vowel, however, you have to mark which one is stressed if it's not the vowel before the final consonant. Second, there are a few polysyllabic words that don't have any vowels before the last consonant in the word, because the only consonant is at the beginning. Words like douâ (two). So we should say that if more than one syllable comes after the final consonant (except for those ignored endings), you stress the penultimate syllable. This would also apply to words like oroiâ (wetlands), which would be stressed on the first syllable under the default rule, but ought to be stressed later.

B2. If the actual stress doesn't follow the Default Rule, it has to be marked.

Absolutely. It may not be necessary to explicitly state this, but I would add that regular stress may optionally be marked. This would eliminate the need for a substantial percentage of the changes to the dictionary and existing texts that would otherwise have to be made. We could update words like Talossán over time, but in the interim they would still be acceptable. I don't see how this could create any real problems. It'd be similar to how the neutral tone is normally not marked in the Pinyin system of Chinese romanization, but it's possible to mark it if you prefer to.

B2a. If ä or ö or ü has to be stressed, mark it by changing the diacritical to a circumflex (â or ô or û).

This makes a lot of sense, although I would still like (and think it's feasible) to preserve the instances of û that I mentioned in my response to section A. Also, it might be a little clearer to say something like "if an ä, or ö, or ü that would not be stressed under the above rules is stressed, it must be marked by changing the diacritical to a circumflex." Based on your proposal example in B4, I understand this to be your meaning. At first glance, though, I thought you meant that all stressed umlauted vowels would have to be changed to circumflexes.

B2b. The other five vowels (a, e, i, o, and u) have their stress marked, if they need it, with an acute accent (‘). We should note that some writers and typesetters prefer the grave accent (`) for marking stress, especially at the end of a word. This is perfectly correct; the choice is a matter of personal style, or even handwriting.

I agree that both acute and grave should be acceptable. In order to reduce unnecessary deviation from tradition, I'd rather say that the grave accent is the default in vowels at the end of words, including in single-letter words (while leaving the acute accent as default elsewhere). This would preserve the appearance of some very common words like à and là and (like the above suggestion) would substantially decrease the burden of updating the dictionary, because there are many words that differ from your proposal only in having a final grave accent.

I agree with B2c, B2d and B2e (both of them). The second B2e resolves the parta-t-o issue that Ben and Sir Tomás were discussing a hair over 4 years ago.

B3. If the actual stress does follow the Default Rule, it should not be marked. This is true even if some other form of the same word does mark it; for instance, apricó (apricot) in the plural is apricons, without the mark.

B3a. EXCEPTION. Even though there’s no need to mark the stress on a one-syllable word, a stress mark is still used on some one-syllable words to avoid confusion with other words pronounced and otherwise spelled the same. There’s no need to continue to mark és or è or frù or tú, because there are no words es or e or fru or tu. But a = she and á (or à) = to; la = the, lá = there; etc. (Even though there is no word mar, már will keep its stress marker so as not to be pronounced mash – see point B3b.)

B3b. ANOTHER EXCEPTION. We need to be able to distinguish between the pronunciation of a verb infinitive ending in -ar (which is pronounced ash), and a non-verb ending in a -ar (pronounced ar). Talossan already marks almost all these non-verbs with a stress marker -ár; I propose that we finish the job, and establish the rule that any non-verb ending in a stressed -ar must have a stress marked (-ár) in order to have its last letter pronounced r. (And if it’s really stressed on some other syllable, both stress marks will be written, one to show the stress and one to show that the r is pronounced r. There are only five such words -- ceáiçár, Máxhár, ísobár, Stáiár, and xhágár.) For example, Noemvár or barbár (barbarian), which without the stress-mark would be pronounced Noemvash or barbash. The change is minimal, affecting only 18 words; the only common one is the preposition par, “by means of”, which I would suggest we simply add to the list of exceptionally-pronounced words rather than spelling it pár. NOTE that a few non-verbs are formed from verbs in such a way that they should be pronounced with the sh sound – sa’star (maybe) and pevar (power), for instance. These take no stress marker. BUT we then need an additional rule that the plural of such a noun, formed with the -en suffix, will also require that the r be changed to rh to maintain the sh sound. So the plural of pevar is pevarhen, not pevaren.)

My comments regarding B2 proper generally apply to B3 as well. I agree with the overall thrust of B3a and B3b in that I think words like là and ísobár should always have their last (or only) vowel marked, but I would permit marking és and è. I agree that the infinitive suffix -ar should never be marked.

B4. If a word has more than one of the “don’t-count” endings, stacked up, don’t count the last one, but do count the others. For instance, ünic is accented on the first syllable, which doesn’t need to be marked because the -ic is a don’t-count ending; but ûnicmint has to have the first-syllable stress marked.

Agree.

B5. Stop marking ïa thus to indicate it’s not the diphthong ia. If the i is stressed, make it ía; if unstressed, it needn't be distinguished from the diphthong, so write it ia (or ja). I believe there will no longer be any words in which ï is required for correct pronunciation; arc’haïsmeu will be written archaísmeu, mhïus as mhíus, etc. If there turn out to be any where í won’t do the trick, we can go on using ï.

Based on some off-Witt discussions, it sounds like the verdict is still out on this one. We may have to go on using ï in some cases.

B6. Go on using éu to distinguish it from the diphthong eu.

Good. This reminds me of something. The suffix éir: would you keep the accent or drop it? I guess it depends on whether you analyze the éi as two vowels or a single diphthong, like the debate over whether English "fire" is one syllable or two. I'd prefer to keep it.

B7. Do not mark more than one stress in a word, except when you have to write ía or éu to distinguish the double-vowel from the diphthong, or when you write -ár to suppress the pronunciation of r as sh (see B3b), and the real stress lies elsewhere in the word, in which case it has to be marked. As an example of this last case, consider mhodernistic (no stress mark needed) and pséudumhodernístic (where the é forces marking of the real stress on the penultimate).

Makes sense.

As an illustration of the kind of impact implied by the adoption of these proposals, consider the following passage from Sir Tomás Gariçéir’s Introduction to the Treisoûr, first given in the old spelling, and then in the new:

. . . .

Still quirky, still Talossan, but much easier to type.

But with 76% fewer accents! It's almost more than I can bear. The RUMP learned the hard way with the anti-diacritic plank in its platform during the 35th Cosâ elections.

To complete the task of predictable pronunciation, we really need a complete list of words that are not pronounced the way the pronunciation tables would lead you to think.

That would be most helpful. A few more that come to mind are the cardinal numbers, many of which have irregular pronunciations. For example, the numbers 30, 40, 50 etc. all end with a silent "e."

Query: Just how is gothique supposed to be pronounced? Surely not goðikwe?

That beats the heck out of me. Maybe that is how you pronounce it!

Query: Shouldn’t “Lesotho” be written “Lesuto”?

I don't think so. The grammar says that local forms are used when no traditional Talossan form exists. Which does make sense, because it's normal for older foreign place names to substantially conform to the language's orthography, while more recently adopted names are less assimilated. And Lesotho only got its current name in the 1960s. Catalan (and I believe some other Romance languages) has "Lesotho," while the more established name of Ethiopia is "Etiopia."

Query: Should “Lutheran” be written “Luteran” or “Lutgeran”? Or is it really pronounced with a ð sound?

Would you be surprised if I were to say it doesn't need to be changed? Really, I can see it being any of those, depending on exactly how it came into the language. My understanding is that even in Luther's day, German th was pronounced t. Luther/Lutheranism got the voiceless dental fricative sound in English just to conform to the pronunciation of other th words in English. So if the same thing happened in Talossan, the ð pronunciation is quite plausible. The Luteran spelling could have happened, like in other Romance languages, but would be less certain in Talossan because Talossan does have that th digraph. Depends whether the pronunciation or the spelling made more of an impact. Similarly, if Talossan got the word through English, the tg spelling is possible due to the influence of the English pronunciation.

In sum, it depends on whether the word came to Talossan 1) directly from German, 2) through English, or 3) through a Romance language. If 1 or 2, it depends on whether the spelling or the pronunciation won out. But all are plausible, I think, and if the current spelling and pronunciation are plausible, I would keep them.

Query: Is the v in vräts silent? The Scúrzniâ Gramáticâ (section 17.33) transliterates ¿non c'è vräts? as “[nonCeræc]”. Is this a mistake, an exceptional pronunciation, or an unknown rule at work?

That's odd. My best guess would be that it's a colloquial pronunciation of that phrase, which would also explain why the transcription is all run together (the grammar normally keeps the words separate). Sort of how some English speakers will run together the phrase "isn't that right?" as something like "innat right?"

[Nic Casálmac'h]

Man, you guys are scary...throwing about all these words I don't know what mean. Here are my thoughts for what they are worth... (Probably about negative two bence...)

My immediate idea was that we want Talossan to be a living language, not the project of language geeks to form as they desire. I think if something is going to be changed it should be changed through use.

On the other hand, we do not have nearly enough people using the language, which makes it more difficult for such a thing to happen. So changing spelling and such could make it easier for people to learn, which could more easily bring about the other.

Mostly I guess I like how it is. Not being nearly so skillful in matters of language and particularly the glhetg, I can't really say much that would be very helpful. I think we shouldn't change much though.

[King John]

Regarding vowels and stress, here's what I'm thinking today. (Leaving consonants aside for now, except for R — see proposals A12 and B3b.)

I've been messing around with my proposed vowel-and-stress system — A1 through 4, A10, and B1 through B6 — and I'm becoming convinced that it's basically sound. (Really good, in fact!) But there are two weaknesses with my scheme as proposed.

1. I haven't done a good job with secondary stresses in long words. (B7 just won't cover the ground.) I'll post something about this within the next few days — a solution exists! —, but it *is* a weakness in my original set of proposals.

2. The biggie — what to do about î. The situation, on deeper study, turns out to be more complicated and difficult than I had thought, or would have believed. My proposal A4c just won't cut the mustard. Sorry. Again, more on this to follow.

It's important that we do things right, and not rush into making changes that will turn out later to be just as wrong as the thing we were trying to fix. If it takes us a few weeks, or even months, to get this right, I'd say it's time well spent.

— John