Proposal: "no"-"nu" Transposition
They say that there is nothing nu under the sun, but here is an idea.
I (.krtisfranks.) opine that it would be beneficial - or, at least, increasedly elegant - to transpose the meanings, selma'o, rafsi, and everything else pertaining to the words "no" and "nu". So, this is a proposal to do just that: old-"no" would now become (new-)"nu" and old-"nu" would now become (new-)"no".
In English, we can just call this proposal the "Nothing Nu (Positive) Convention"; in Lojban, it can be called ".nonu(ja'ai)s.".
- selma'o PA simplifies and becomes more regular: For the first ten integer digits, excluding that for '0', each of the cmavo is exactly of form 'CV', where "V" represents a single vowel and the vowels follow the ordered sequence ("a", "e", "i", "o", "u"), which I shall now call the 'A E I O U' sequence, for consecutive terms (and where "C" represents a single consonant); thus "pa" is followed immediately by "re", which is in turn followed immediately by "ci", and so forth in normal counting. This scheme breaks down for '0'. If 0 is counted before 1, then it would be reasonable to extend the pattern backward, choosing cmavo of form 'Cu' (where "u" is the vowel "u" and "C" is a single consonant still). If 0 follows 9, as happens on telephones/telephone menus or when counting modulo (2*5), then the pattern should extend forward as normal, thereby producing the same result (a 'Cu'-form cmavo). These expectations, however, do not hold. Instead of 'Cu', we have 'Co'; namely, instead of - say - "nu", we have "no". Moreover, the sequence runs through the five vowels in the aforementioned order completely once and almost does so for a second time - but it is one term short. This proposal would fix these problems, completing the second suffixal 'A E I O U' sequence of PA cmavo. Instead of "(no,) pa re ci, vo, mu, xa, ze, bi, so(, pano)" or "pa re ci, vo, mu, xa, ze, bi, so, no" (which feel weird and ugly when you say them - go on, try it: your tongue wants to complete the pattern and your brain is upset by its being broken), we would have "(nu,) pa re ci, vo, mu, xa, ze, bi, so(, panu)" or "pa re ci, vo, mu, xa, ze, bi, so, nu" (which are much nicer to say).
- There is a certain amount of elegance to having "ni'u" and "ma'u" bridged by "nu" (rather than "no").
- There would be a minor, but nonetheless positive, pedagogic value to the exchange. The pattern would have no exceptions in the decimal convention and the series can be memorized more easily and more pleasantly - and with fewer questions asked.
- The rafsi of each word is just the word itself followed immediately by "n". Thus, memorability is equal and the rafsi can be used in the same way in all contexts as before. (And both are fairly versatile due to the relatively high compatibility of "n" with other consonants in consonant clusters according to Lojban phonotactics).
- The word "nu" still has approximately the same phonosemantic associations as "no" in the context of nonpositives and numeric values.
- In Lojban, there is no reason to really guarantee or care about the culturally non-neutral sequence of vowels in the Latin order ('A E I O U'). Internally, there is no especial reason to not precede "a" by "o". However, Lojban does have vowel patterns (for example: 'I A U', 'E O', 'A E I O U' are common ones amongst monophthongs), and the decimal PA sequence is (mostly) one of them (suffixally).
- "fasnu" no longer corresponds with "nu". I think that this is a relatively minor cost since the connection is not especially strong and we can always informally replace "fasnu" with "fasno" (except for its cmarafsi "-fau-").
- We would have to rename selma'o NU to "selma'o NO" or choose another name for it (such as "selma'o SUhU") - everywhere. Note, though, that Lojban texts using "ma'oi" would not need to be changed if the word immediately following "ma'oi" were any cmavo of selma'o NU or selma'o PA unless that word were either "no" or "nu" (thus, a simple find-and-replace routine would be sufficient for those); it is only foreign language (such as English) texts/references which would need to be combed slightly more carefully (still, a simple find-and-replace routine should do the trick, especially if it is case-sensitive).
- This proposal is not back-compatible, so all old documents need to be updated so as to comply with this proposal or otherwise become marked/disclaimed for noncompliance therewith if this proposal were to be adopted. However, this should be relatively easy since the change is just in the vowel of these two words in all cases (including rafsi/lujvo, exempting cmevla, and providing for cmavo clusters (such as "lonu" and "lino")).
- It is not a pressing concern or issue with the language and would not improve functionality or anything like that in any way. It would, at best, make the language seem more systematic and could, very marginally, improve pedagogic practice/ease. The work necessary for the update would be rather large upfront for such payoff.
However, if we just enable the utterer/user to specify which version of Lojban they are using (or just assume cultural defaults), then many of these concerns evaporate.