type 4 fu'ivla: Difference between revisions

From Lojban
Jump to navigation Jump to search
m ((‘Table source => {and examples}’.))
Tag: 2017 source edit
(Update comprehensive form section after update that adds the missing slinku'i test part of fu'ivla (condition #3); adding it changes the tables a bit.)
Tag: 2017 source edit
 
Line 29: Line 29:


===Comprehensive Form===
===Comprehensive Form===
CLL section 4.7 defines the morphology of fu'ivla. They are brivla that are not gismu or lujvo and that satisfy the other 4 criteria. This effectively means that fu'ivla consist of valid ‘y’-less brivla with either 1) a consonant/vowel cluster that cannot appear in lujvo (this is why stage 3 fu'ivla are necessarily fu'ivla: the 4th, 5th, and 6th letters are consonants, but (section 4.1) no valid ''initial'' consonant pair starts with ‘r’, ‘n’, or ‘l’, and in lujvo consonant triples must have both pairs be valid ''initial'' consonant pairs (i.e. what section CLL 4.1 denotes as ‘CC’ rather than ‘C/C’) (see CLL section 3.7 and also CLL section 4.7's 2 ‘triple’ sentences), but fu'ivla do not have this restriction), or 2) including, if hyphenization is ignored, the right cv form, with 1 of the 4 normal fu'ivla forms (v, ccc, ccvv, and cvcv), either at the beginning or with a prefix that has the right rafsi decomposition parser state (one that leaves no possibility of a valid rafsi decomposition, by following the reductions listed in the table), according to the following table, or 3) ending in 1 of the 4 fu'ivla tail forms (c, cv, cc, ccvcc) (only ‘cv’ ends in ‘v’, so only it is conseqential for fu'ivla determination), if hyphenization is ignored:
CLL section 4.7 defines the morphology of fu'ivla. They are brivla that are not gismu or lujvo and that satisfy the other 4 of the 5 CLL conditions. This effectively means that fu'ivla consist of valid ‘y’-less brivla with either 1) a consonant/vowel cluster that cannot appear in lujvo (this is why stage 3 fu'ivla are necessarily fu'ivla: the 4th, 5th, and 6th (+1 if the rafsi isn't the CVC variant) letters are consonants, but (section 4.1) no valid ''initial'' consonant pair starts with ‘r’, ‘n’, or ‘l’, and in lujvo consonant triples must have both pairs be valid ''initial'' consonant pairs (i.e. what section CLL 4.1 denotes as ‘CC’ rather than ‘C/C’) (see CLL section 3.7 and also CLL section 4.7's 2 ‘triple’ sentences), but fu'ivla do not have this restriction), or 2) starts with a vowel (‘.v’) that necessarily morphologically separates it from anything preceding it, or 3) including, if hyphenization is ignored, the right cv form, with 1 of the 4 normal fu'ivla forms (v, ccc, ccvv, and cvcv), either at the beginning or with a prefix that has the right rafsi decomposition parser state (one that leaves no possibility of a valid rafsi decomposition, by following the reductions listed in the table), according to the table, or 4) ending in 1 of the 4 fu'ivla tail forms (c, cv, cc, ccvcc) (only ‘cv’ ends in ‘v’, so only it is conseqential for fu'ivla determination), if hyphenization is ignored.  (For cases #3 and #4, the table includes the slinku'i test (condition #3 in CLL 4.7).)


<code><pre>
<code><pre>
(v)
v
c *
c
cv *
(vv)
cc *
vc *
ccv
vcv *
(ccc)
vcc *
cvv ->  
vccv
cvc
(vccc)
(ccvv)
vcvv -> v
ccvc
vcvc
(cvcv)
(vccvv)
cvcc
vccvc
ccvcv
(vcvcv)
ccvcc *
vcvcc
ccvccv
vccvcv ->
ccvccc -> cc
vccvcc *
ccvcvv ->  
vccvccv
ccvcvc
vccvccc -> vcc
ccvccvv ->  
vccvccvv -> v
ccvccvc
vccvccvc
cvccv -> ccvcv
vcvccv ->  
cvccc -> ccvcc
vcvccc -> vccvcc
ccvcvcv -> cv
cv -> vcv
ccvcvcc
cc
ccvccvcv -> ccvcv
vccvccvcv ->  
ccvccvcc -> ccvcvcc
vccvccvcc -> cc
ccvcvccv -> ccvccv
ccv -> vccvccv
ccvcvccc -> ccvcc</pre></code>
ccc -> vccvcc</pre></code>


[https://github.com/bairyn/lujvo-forms (Table source and examples.)]
[https://github.com/bairyn/fuhivla-forms (Table source and examples.)]

Latest revision as of 23:56, 11 September 2024

fu'ivla which, unlike type 3 fu'ivla, do not have a rafsi prefixed to them to provide a semantic hint.

See also Making type-4 fu'ivla, Exhaustive list of short fu'ivla forms.

Discussion

See also Positive characterization of fu'ivla space.

  • And Rosta:
    • What opponents of type-4 fu'ivla doesn't seem to appreciate, and what the rules on the Making type-4 fu'ivla page don't seem to recognize, is that fu'ivla are not necessarily borrowings. Fu'ivla are simply a morphophonologically-defined word class, along with cmavo, gismu and cmene. Type 4 fu'ivla allow for new elegant disyllabic wordforms for new concepts. If you need a new word and no obvious elegant lujvo is available, then a nice option is to manufacture a type 4 fu'ivla.
      • nitcion:
        • What proponents of type-4 fu'ivla don't seem to appreciate is that the en masse introduction of a novel morphophonological class into a language as pernickety, intractable, and all-round brain-sapping as Lojban is impolitic; human lexing of Lojban is hard enough already. The primary responsibility of the Lojbanist remains to be intelligible - not cute.
        • xod:
          • But presumably these are widely-known within in the community, in common usage, and not created nonce. Hence, the stage-3 hazing period.
            • mark:
              • On the contrary, nonce-ness can be a Good Thing. As And points out, fu'ivla are just another class of words. They're essentially a boundless source of broda-style brivla variables. Just like we can use any string of lerfu as a ko'a-style sumti variable (which makes me think that there's practically no reason ever to use the ko'a series at all), we can use any brivla-that's-not-anything-else (i.e. not a gismu, cmavo, cmene, or lujvo) as a broda-type variable and assign with cei, for the course of a given discussion. People are always redefining their words for a discourse at hand. Fu'ivla have the advantage that they are a strong memory hook/hint that you have something specific in mind that the audience might understand; kelcrgo defined as a strategy game is more likely to be understood as Go than broda defined the same way. In actual conversation, you'd probably check that your interlocutor knows what you're talking about. See my similar rant under terki. Then again, most of the proponents of type-4 fu'ivla aren't after quite that result.
              • Jay:
                • hrm. i like that quite a bit, fwiw.
                • nitcion:
                  • .i ku'i doi mark ledo selsku na steci lei fu'ivla pe li vo .i mi xanku tu'a lei fu'ivla pe li vo ge'u .enai ro fu'ivla
  • phma:
    • Here's a type-4 I made up which isn't copied from another language: tsaparatsa'i, "ratamacue" (a triplet pattern played on a drum). Both the Lojban word and the English word are imitative of the rhythm pattern.
  • la .kreig.daniyl.:
    • Those that love type 4 fu'ivla and love using the slinku'i test every time they want to mention an animal are hereby encouraged to translate 20,000 Leagues Under the Sea. If you can do that and have every fu'ivla be recognizable as the name mentioned in the book, as well as not being confusing in that, for instance, triggerfish when called by the common name should be different from the Linnean name - if you can do that I will never complain about fu'ivla of any variety again.
    • phma:
      • Whether the common name for a particular animal or plant has any resemblance to the Linnean name depends on the language. What in English is called "purple pitcher plant" is called in French "sarracénie pourpre", which differs by only a few letters from the scientific name. The French words "orge", "seigle", and "avoine" (but not "blé", but Spanish "trigo") are derived from the Latin words which are those grains' genus names. Most of the fu'ivla I've been coining as common names are from Linnean, but the latest one, "xarcufu", isn't - it's from the Arabic word that the English common name comes from. (It got mangled in Italian: arcicioffo -> arciciocco.) For a triggerfish, how about jestelfi'e?
      • nitcion:
        • It matters because the book talks several times about how the common names are more interesting than the Linnean. How would you translate into French a book saying that the French names of animals are more interesting than the English ones? The same way you deal with this, o ye of little faith. (applying xod's "Lojban needn't convey all nuances of all languages" where it belongs) to vocab. Actually, the English version talks about how the English common names are more colorful than the scientific names, and at least once says something about how "Fishermen are better poets than naturalists."

Comprehensive Form

CLL section 4.7 defines the morphology of fu'ivla. They are brivla that are not gismu or lujvo and that satisfy the other 4 of the 5 CLL conditions. This effectively means that fu'ivla consist of valid ‘y’-less brivla with either 1) a consonant/vowel cluster that cannot appear in lujvo (this is why stage 3 fu'ivla are necessarily fu'ivla: the 4th, 5th, and 6th (+1 if the rafsi isn't the CVC variant) letters are consonants, but (section 4.1) no valid initial consonant pair starts with ‘r’, ‘n’, or ‘l’, and in lujvo consonant triples must have both pairs be valid initial consonant pairs (i.e. what section CLL 4.1 denotes as ‘CC’ rather than ‘C/C’) (see CLL section 3.7 and also CLL section 4.7's 2 ‘triple’ sentences), but fu'ivla do not have this restriction), or 2) starts with a vowel (‘.v’) that necessarily morphologically separates it from anything preceding it, or 3) including, if hyphenization is ignored, the right cv form, with 1 of the 4 normal fu'ivla forms (v, ccc, ccvv, and cvcv), either at the beginning or with a prefix that has the right rafsi decomposition parser state (one that leaves no possibility of a valid rafsi decomposition, by following the reductions listed in the table), according to the table, or 4) ending in 1 of the 4 fu'ivla tail forms (c, cv, cc, ccvcc) (only ‘cv’ ends in ‘v’, so only it is conseqential for fu'ivla determination), if hyphenization is ignored. (For cases #3 and #4, the table includes the slinku'i test (condition #3 in CLL 4.7).)

v
c
(vv)
vc *
vcv *
vcc *
vccv
(vccc)
vcvv -> v
vcvc
(vccvv)
vccvc
(vcvcv)
vcvcc
vccvcv -> 
vccvcc *
vccvccv
vccvccc -> vcc
vccvccvv -> v
vccvccvc
vcvccv -> 
vcvccc -> vccvcc
cv -> vcv
cc
vccvccvcv -> 
vccvccvcc -> cc
ccv -> vccvccv
ccc -> vccvcc

(Table source and examples.)