Morphology proposals

From Lojban
Jump to navigation Jump to search

This is a grab bag of changes to the morphology grammar that people have considered over the years. Unless noted otherwise, BPFK's working morphology, as implemented by ilmentufa, is used as a baseline for describing the proposals. See also this discussion from when the BPFK morphology was first being developed.

CyC syllables as alternative or replacement for consonantal syllables

Proposed by la xorxes and others. Rationale: some speakers find syllabic consonants difficult to pronounce or to distinguish from -yC- sequences.

xorxes's original proposal was to allow these syllables only for type 3 fu'ivla hyphens; more recent variants have no such restriction.

Newly grammatical: sincyrboua, sipnybzu

brivla_head = !cmavo !slinkuhi !h &syllable unstressed_syllable*
# Change y to y_hyphen in every y-rafsi rule
y_hyphen = y &onset
stress = (consonant / glide)* h? y? coda? syllable pause
consonantal_syllable = onset y coda / consonant &syllabic coda

In the wild

durka (2008) xu cirla cidjrbyrga
jukni (2010) lo juknynba cu mutce kukte
tsani (2012) i do co'e ca lo xo'yldei
selpahi (2014) pe'i lo saltyrkaisara mi cnino .i ca lo cabdei mi citka lo salta be lo laktuka jo'u lo tamca jo'u lo clazme .i no'e cinri .i ku'i mi na pante
lablanu (2015) mi ca gugylsisku .i mi na facki no da
Amnadya (2015) .i so'o gernyrkasu cu srana zo ca
fegmese (2016) ti'e zmadu so'a drata linyksu lo ka lo pilno be ce'u cu nitcu lo ka kurji ... ri xu

Removing or simplifying the slinkuhi rules

Proposed by many and made official in TLI Loglan. Rationale is that slinku'i in its current form makes the morphology unnecessarily hard to learn and use in fluent speech. Also often part of calls to do away with assigned rafsi, since slinkuhi exists to favour words made of those.

Killing .oisai'yvla (brivla rafsi merger)

The least radical of the slinku'i proposals. sai'yvla would become a valid word, removing the only instance where a string beginning with a single consonant can be a slinku'i.

Newly grammatical: sai'yvla

No longer grammatical: .oisai'yvla (would parse as two words)

initial_rafsi = !slihykru extended_rafsi / y_rafsi / !any_fuhivla_rafsi y_less_rafsi !any_fuhivla_rafsi
any_fuhivla_rafsi = fuhivla / fuhivla_rafsi / stressed_fuhivla_rafsi
stressed_extended_rafsi = stressed_hy_rafsi / stressed_fuhivla_rafsi
extended_rafsi = hy_rafsi / fuhivla_rafsi
stressed_fuhivla_rafsi = fuhivla_head stressed_syllable consonantal_syllable* onset y
fuhivla_rafsi = &unstressed_syllable fuhivla_head onset y h?
slihykru = (CCV_rafsi / stressed_CCV_rafsi) h y onset
hy_rafsi = (long_rafsi unstressed_vowel / CCV_rafsi / CVV_rafsi) h y h?


slinku'i only on word-final or -'y short rafsi

This is what Loglan has implemented, and what Lojbanists usually mean by "getting rid of slinku'i". sporte would become a valid word, but not sli'a, so that CVV and CV'V rafsi behave identically with regards to hyphen requirements. Some also propose that sequences like dja'y- remain as slinku'i so that brivla rafsi like -uidja'y- stay valid.

Newly grammatical: sporte zdanco'o klesnai krinju

No longer grammatical: .esporte .adjunte .asfalte bavlamdei faglitki .islenska lojbangirzu

slinkuhi = consonant CVV_final_rafsi
# Plus the changes above and optionally removal of the slihykru rule

In the wild

dbrock (2007) mi nelci lo lenku djarpitsa
timonator (2008) i bazibo lo cridrimpe cu renro lo bolci be lo fagri mi
latros (2011) .i ro da poi ke'a cmacrordra zo'u ro de zo'u de me'au da de
klaku (2012) ko kibrolinki plokso
danr (2014) .i mi ckaklyzu'e mu'i lo nu tatpi ru'e .i ku'i za'u re'u casnu cu'e
menli (2014) lo nu da'i la datnymu'e remna cu simsa lo drata brotkaste pe BBC cu se cumki lo nu lo vreji be lo nu la .robin. cu fonxa tavla cu nenri lo brotkaste vau :P
emna (2015) no'i xu do jinvi co jbofnu mapti co selcme zo .emna
gleki (2015) ie i xu do ba favgau lo frontende
omni___ (2015) lo ka sruri na dzistribu .i lo du'u lo tadni cu sruri lo ckule na nibli lo du'u ro lo tadni cu sruri lo ckule
clakre (2016) mi na se cmene zo clarke .i ku'i se cmene zo clakre
uakci (2016) .i je lo nu tai zukte cu se jalge lo nu do'e lo cmalu skrine cu za'u to zmadu lo na se me'oi br toi mei co linji
la_aLEKsolas (2017) ju'opei lo tsaibaklo lo vajraifla cu citka

No slinku'i at all

Under this option, vla'i would become a word and the lujvo for balvi + lamji would come out as bavyla'i.

Transitional dialects

la ctefa'o and others propose ways for slinku'i-ful and slinku'i-less dialects to share the same lexicon. slinku'i-ful dialects could import what they see as slinku'i by prepending a glottal stop. Relative to the lujvo-zi'evla merger proposal below:

brivla = !cmavo brivla_head brivla_core &post_word
post_word = pause / !nucleus !slinkuhi lojban_word

slinku'i-less dialects could keep words that used to depend on slinku'i glue together using an extended CVCy-lujvo rule.

In the wild

Ilmen (2017) sei ca lo nu viska zo ckt cu pensi be lo. tcuktci bangu
solpahi (2017) pu lo nu casnu cu pensi zo temkracabdu'i .i lo nu na birti pu mukti lo nu. pretsku

Forbidding certain initial clusters

la solpahi proposes to remove all consonant triples as well as ml- and mr- as initial clusters. Rationale is that regular prefixes such as kam- would be less likely to cause tosmabru, and that some existing words (notably of CVCCCVCV shape) would be preserved if slinku'i were to be abolished.

Eight gismu would need to be regenerated: mlana mlatu mleca mledi mluni mraji mrilu mruli.

Some other words no longer grammatical: citmle mlatejami mlongena mlumitu mri'au skripto skrolbake spranto strelka stribilo

onset = h / glide / !(consonant consonant consonant) initial
other <- p / t !l / k / f / x / b / d !l / g / v / m !liquid / n !liquid

In the wild

solpahi (2016) su'a mi na'o nelci lo danlu .i lo maltu jo'a (to .u'i pei toi) cu mupli
solpahi (2016) mi nelci lo malna si frikative
jelca (2017) .i melbi fa lo tarci maltu

Forbidding certain clusters altogether

la ctefa'o proposes to forbid clusters md, mt, nb, np, which are prone to assimilation.

27 gismu would need to be regenerated: canpa cenba cimde cinba denpa famti femti finpe janbe jimte kanba kumte lanbi lunbe mamta nanba panpi penbi semto senpi sumti tanbo tinbe tunba vamtu vanbi zenba

Some other words no longer grammatical: bavlamdei cinpa'i donpre kinpau menbi'a prulamdei runbau samda'o vimku'a

In the wild

timthelion_ (2006) do jinga lo titnamba
dbrock (2007) ta'o mi ro roi cusku zo titnamba
codrus (2009) .i mi citka lo grute titnamba

Allowing some new medial clusters

Calls to make mz a valid cluster have always been commonplace. la lalxu also proposes to allow n plus affricate clusters.

m <- comma* [mM] !h !glide !m
n <- comma* [nN] !h !glide !n

In the wild

rlpowell (2005) je'e .i ca lo mergu'e so'a mi'a salci lo bavlamzeldei po'o
timonator (2008) mi badri lo nu do na samzva
menli (2014) .i ua la .randol.xolmz. noi loglo jatna cu ba'o se micka'a
selpahi (2015) lo larkamzi'e cu krali ro larfi'i
Ilmen (2015) ni'o mi ca'o tcidu fi tau lo .itku'ile genkasu noi ba'u memzma la pa no no mei
mungojelly (2015) ni'o mi mumblenomzvati ca doi ro do
solpahi (2016) .oi se'i mi za'u re'u xaksu lo temze'u si ze'u zei temci fau lo nu da'arta'a co palcau
bongnanba (2016) .i mu'a sei jijnu ti noi tabypinsi je pavmei jo'u ti noi penbi je pavmei va'o lo nu mi co'a ponse su drata pinsi cu na memze'a


Adding a new diphthong

This is often part of proposals to add new series of cmavo, so that they can both stay short and fit into the existing initial consonant patterns. la zipcpi has proposed eu, la solpahi ui [uj].

-y syllables within non-lujvo (lujvo–zi'evla merger)

Proposed by many over the years, especially together with removal of assigned rafsi. Usually implemented by removing any detection of rafsi that isn't related to slinku'i. slinku'i itself only applies at the beginning of a word, not after hyphens.

BRIVLA = brivla
brivla = !cmavo !slinkuhi !slihykru !h brivla_head brivla_core &post_word
brivla_head = &syllable (unstressed_syllable / medial_syllable)*
brivla_core = stressed_syllable medial_syllable* final_syllable
medial_syllable = &(onset !vowel !diphthong) any_syllable
slihykru = (CCV_rafsi / stressed_CCV_rafsi) (h / initial_pair) y onset
rafsi_string = y_less_rafsi* (gismu / CVV_final_rafsi / stressed_y_less_rafsi short_final_rafsi / y_rafsi / stressed_y_rafsi / hy_rafsi / stressed_hy_rafsi))

For a slinku'i-less variant, change rafsi_string to the one shown here and remove initial_pair from slihykru.

In the wild

kamsen (2007) lo tavlrgedyflai be le lojbo cecmu
timonator (2008) i ki'u bo la anar dasni lo nukni se dasni be lo prenrpairyto
tsani (2012) i ji'asa'e mibykemymo'ydji
zahlman (2012) u'i pe'i lonu nanbrdonyto bevri cu no'e nu jarco loka tsali
gleki (2014) lo'e banfyracolotli cu nelci lo ka citka lo me'oi fpcalep vau sei mi pu reisku
mukti (2014) na nitcu lo cnino bauplyfuzykamni .i na manku .i lo jbopre cu lazni lo ka viska
akmnlrse (2015) .i mo .i la'a ca sipnybi gi'e ba ke fau gi ciknybi gi di'a gunrizu
selpahi (2015) .i .au pei mi merdjygau
durka42 (2015) plisyta'oi je tcati
Ilmen (2016) mi pu zu pensi lo nu zo sfataba da'i valsi lo si'o za'e xorsyflai
solpahi (2017) pe'i lo nu zi'evlygau tai tu'a di'u cu ji'i no roi xagzenri'a

Expansion of the CVCy-lujvo rule to non-lujvo

Proposed by la solpahi and others. One of two ways to remove this last bit of privilege given to traditional lujvo after unifying brivla and removing the slinku'i rule.

This would turn streams of sound currently parsed as multiple words -- any sequence of the form [cmavo not ending in y] [one or more cmavo ending in y] [non-cmevla word] -- into single brivla. Notably, .a'y-style letterals would very often require glottal stops after them.

CVCy_lujvo = onset nucleus medial_syllable+ (brivla / final_syllable)

In the wild

solpahi (2016) mi pu mojyko lo du'u lo cabdei cu xavdei

CV-shaped rafsi

Proposed by la solpahi as part of logbanu, to shorten regular compounds ending in certain gismu that officially lack one-syllable terminal rafsi.

For example, binxo would gain -bi as a rafsi, shortening ciknybi'o to ciknybi.

This proposal does not necessarily affect the morphology PEG; unifying brivla and expanding CVCy-lujvo would already allow such words. It is rather a proposal to give these words regularly derivable meanings.

CCy-shaped cmavo

Proposed by la mudri as a way to expand letteral space.

ziren (2009) doi djy. mi pu kibro ciska lo ve ciksi fe be la tenuar. pe mi

Restricting cmevla to valid syllables

Proposed by la xorxes and others, as the simplest way of reducing the amount of cmevla that the parser accepts but speakers find ungrammatical.

cmevla = jbocme

In the wild

tsani (2015) .i xebni mu'i ra'u lo nu iy. jai nandu fai lo ka ce se pilno ci'e la .linukys.
solpahi (2016) .i do'a nai mi sinma la .kymir. ki'u lo nu ri troci lo ka tcidu

Allowing glides after consonants in cmevla (partial revert of BPFK decision)

Proposed by la .ilmen. and others, as sort of a counterpoint to the above proposal.

Ilmen (2016) ua la idiot.bot. co'a zifre
lynn (2016) .ua mi tirna fi tu'a la .teilr.suift. fe tu'a la'o xy. file_94.oga .xy.
ilmen (2016) .i xu zo .period. cmene fe ji'a lo namcu mocle'u lo glibau baupli

Allowing glides after consonants, but not within initial clusters (partial revert of BPFK decision)

Proposed by la tsani and others.

CGV syllables as alternative or replacement for CV'V cmavo and rafsi (La Alta)

Proposed by la gleki and la .aleksolas.

Treatment of VGV sequences similarly to V'V (as in CLL)

Proposed by la gleki.