!Distributed under the terms of the GNU General Public License version 2 ! or any later version. ! ========================================================================== ! ! North Sámi morphophonological rule set ! ! ========================================================================== ! ! twol-sme.txt ! The file contains the rule set for the non-segmental ! North Sámi morphophonological rules ! m4 rules for selecting parts of the file to a certain application. changequote(<<,>>) changecom(<>,<< >>) define(<>, <>,<>,1)>>,<>,0)>>)>>)dnl dnl define(<>,0)dnl define(<>,<>)divert(DIVNUM)dnl>>)dnl define(<>,<>)divert(DIVNUM)dnl>>)dnl define(<>, <>,ifelse(eval($1),0,0,-1))divert(DIVNUM)dnl>>)dnl define(<>, <>,ifelse(eval($1),0,-1,0))divert(DIVNUM)dnl>>)dnl define(<>, <>,ifelse(DIVNUM,0,-1,0))divert(DIVNUM)dnl>>)dnl FLAGINIT(<>) FLAGINIT(<>) FLAGINIT(<>) FLAGINIT(<>) FLAGINIT(<>) ! HYPHENATE = To be used when making hyphenators - it allows surfacing of # and ^ ! SHORTCOMP = Allows/encloses subnorm short compounds ! DIPHSIMPL = Allows/encloses subnorm diphthong simplification Alphabet a b c d e f g h i j k l m n o p q ! small r s t u v w x y z å ä ö ø æ %- á é ó ú í à è ò ù ì ë ü ï â ê ô û î ã ý þ ñ ð ß ª ß ç A B C D E F G H I J K L M N O P Q ! capital R S T U V W X Y Z Å Ä Ö Ø Æ Á É Ó Ú Í À È Ò Ù Ì Ë Ü Ï Â Ê Ô Û Î Ã Ý þ Ñ Ð e7:e e9:e i7:i o7:o o9:o u7:u ! Morphophonemes č đ ŋ š ŧ ž ! Sámi letters. Č Đ Ŋ Š Ŧ Ž ! KEEPIF(HYPHENATE) # %^ ELSE #:0 %^:0 ENDKEEP ':0 %/ ¤:0 '7:' ! ' is for CnsGrad of the lg:lgg and l'l:ll type ! ¤:0 prevents ConsGrad in certain words ! '7 is the real apostroph ! # is used to mark both lexicalised and ! derived compounds h7:h h8:h g8:g m8:m n8:n ! the x8 ones are consonants that alternate in H7:H H8:H G8:G M8:M N8:N ! stem-final positions. j9:j b9:b d9:d g9:g h9:h k9:k m9:m n9:n p9:p s9:s t9:t z9:z š9:š r9:r J9:J B9:B D9:D G9:G H9:H K9:K M9:M N9:N P9:P S9:S T9:T Z9:Z Š9:Š R9:R ! The x9 ones are consonants that never alternate. ! The capital J9 etc. do not work. ! Added to be able to easily see if there are bugs in the twolc rules X1:0 X2:0 X3:0 X4:0 X5:0 X6:0 X7:0 X8:0 X9:0 ! diacritics Q1:0 Q2:0 Q3:0 Q4:0 Q5:0 Q6:0 Q7:0 Q8:0 Q9:0 ! They trigger morphophono- Y1:0 Y2:0 Y3:0 Y4:0 Y5:0 Y6:0 Y7:0 Y8:0 Y9:0 ! logical rules W1:0 W2:0 W3:0 W4:0 W5:0 W6:0 W7:0 W8:0 W9:0 ! Make stem borders visible when using the M4FLAGS=-DHUNSPELL when invoking make OMITIF(HUNSPELL) %>:0 %>7:%> »:0 ! stem-suffix border mark ELSE %> ! stem-suffix border mark, only visible in Hunspell ENDOMIT ; Sets Vow = a á e e7 e9 i i7 o o7 o9 u u7 y æ ø å ä ö A Á E I O U Y Æ Ø Å Ä Ö Å é ó ú í à è ò ù ì ë ü ï â ê ô û î ã ý É Ó Ú Í À È Ò Ù Ì Ë Ü Ï Â Ê Ô Û Î Ã Ý ; Cns = b b9 c č d đ d9 f g g8 g9 h h8 h9 j j9 k k9 l m m8 m9 n ŋ n8 n9 B B9 C Č D Đ D9 F G G8 G9 H H8 H9 J J9 K K9 L M M8 M9 N Ŋ N8 N9 p p9 r s š t ŧ v z ž z9 ' '7 P P9 R S Š T Ŧ V Z Ž Z9 ; StemCns = b b9 d d9 g g8 g9 h h8 h9 j j9 l m m8 m9 n n8 n9 r s š B B9 D D9 G G8 G9 H H8 H9 J J9 L M M8 M9 N N8 N9 R S Š ; WeG = X4 X5 X6 X8 X9 Q4 Q5 Q6 Q7 Q8 W1 W4 W5 W7 ; Len = Y1 Y2 Y3 Y4 Y5 Y6 Y7 Y9 ¤ ; Dummy = X1 X2 X3 X4 X5 X6 X7 X8 X9 Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Y1 Y2 Y3 Y4 Y5 Y6 Y7 Y8 W1 W2 W3 W4 W5 W6 W7 W8 W9 ¤ %> %^ ; FinalDummy = X2 X4 X7 X8 %> %^ ; ! This set is not in use !! NonMPDummy = X1 X4 X7 X8 ¤ %> %^ ; NonMPDummy2 = X1 X4 ¤ %> %^ ; HighVow = i i7 u u7 I U ; NonHighVow = a e e7 e9 o7 o9 o A E O ; Stop = k p t c č K P T C Č ; DelCns = g8 m8 n8 h8 G8 M8 N8 H8 ; CntrCns = s j h ž S J H Ž ; Hyph = %- ; Definitions ! G3 = Vow Cns Cns Cns Vow ; ! G2 = Vow Cns Cns Vow ; ! G1 = Vow Cns Vow ; ! G12 = G1 | G2 ; ! The G1 gang ! The G2 gang ! The G3 gang ! đ'đ f'f l'l hl'l m'm n'n n'nj ŋ'ŋ r'r hr'r s's š'š ŧ'ŧ v'v ! bm dn dnj gŋ ! hcc hčč hpp hkk htt ddj ! llj ! bb dd gg zz žž ! r'bm r'dn r'dnj 'rgŋ !đ'b, đ'b, đ'j, đ'v, i'b, i'c, i'd, i'f, i'g, i'k, i'l, i'hl, i'hm, i'hn, i'p, i'r, i's, i't, i'v, i'z, !l'b, l'c, l'č, l'd, l'f, l'g, l'j, l'k, l'p, l's, l'š, l't, l'v, l'ž, m'b, m'p, m's, m'š, n'c, n'č, n'd, !n's, n't, n'z, n'ž, ŋ'g, ŋ'k, r'b, r'c, r'č, r'd, r'f, r'g, r'j, r'k, r'p, r's, r'š, r't, r'v, r'z, r'ž, !s'k, s'm, s'p, s't, š'k, š'm, š't, š'v, t'k, t'm, v'd, v'g, v'j, v'k, v'l, v'hl, v'p, v'r, v't, v'z, v'ž !đ'bm d'gŋ i'bm i'dn i'gŋ l'bm l'dn l'gŋ v'dn v'dnj !i'sk i'st m'šk n'sk n'st r'sk r'st v'sk !k'c k'č k's k'st k'š k't ! G3 = Cns: Cns: Cns (Cns) | b m | d n | g ŋ | b b | d d | g g | z z | ž ž | ! Intuition: ! Have the => rule work in case of (upper or lower or both?) G3. ! Have the <=> rule work in al ! Define G12: Here we have <=> ! Define G3 (or not): Here (i.e. elsewhere) we have =>. !Defining G1, G2, G3. !Auxiliary definitions G1 = Cns | h [l|r] | n j ; !UCns = Cns: ; !RLCns = :Cns; !RealCns = UCns:LCns - ':0 ; RealCns = Cns - ':0 ; all2Cns = Cns Cns ; real2CnsG3 = b: m | d: n | g: ŋ | b: b: | d: d: | g: g: | z: z: | ž: ž: ; ! This could be the problem 2CnsG2 = all2Cns - real2CnsG3 ; 3CnsG2 = Cns RealCns Cns ; !!!!!!!!!!!!!! ! đ'b:đbb, đ'g:đgg, ... fake8CnsG3 = Cns: ': Cns ; ! đ'b:đ0b 3CnsG3 = h [ c c | č č | p p | k k | t t ] | [ d d | l l ] j ; !G2 = 2CnsG2 | 3CnsG2 ; ! 2CnsG2 is well-defined, 3CnsG2 is not. :-( !G3 = fake8CnsG3 | real2CnsG3 | 3CnsG3 ; !G3 = fake8CnsG3 ; !err !G3 = real2CnsG3 ; !err !G3 = 3CnsG3 ; !err ! => All 3 parts of the G3 definition are flawed !G12 = G1 | G2 ; ! This is what we waht G12 = G1 | 2CnsG2 ; ! testing G2 parts ! This we do while waiting for 3CnsG2. ! The second part of the G2 definitions is flawed. ! Pseudocode (see "Going for 2" below) ! Diphthongsimpl <=> _ G12 i j: ; ! OptDiphtsimpli => _ G3 i j: ; ! ============ ! ! Rule section ! ! ============ ! Rules !! Remainder! Change all #: to (Hyph) #: in order to account for ealáhus- ja ... ! Introduction to the set of testwords (introduced with !€ and !$ that are added) ! ------------------------------------------------------------------------------- ! (this documentation will be enlarged and moved to the documentation, and a trimmed version ! will be left here) ! For each rule, there is a set of word pairs to test. The words will be extracted into ! two files (the !€ words for positive tests, the !$ ones for negative tests, i.e. pairs that ! the parser shall reject. ! Basic usage is: (this paragraph shall be removed, as we want the one in the next paragraph) ! open twolc -utf8. Issue the following commands: ! install-binary bin/twol-sme.bin ! pair-test-file filename (where "filename" is the name of the file containing the test words) ! Usage (this is what we are aiming at): ! In the testing catalogue, issue one of the commands ! make twol-positive ! make twol-negative ! make twol-test (do both) ! This will extract the test pairs from this file, and print a report for each of the tests. ! The reports will be printed to report files named twol-pos.out and twol-neg.out. The reports ! are finally scanned by another script, which gives a warning if anything unexpected was found. ! We may also use other testfiles than the ones generated from this file (twol-sme.txt file). ! make twol-test TYPE=pos FILE=testfile (when the file contains positive test pairs) ! make twol-test TYPE=neg FILE=testfile (when the file contains negative test pairs) ! make twol-test FILE=testfile (the default is TYPE=pos) ! make twol-test HELP (prints out these lines) ! The test cases are of two types: ! - positive tests ! - negative tests ! ! Positive tests are such that they require the pairs to match - if not, the rule (or ! another conflicting rule) is broken. These test pairs should be designed to check that ! the rule allways give the wanted mapping. ! ! Positive tests are marked with a € sign (and as a comment, to not infer with the ! regular rule compilation): ! !€ áj'gi !€ ái0gi ! ! Negative tests are pairs that should NOT match. The pairs should be designed such that ! they are likely but illegal pairs, so as to identify possible holes in the rules. Such ! holes will lead to overgeneration, and spurious acceptance of misspelled words, wrong ! analyses, bad generation, etc. Also, errors in the rules can lead to an inefficient ! analyser. ! ! Negative tests are also very important for regression testing: all bugs found and ! corrected should be added as negative pairs illustrating the bug that was corrected. ! This way, we can always check whether the bugs are still fixed or reintroduced by ! other bug fixes. ! ! Negative tests are marked with a dollar sign: ! !$ mánnáX5jd !$ má0ná0jd ! ! You may have more than one test word (pair) on each line, that is, you can have several ! pairs after each other on two lines separated by one or more spaces. This will save ! precious vertical space in the file, and make the rule file easer to read. ! ! The result/output is: ! REJECT ! ACCEPT ! Error ! and will be put into a separate test result file. A final script checks that the test ! result is what it should be, and reports the findings (ok/error(s) found). ! ============================ ! ! Here begins the rule section ! ! ============================ ! ! j rule ! ------ "Postvocalic j Surfacing" j:i <=> Vow: ( :0 ) ( %>: ) ( Dummy: ) _ [ Cns | ': | ( :0 - Y5: - Y6:)+ (Hyph) [# | #:] ] ; !more and more dummies... !€ áj'gi !€ ái0gi !€ majdege !€ maidege !$ almmajX4in !$ almmai0in !€ mánnáX5jd !€ má0ná0id !€ almmájX4 !€ almmái0 !€ almmájX4X7-# !€ almmái00-0 !€ almmájX4in !€ almmáj0in ! Added ( :0 ) because of olmmoš ! new version for Troandin-bisma ! All j:i alternation is conflated into one rule. LC is vowel and optional ! (suffixinitial) dummy symbol, RC is consonant, dummy or word boundary, ! The intention is to have the strict operator (<=>), but catch the ! variation with a lot of (=all) contextual alternatives. ! exchanging DummyNotY5 with :0 - Y5: , better luck this time. ! Word final consonant alternations ! --------------------------------- ! Only alveolar consonants word-finally: "Word Final Consonant Neutralization 1" !smirezit : smires, Troandin-bisma Cx:Cy <=> Vow: (( CntrCns:) Dummy:+) _ (':0 k:0) ( :0 - Y5: - Y6: )( :0 - Y5: - Y6: ) ( %>: ) [ Hyph | [ # | #: ] ] ; where Cx in (z ž m h p g b d) Cy in (s š n t t t t t) matched ; !Testcases for the above rule: !€ smirez# !€ smires0 !€ Troandim# !€ Troandin0 !€ Troandim- !€ Troandin- !€ muhtum# !€ muhtun0 !€ skoalkkuh# !€ skoalkkut0 !€ nagod# !€ nagot0 !€ bávččag# !€ bávččat0 !€ nuorab# !€ nuorat0 !€ bávččag# !€ bávččat0 !€ eamid# !€ eamit0 !€ alih# !€ alit0 !$ vejolažž# !$ vejolaž00 !€ vejolažž# !€ vejolaš00 !€ olmmožX4X7- !€ olmmoš00- ! This test pair irrelevant, as line commented out. !$ olmmožX7- !$ olmmoš0- !€ olmmožX4 !€ olmmoš0 "Word Final Consonant Neutralization 2" Cx:Cy <=> [ # | #: ] Cns:* (Vow:) Vow: Cns:+ Vow: j: _ ( %>: ) [ # | #: ] ; !háliid/háliit where Cx in (d) Cy in (t) ; "Deletion: Final h s j ž in Contracted Stems" Cx:0 <=> Vow _ ( %>: ) [ Y5: | Y6: j: ] ; !1.138 !Cx:0 <=> Vow _ [ Y5: | Y6: j: ] ( [ Hyph | #: ] ) ; !Cx:0 <=> Vow _ [ Y5: | Y6: j: ] [ Hyph | #: | Cns: ]; where Cx in (h s j ž) ; ! Debugging, 142 for reference. ! Changed because:- !€ fijdnisY5t !€ fiidná00t !$ fijdnisY5t !$ fiidnás0t !$ fijdnisY5t !$ fiidnis0t !€ albmájY5 !€ albmá00 !€ olbmožY5 !€ olbmo00 !€ fijdnisY5t- !€ fiidná00t- !€ albmájY5- !€ albmá00- !€ olbmožY5- !€ olbmo00- "Word Final Cluster Simplification 1" Cy:0 <=> Vow Cx _ (Hyph) [ # | #: ] ; where Cx in (l r) Cy in (d g) ; !€ vuordild# !€ vuordil00 !$ vuordild# !$ vuordild0 "Word Final Cluster Simplification 2" ! attestit : attes Cx:0 <=> Vow (W4:) s _ ( %>: ) (Hyph) [ # | #: ] ; where Cx in (t k m) ; !€ attest# !€ attes00 !$ attest# !$ attest0 "Word Final Cluster Simplification 3" ! beroštit : beroš t:0 <=> Vow š _ ( %>: ) (Hyph) [ # | #: ] ; !€ berošt# !€ beroš00 "Word Final Cluster Simplification 4" ! bearjadahka : bearjadat Cx:0 <=> Vow h: ':0 _ ( %>: ) (Hyph) [ # | #: ] ; where Cx in (t k) ; ! Changing bearjadahk to bearjadah'k, adding ':0 !€ bearjadah'k# !€ bearjadat000 !$ bearjadah'k# !$ bearjadat0k0 !$ bearjadah'k# !$ bearjadat'00 !$ bearjadah'k# !$ bearjadah000 !$ bearjadah'k# !$ bearjadat'k0 !$ bearjadah'k# !$ bearjadah0k0 !$ bearjadah'k# !$ bearjadah'k0 !"Word Final Cluster Simplification 5" ! bearjadahka : bearjadat ! h:t <=> Vow _ ':0 [t:0 | k:0] (Hyph) #: ; ! Splitting WFCF 4 into 2 to handle the inserted ':0 symbol. ! But this seems already covered by "Word Final Consonant Neutralization 1" "Word Final Double Consonant Simplification" ! muitaluss : muitalus Cy:0 <=> Vow Cx: _ ( %>: ) (Hyph) [ # | #: ] ; where Cx in (s š d ž) Cy in (s š d ž) matched ; !€ muitaluss# !€ muitalus00 !$ vejolažž# !$ vejolažž0 !€ vejolažž# !€ vejolaš00 ! Consonant gradation rules ! ------------------------- !Now that long double cons. have been analyzed having lexical triple consonants, !many of the grad. rules of the form ':Cx may be collapsed to a single rule !without left context: _ Cx Vow: WeG: ; where XXXX matched ; . This has not !been done yet (PS 29.11.1993). ! But the long double cons have not been analyzed as triple consonants! "Gradation: h Loss" ! johka : joga (h:0) h:0 <=> _ ': Cy: ( %>: ) Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cy in (p t k c č) ; !€ joh'kaX4 !€ jo00ga0 !$ joh'kaX4 joh'kaX4 joh'kaX4 joh'kaX4 joh'kaX4 jo0'kaX4 !$ joh0ga0 jo0'ga0 jo00ka0 joh'ga0 joh0ka0 joh0ga0 "Gradation: Prenasal Stops" ! sápmi : sámi (p:0) Cx:0 <=> _ Cy Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in (p t k) Cy in (m n ŋ) matched ; !€ sápmiX4 !€ sá0mi0 !$ sápmiX4 !$ sápmi0 "Gradation: tnj" ! latnja : lanja (t:0) t:0 <=> _ n j Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; !€ latnjaX4 !€ la0nja0 "Gradation: dj" ! vuodjit:vuoján d:0 <=> Vow: _ ': j Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; !€ vuod'jiQ4n !€ vuo00já0n "Gradation: Double Consonant" ! káffe:káfes Cx:0 <=> Vow: _ Cy ( %>: ) Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in (đ f l m n ŋ r s š ŧ v) Cy in (đ f l m n ŋ r s š ŧ v) matched ; !€ káffeX4s !€ ká0fe0s ! Note: I added a t rule, due to substandard Ruotta:Ruotas "Gradation: Double t" t:0 <=> u o _ t a: ( StemCns:) (:StemCns) ( %>: ) WeG: ; !€ RuottaX4s !€ Ruo0ta0s !"Gradation: šž" !Commented out because it causes wordforms like *muitalepmožis. Instead we now have !this in lexicon: !nju0nuš:njunnuš MALIS ; !nju0nuš:njunnuž MALIS ; !SUB ! Cx:Cy <=> Vow: _ X2: Vow ( StemCns:) (:StemCns) ; ! njunnoša:njunnožis ! Cx:Cy => Vow: _ X2: Vow ( StemCns:) (:StemCns) ; ! njunnoša:njunnožis ! where Cx in (š) ! Cy in (ž) ! matched ; !!€ njunnošX2a !!€ njunnož0a "Gradation: Preaspirated Geminates" ! áhkku:áhkus Cx:0 <=> h _ Cy Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in (k p t c č l r) Cy in (k p t c č l r) matched ; !€ áhkkuX4s !€ áh0ku0s "Gradation: Jodded Double Consonants" ! viellja:vieljas Cx:0 <=> Vow: _ Cy j Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; ! mannji:manjis where Cx in (l d n) ! áddjá:ádjás Cy in (l d n) matched ; !€ vielljaX4 !€ vie0lja0 !€ mannjiX4 !€ ma0nji0 !€ áddjáX4 !€ á0djá0 "Gradation: Cluster Non-nasal Sonorant + Non-sonorant" ! láj'bi:láibbi ':Cx <=> Vow: [ l | r | đ | v | j: ] _ Cz Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in (f g d b k p t z ž c č s š) Cz in (f g d b k p t z ž c č s š) matched ; ! Added (¤:), evaluate! XXX. (Removed it since it caused CG in some PrsPrc.TO) !€ láj'biX4 !€ láibbi0 "Gradation: Cluster ŋ + Non-sonorant" ! seaŋ'ga:seaŋgga ':Cx <=> Vow: ŋ _ Cz Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; ! boŋ'kit:boŋkkii where Cx in ( g k ) Cz in ( g k ) matched ; !€ seaŋ'gaX4 !€ seaŋgga0 !€ boŋ'kiX4j# !€ boŋkki0i0 "Gradation: Cluster ft:ftt" ! sáf'ta:sáftta ':t <=> Vow: f _ t Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; !€ sáf'taX4 !€ sáftta0 "Gradation: Cluster bn:bnn" ! oab'na:oabnna ':n <=> Vow: b _ n Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; !€ oab'naX4 !€ oabnna0 "Gradation: Cluster m + Non-sonorant" ! lám'pa:lámppas ':Cx <=> Vow: m _ Cz Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; ! rán'ska:ránskkas where Cx in (f b p s š t) Cz in (f b p s š t) matched ; ! mt : mtt added bec. of ámta:ámttas, although not in grammar. Check norm! !€ ám'taX4 !€ ámtta0 ! This rule only applies to lexical entries written with non-standard n. ! The rule itself is ok, and requried for proper gradation of the lexical ! entries. Thus, the rule is not in itself !SUB, but the entries in the ! lexicon are, and should be marked accordingly. "Gradation: Cluster n + Non-sonorant" ! In'gá:Inggás ':Cx <=> Vow: n _ Cz Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in ( g k ) Cz in ( g k ) matched ; !€ In'gáX4 !€ Inggá0 ! The proper way of writing is Iŋgá/Iŋggás. We add a substandard pattern with n here, ! as we have seen it in use (56 Ingá, 18 Inggá!). "Gradation: Cluster n + Non-sonorant" ! gán'da:gánddat ':Cx <=> Vow: n _ Cz Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in (d t z ž c č s š) Cz in (d t z ž c č s š) matched ; !€ gán'daX4 !€ gándda0 "Gradation: Cluster p + t" ! konseap'ta:konseapttat ':t <=> Vow: p _ t Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; ! Added because of this word. The pattern is not in the grammar. !€ konseap'taX4 !€ konseaptta0 "Gradation: Cluster j/v + Liquid" ! áj'ru:áirrut ':Cx <=> Vow: [ j: | v ] _ Cz Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in (l r) Cz in (l r) matched ; !€ áj'ruX4 !€ áirru0 !€ biev'laX4 !€ bievlla0 "Gradation: Cluster r + l" ! áj'ru:áirrut ':l <=> Vow: r _ l Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; !€ jar'laX4 !€ jarlla0 "Gradation: Clusters lv, rj etc." ! ol'ju:oljjus ':Cx <=> Vow: [ đ | l | r ] _ Cz Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in (v j) Cz in (v j) matched ; !€ ol'juX4 !€ oljju0 "Gradation: Clusters tk, tm" ! mát'ki:mátkkis ':Cx <=> Vow: [ t | ŧ ] _ Cy Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in (k m) Cy in (k m) matched ; ! ŧ is missing in the grammar. example? !€ mát'kiX4 !€ mátkki0 "Gradation: Clusters ts, ks" ! 0:s <=> Vow: [ t | k ] _ s Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; ! I cannot find any such clusters. "Gradation: Clusters sk, sm etc." ! beas'ka:beaskka ':Cx <=> Vow: [ s | š ] _ Cz: Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in (b k p t l m v) Cz in (b k p t l m v) matched ; !€ beas'kaX4 !€ beaskka0 "Gradation: Cluster vj" ! čoav'ji:čoavjji ':j <=> Vow: v _ j Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; !€ čoav'jiX4 !€ čoavjji0 "Gradation: Cluster jv" ! beaj'vi:beaivvit ':v <=> Vow: j: _ v Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; !€ beaj'viX4 !€ beaivvi0 "Gradation: ihm, vhl etc." ! dujh'mi:duihmmit ':Cx <=> Vow: [ j: | v ] h _ Cy Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in (l m n) Cy in (l m n) matched ; !€ dujh'miX4 !€ duihmmi0 "Gradation: Clusters lbm, jdn, vdn etc." ! čuolbma:čuolmma Cx:Cy <=> [ l | j: | v | đ ] _ Cz Vow (Cns) ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in (b d g ) Cy in (m n ŋ) Cz in (m n ŋ) matched ; !€ čuolbmaX4 !€ čuolmma0 "Gradation: Clusters lm, jn, vn etc." ! Dálma:Dálmma ':Cx <=> [ l | j: | v ] _ Cy Vow (Cns) ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in (m n) Cy in (m n) matched ; !€ DálmaX4 !€ Dálmma0 ! "(Cns)" is added. I do not like it. But the rule as it was blocked the ! "olmmožin" essive of "olmmoš", since the "ž" was neither an underlying ! nor a surface stem consonant, I had to do something. The rule as it stands ! is meaningless, "(Cns)" contains both (StemCns:) and (:StemCns). ! Think through this. "Gradation: Clusters ldnj, vdnj etc." ! sávdnji:sávnnji d:n <=> [ l | v ] _ n j Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; !€ sávdnjiX4 !€ sávnnji0 "Gradation: Clusters rbm, rdn rgŋ" ! čorbma:čorpma Cx:Cy <=> r _ Cz Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in (b d g) Cy in (p t k) Cz in (m n ŋ) matched ; !€ čorbmaX4 !€ čorpma0 "Gradation: Clusters rdnj" ! skurdnji:skurtnji Cx:Cy <=> r _ Cz j Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in (b d g) Cy in (p t k) Cz in (m n ŋ) matched ; !€ skurdnjiX4 !€ skurtnji0 "Gradation: ist, vsk etc." ! návs'tu:návsttu ':Cx <=> Vow: [ j: | v | l | r | m | n ] s _ Cy Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in (k t) Cy in (k t) matched ; !€ návs'tuX4 !€ návsttu0 "Gradation: mšk, ršk" ! limš'ki:limškki ':k <=> Vow: [ m | r ] š _ k Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; !€ limš'kiX4 !€ limškki0 "Gradation: Clusters kc ks etc. 1" ! uk'sa:uvssa k:v <=> Vow: _ ': Cx Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in (t c č s š) ; !€ uk'saX4 !€ uvssa0 "Gradation: Clusters kc ks etc. 2" ! uk'sa:uvssa ':Cx <=> Vow: k: _ Cy Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in (t c č s š) Cy in (t c č s š) matched ; "Gradation: Cluster kst 1" ! teaks'ta:teavstta k:v <=> Vow: _ s ': t Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; !€ teaks'taX4 !€ teavstta0 "Gradation: Cluster kst 2" ! teaks'ta:teavstta ':t <=> Vow k: s _ t Vow: ( StemCns:) (:StemCns) ( %>: ) WeG: ; "Gradation: bb dd etc.1" ! spábba:spáppat Cx:Cy <=> Vow: _ Cz: ( %>: ) Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in (b d g z ž) Cy in (p t k c č) Cz in (b d g z ž) matched ; !€ spábbaX4 !€ spáppa0 "Gradation: bb dd etc.2" ! spábba:spáppat Cx:Cy <=> Vow: :Cy _ ( %>: ) Vow ( StemCns:) (:StemCns) ( %>: ) ; where Cx in (b d g z ž) Cy in (p t k c č) matched ; !$ Szczecin !$ Szccecin !€ Szczecin !€ Szczecin "Gradation: dn to tn, etc." ! eadni:eatni Cx:Cy <=> Vow: _ Cz: Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; where Cx in (b d g ) Cy in (p t k ) Cz in (m n ŋ) matched ; !€ eadniX4 !€ eatni0 "Gradation: dnj to tnj" ! boadnji:boatnji d:t <=> Vow: _ n j Vow ( StemCns:) (:StemCns) ( %>: ) WeG: ; !€ boadnjiX4 !€ boatnji0 ! Lengthening of consonants ! ------------------------- ! This takes place in certain verb forms, e.g. imperative and certain ! nominalisation forms. "Lengthening of Central Consonants hk etc." ! boah'tiY1:boahtti ':Cy <=> h: _ Cz Vow ( CntrCns:) ( %>: ) Len: ; where Cy in (p t k c č) Cz in (p t k c č) matched ; !€ boah'tiY1 !€ boahtti0 !€ dah'kaY7j !€ dahkku0i "Lengthening of Central Consonants pm etc." ! dapmaY1:dabmi Cx:Cy <=> Vow: _ Cz Vow ( CntrCns:) ( %>: ) Len: ; where Cx in (p t k ) Cy in (b d g ) Cz in (m n ŋ) matched ; !€ dapmaY1 !€ dabmi0 "Lengthening of Central Consonants tnj" t:d <=> Vow: _ n j Vow ( CntrCns:) ( %>: ) Len: ; !€ bitnjuY1 !€ bidnju0 "Lengthening of Central Consonants dj" ! dad'jaY1:daddji ':d <=> Vow: d _ j Vow ( CntrCns:) ( %>: ) Len: ; !€ dad'jaY1 !€ daddji0 ! Lenition ! -------- "Lenition: Intervocalic Stops and Affricates" ! joh'kaX4:jo00ga Cx:Cy <=> h:0 ':0 _ ( %>: ) Vow ( %>: ) ; where Cx in (k p c č t) Cy in (g b z ž đ) matched ; !€ joh'kaX4 !€ jo00ga0 ! Here, the k is reduced to g as an automatic result of the h deletion in the ! "Gradation: h loss" rule. !"Three-part compounding: Consonant geminates" ! giella:giel ! Cx:0 <=> Vow: _ Cy: Vow WeG: ; ! where Cx in (l r) ! Cy in (l r) ! matched ; !"Three-part compounding: Consonant geminates II" ! giella:giel ! Vx:0 <=> Vow: Cx Cx _ ; ! where Vx in (a e i o u) ! Cx in (l r) ; !"Three-part compounding: -iv- consonant center" ! oaivi:oai ! v:0 <=> i: _ Vow WeG: ; !"Three-part compounding: -iv- consonant center II" ! oaivi:oai ! i:0 <=> i: v:0 _ WeG: ; !!!!!!!!!!!!!! Start of non-normative compound shortening section !!!!!!!!!!!!!!!!! KEEPIF(SHORTCOMP) ! Compound shortening starts here ! ------------------------------- "d-deletion in compounding" d:0 => _ ': j: Vow:0 W3: ; ! vuod'ja:vuoj- _ d j Vow:0 W3: ; ! áddjá:ádj- [ l | j: | v ] _ [ m: | n: | ŋ: ] Vow:0 W3: ; !geaidnu:geain- Vow: _ d: Vow:0 W3: ; ! gáddi:gád- "j:i in compounding" j:i => Vow: d:0 ': _ Vow:0 W3: ; ! vuoj:vuoi- Vow: _ [ b:0 | d:0 | g:0 ] [ n: | m: | ŋ: ] Vow:0 W3: ; !geajdnu:geain- "t-deletion in compounding" t:0 => _ n Vow:0 W3: ; ! vuotna:vuon- Vow: [ s | š ] ': _ Vow:0 W3: ; !Clusters st, št h: ': _ Vow:0 W3: ; !Cluster ht "Double:single cons. in compounding" ! káffe:káf(f)- Cx:0 => Vow: _ Cy Vow:0 W3: ; where Cx in (đ f l m n ŋ r s š ŧ v) Cy in (đ f l m n ŋ r s š ŧ v) matched ; "Preaspirated Geminates in compounding part 1" !áhčči:áhč(č)- Cx:0 => h: Cy: _ Vow:0 W3: ; where Cx in (k p t c č l r) Cy in (k p t c č l r) matched ; "Preaspirated Geminates in compounding part 2" !áhčči:áh(č)- Cx:0 => h: _ Cy:0 Vow:0 W3: ; where Cx in (k p t c č l r) Cy in (k p t c č l r) matched ; "Preaspirated Geminates in compounding part 3" !geahči:geaš- h:š => _ (':) č: (č:) Vow:0 W3: ; "l:0 in compounding" ! viellja:vie(l)lj- l:0 => Vow: _ l j Vow:0 W3: ; Vow: [ s | š ] ': _ Vow:0 W3: ; ! Clusters sl, šl Vow: [ v: | j: ] ': _ Vow:0 W3: ; !Cluster v/j + l "n:0 in compounding" n:0 => Vow: _ n j Vow:0 W3: ; ! mannji:ma(n)nj- [ l | j: | v ] [ b:0 | d:0 | g:0 ] _ Vow:0 W3: ; !jdn, vdn etc. "Non-nasal Son. + Non-son. compounding part 1" ! láj'bi:lái(b)- Cx:0 => [l|r|đ|v|j:] ': _ Vow:0 W3: ; where Cx in (f g d b k p t z ž c č s š) ; "Non-nasal Sonorant + Non-sonorant compounding part 2" ! láj'bi:láib(b)- ':Cx => Vow: [ l | r | đ | v | j: ] _ Cz Vow:0 W3: ; where Cx in (f g d b k p t z ž c č s š) Cz in (f g d b k p t z ž c č s š) matched ; "Non-nasal Son. + k compounding" ! guoj'ka:guoih- k:h => [l|r|đ|v|j:] ': _ Vow:0 W3: ; "Non-nasal Son. + p compounding" p:h => [l|r|đ|v|j:] ': _ Vow:0 W3: ; "Non-nasal Son. + t compounding" t:h => [l|r|đ|v|j:] ': _ Vow:0 W3: ; "Non-nasal Son. + c compounding" c:h => [l|r|đ|v|j:] ': _ Vow:0 W3: ; "Non-nasal Son. + č compounding" č:h => [l|r|đ|v|j:] ': _ Vow:0 W3: ; "Cluster ŋ + g in compounding" ! seaŋ'ga:seaŋg(g)- ':g => Vow: ŋ _ g Vow:0 W3: ; "':k in compounding" ':k => Vow: ŋ _ k Vow:0 W3: ; ! báŋ'ku:báŋk(k)- Vow: [ t: | ŧ: ] _ k Vow:0 W3: ; ! mát'ki:mátk(k)- Vow: [ s | š ] _ k: Vow:0 W3: ; ! beas'ka:beask(k)- "Cluster m + f in compounding" ':f => Vow: m _ f Vow:0 W3: ; "Cluster m + b in compounding" ':b => Vow: m _ b Vow:0 W3: ; "':p in compounding" ! lám'pa:lámp(p)- ':p => Vow: m _ p Vow:0 W3: ; Vow: [ s | š ] _ p: Vow:0 W3: ; !Clusters sp, šp "':s in compounding" ':s => Vow: m _ s Vow:0 W3: ; Vow: k: _ s Vow:0 W3: ; ! uk'sa:uvss- "':š in compounding" ':š => Vow: m _ š Vow:0 W3: ; Vow: k: _ š Vow:0 W3: ; "':t in compounding" ':t => Vow: m _ t Vow:0 W3: ; Vow: [ s | š ] _ t: Vow:0 W3: ; Vow: k:v _ t Vow:0 W3: ; Vow k:v s _ t Vow:0 W3: ; ! teaks'ta:teavstt- "Cluster n + Non-sonorant in part 2" ! gán'da:gánd(d)- ':Cx => Vow: n _ Cz Vow:0 W3: ; where Cx in (d t z ž c č s š) Cz in (d t z ž c č s š) matched ; "Cluster v/j + r in compounding 1" ! áj'ru:ái(r)- r:0 => Vow: [ v: | j: ] ': _ Vow:0 W3: ; "Cluster v/j + r in compounding 2" ! jáv'ri:jávr(r)- ':r => Vow: [ j: | v ] _ r: Vow:0 W3: ; ! "':l in compounding" ! ':l => Vow: [ j: | v ] _ l Vow:0 W3: ; ! biev'la:bievl(l)- ! Vow: [ s | š ] _ l: Vow:0 W3: ; !Clusters sl, šl ! biev'laW3 ! bievlla0 "Clusters lv, rv, đv in compounding 1" v:0 => Vow: [ đ | l | r ] ': _ Vow:0 W3: ; ! bal'va:bal(v)- Vow: [ s | š ] ': _ Vow:0 W3: ; !Clusters sv, šv Vow: j: ': _ Vow:0 W3: ; ! beaj'vi:beai(v)- "Clusters lj, rj, đj in compounding 2" ! ol'ju:ol(j)- j:0 => Vow: [ đ | l | r | v ] ': _ Vow:0 W3: ; "':v in compounding" ':v => Vow: [ đ | l | r ] _ v Vow:0 W3: ; ! bal'va:balv(v)- Vow: [ s | š ] _ v: Vow:0 W3: ; !Clusters sv, šv etc. Vow: j: _ v Vow:0 W3: ; ! beaj'vi:beaiv(v)- "':j in compounding" ':j => Vow: [ đ | l | r ] _ j Vow:0 W3: ; ! ol'ju:olj(j)- Vow: v: _ j: Vow:0 W3: ; ! čoav'ji:čoavj(j)- "':m in compounding" ':m => Vow: [ t: | ŧ: ] _ m Vow:0 W3: ; ! fát'mi:fátm(m)- Vow: [ s | š ] _ m: Vow:0 W3: ; !Clusters sm, šm "k:0 in compounding" k:0 => Vow: [ s | š ] ': _ Vow:0 W3: ; ! beas'ka:beas(k)- h: ': _ Vow:0 W3: ; !joh'ka:joh(k) "p:0 in compounding" p:0 => Vow: [ s | š ] ': _ Vow:0 W3: ; h: ': _ Vow:0 W3: ; "m:0 in compounding" m:0 => Vow: [ s | š ] ': _ Vow:0 W3: ; [ l | j: | v ] [ b:0 | d:0 | g:0 ] _ Vow:0 W3: ; ! čuolbma:čuol(m)- "b:0 in compounding part" b:0 => [ l | j: | v ] _ [ m: | n: | ŋ: ] Vow:0 W3: ; ! čuolbma:čuol(b)(m)- Vow: _ b: Vow:0 W3: ; ! spábba:spáb- "g:0 in compounding" g:0 => [ l | j: | v ] _ [ m: | n: | ŋ: ] Vow:0 W3: ; Vow: _ g: Vow:0 W3: ; "Clusters lbm, jdn, vdn etc. in compounding part 6" ŋ:0 => [ l | j: | v ] [ b:0 | d:0 | g:0 ] _ Vow:0 W3: ; "Clusters lbm, jdn, vdn etc. in compounding part 7" ! čuolbma:čuolm(m)- Cx:Cy => [ l | j: | v ] _ Cz Vow:0 W3: ; where Cx in (b d g ) Cy in (m n ŋ) Cz in (m n ŋ) matched ; "k:v in compounding" k:v => Vow: _ ': [t:|c:|č:|s:|š:] Vow:0 W3: ; ! uk'sa:uvs- Vow: _ s ': t Vow:0 W3: ; ! teaks'ta:teavst- "Cluster kc in compounding" ':c => Vow: k: _ c Vow:0 W3: ; "Cluster kč in compounding" ':č => Vow: k: _ č Vow:0 W3: ; "Clusters bb dd etc. in compounding 1" ! spábba:spább- Cx:Cy => Vow: _ Cz: Vow:0 W3: ; where Cx in (b d g z ž) Cy in (p t k c č) Cz in (b d g z ž) matched ; "Clusters bb dd etc. in compounding 2" ! spábba:spápp- Cx:Cy => :Cy _ Vow:0 W3: ; where Cx in (b d g z ž) Cy in (p t k c č) matched ; "Cluster zz in compounding" z:0 => Vow: _ z: Vow:0 W3: ; "Cluster ž in compounding" ž:0 => Vow: _ ž: Vow:0 W3: ; "Cluster bb in compounding 2" ! spábba:spáp- b:p => Vow: b:0 _ Vow:0 W3: ; "Cluster dd in compounding 2" d:t => Vow: d:0 _ Vow:0 W3: ; "Cluster gg in compounding 2" g:k => Vow: g:0 _ Vow:0 W3: ; "Cluster zz in compounding 2" z:c => Vow: z:0 _ Vow:0 W3: ; "Cluster žž in compounding 2" ž:č => Vow: ž:0 _ Vow:0 W3: ; "Cluster hc in compounding 1" c:0 => h: ': _ Vow:0 W3: ; "Cluster hč in compounding 1" č:0 => h: ': _ Vow:0 W3: ; "Cluster hk:g etc in compounding" ! johka : jot- h:t => _ (':) [p:0|t:0|k:0|c:0|č:0] ([p:0|t:0|k:0|c:0|č:0]) Vow:0 W3: ; "Cluster hp in compounding 2" p:t => h (':) _ (p:0) Vow:0 W3: ; "Cluster hk in compounding 2" ! johka : joht- k:t => h (':) _ (k:0) Vow:0 W3: ; "Cluster hc in compounding 2" c:t => h (':) _ (c:0) Vow:0 W3: ; "Cluster hč in compounding 2" č:t => h (':) _ (č:0) Vow:0 W3: ; ENDKEEP !!!!!!!!!!!!!! End of non-normative compound shortening section !!!!!!!!!!!!!!!!! !Vowel shortening !---------------- ! Nouns "Vowel Shortening in Compounds of Contract Stems" Vx:Vy <=> [ (Vow) Vow | á ] Cns:+ Cns: _ (StemCns:*) ( %>: ) [X4: X7:|X8:] ; where Vx in (u i) Vy in (o e) matched ; "Allegro Forms in Compounds" ! gáv'piX7:gáv0pe Vx:Vy <=> Cns:+ _ ( %>: ) (¤:) (StemCns:*) ( %>: ) X7: ; where Vx in (i u á) Vy in (e o a) matched ; !€ gáv'piX7 !€ gáv0pe0 !€ bassiX7 !€ basse0 ! Overlapping => constraint wrt. 'i:e', Optional Vow Shortening ! After short 1st Syllable ! The ¤: condition was added in order to account for VIVVA nouns. ! First, it was added as optional, with parentheses, but that did not ! work, for some strange reason. Then the parentheses were removed, and to our ! surprise both forms with and without ¤ worked. ! And parantheses in again. Sigh. !"Optional Vowel Shortening after Short 1st Syllable" ! Vx:Vy => [ (Hyph) #: | Cns ] [a|i|o|u] [ Cns Cns:+ | j: Cns:+ | h: ': Cns: ] ! Vx:Vy => [ (Hyph) #: | Cns ] [a|i|o|u] [ Cns Cns:+ | h: ': Cns: ] ! _ (StemCns:*) [ X7: | X8: ] ; ! where Vx in (u i) ! Vy in (o e) ! matched ; ! Added + to Cns:+ and made StmCns: optional in LC, for biv'du:bivdo ! Onset relevant? !"Shortening of á in Compounds after Long 1st Syllable 1" ! á:a <=> [ Vow Vow | á ] Cns:+ Cns _ StemCns:* [ X7: | X8: ] ; !"Shortening of á in Compounds after Long 1st Syllable 2" ! á:a <=> [ (Hyph) #: | Cns ] Vz Cns+ Cns: Cns+ _ StemCns:* [ X7: | X8: ] ! where Vz in (a i o u) ; !"Vowel lowering for buorre" ! Vx:Vy <=> [ Vow Vow | á ] Cns:+ Cns _ X8: ; ! where Vx in (u i) ! Vy in (o e) ! matched ; ! This rule was actually made for one word only, buorre:buoret ! It replaces the missing "| X8:" sequence that is commented out in the ! "Vowel shortening ... 1" rule above. Why X8: does not work there is unclear. ! It seems to work in certain other cases, though. !€ buorriX7# !€ buorre00 !€ buorriX8# !€ buo0re00 !"Vowel Shortening in Vowel-Final Compounds after Long 1st Syllable 1" ! Vx:Vy <=> [ Vow Vow | á ] Cns:+ Cns: _ X7: ; ! where Vx in (u i) ! Vy in (o e) ! matched ; ! This rule was difficult. Look at *)footnote 1 below for comments. ! *) footnote 1 ! Older, removed versions: ! Vx:Vy <=> [ Vow Vow | á ] Cns:+ Cns: _ StemCns:* X7: ; ! This one gave vuovden# as only possible form of vuovdin in compounds. ! Vx:Vy <=> [ Vow Vow | á ] Cns:+ Cns _ StemCns:* [ X7: | X8: ] ; ! the problem with this one was that it was not restricted to compounds !! ! The X7: | X8: line was commented out since no lexicon in sme-lex.txt ! pointed from X8 to R. This move should be checked for compounds later. ! The testcases are kept here to ensure that its workload is being taken care of. !€ várriX7gir'ku# !€ várre0gir0ku0 !"Vowel Shortening in Compounds after Long 1st Syllable 2" ! Vx:Vy <=> [ (Hyph) #: | Cns ] Vz Cns+ Cns: Cns+ _ StemCns:* [ X7: | X8: ] ; ! where Vz in (a i o u) ; ! where Vx in (u i) ! Vy in (o e) ! matched ; ! XXX Remainder! This rule is commented out since including it ! spoils the i:á alternation of verbs (boahtit:boađán, for some reason. ! This should be looked into. !"Vowel Shortening in Cns-Final Compounds" ! Vx:Vy <=> Cns:+ _ StemCns:+ X7: ; ! where Vx in (u i) ! example? ! Vy in (o e) ! matched ; ! This is for the vuovdin# and vuovden#, if vuovden# exists. ! If vuovden# etc. is not possible, this rule should be removed. !"Vowel Shortening in Compounds of Contract Stems" ! Vx:Vy <=> [ Vow Vow | á ] Cns:+ Cns: _ StemCns:* X4: X7: ; ! where Vx in (u i) ! Vy in (o e) ! matched ; !"Compulsory Vowel Shortening after ij" ! liigi : liigeruđaid ! Vx:Vy <=> [ (Hyph) #: | Cns ] i j:i Cns:* _ [ X7: | X8: ] ; ! where Vx in (u i) ! Vy in (o e) ! matched ; ! The rule is too general at the moment, since we do not have illu : illosáhka, ! Now, we have "Cns:*", the earlier version said "': Cns", and that was wrong, ! it did not give liidni > liidne-. Still, *illo- must be a different problem, ! since it has short vowel before the consonants. I leave the rule as it is now ! and reconsider the illu case. !€ lijgiX7#ruh'taX4jd# !€ liige00ru00đa0id0 !FOLLOWING TwO RULES COMMENTED OUT BECAUSE THEY ARE TOO CATEGORICAL, new rules under optional vowelshortening e:i, u:o !"Optional Vowel Shortening in Caritives" ! Vx:Vy => [ (Hyph) #: | Cns ] [a|i|o|u] Cns:+ _ ( DelCns:) X9: ; ! where Vx in (u i) ! Vy in (o e) ! matched ; ! Hmm. X9 is used only in the DenominalAdjsV lexicon., in front of ! heapmi, huvva, huhtti, hudda. !"Vowel Shortening in Caritives" ! Vx:Vy <=> [ Vow Vow | á ] Cns:+ _ ( DelCns:) X9: ; ! where Vx in (u i) ! Vy in (o e) ! matched ; ! Verbs "Vowel Shortening in Bisyllabic Sg2 Imperatives and Connegatives" Vx:Vy <=> Cns _ ( %>: ) Q6: ; where Vx in (u i) ! čuor'vu:čuorvvo Vy in (o e) ! boah'ti:boađe matched ; !€ čuor'vuQ6 !€ čuorvvo0 !€ boah'tiQ6 !€ boa00đe0 ! Adjectives "Vowel Shortening in Attributive Forms, V-final Adjectives" ! láj'kiW1s:láikkes Vx:Vy <=> Cns:+ _ ( %>: ) [ W1: | W2: ] s [(Hyph) [ # | #: ]|(Hyph)] ; ! ál'kiW2s:álkes where Vx in (u i) Vy in (o e) matched ; !€ láj'kiW1s# !€ láikke0s0 !€ ál'kiW2s# !€ ál0ke0s0 ! W1 for WeG Attr, W2 for non-CG Attr form. "Vowel Shortening in Attributive Forms, s-final Adjectives" ! váttisW1:váttes W2?? Vx:Vy <=> Cns:+ _ s ( %>: ) W1: [ Hyph | [ # | #: ] ] ; ! trying to get heajos-biila as well as heajosbiila where Vx in (u i) Vy in (o e) matched ; ! Vx:Vy <=> Cns:+ _ ( %>: ) s W1: ; ! (Hyph) #: ; ! orig !€ váttisW1# !€ váttes00 !€ head'jusW1- !€ hea00jos0- !€ head'jusW1# !€ hea00jos00 !€ váttisW1- !€ váttes0- ! The rule now works for all "s W1:" right contexts. This means that ess of ! guorus becomes guorrosin, as required. Check for side-effects. "Optional vowel-shortening i:e" i:e => Cns:+ _ (Cns:) ( %>: ) (X4:) W4: ; !in diminutive verbs, caritives "Optional vowel-shortening u:o" u:o => Cns:+ _ (Cns:) ( %>: ) (X4:) W4: ; !Diminutive geaffut:geafustit/geafostit, caritives ! Diphtong simplification ! ----------------------- "Diphthong Simplification in i-Stems before Suffixes Beginning with j:" ! Vx:0 <=> Vow _ Cns:+ [ i | e7:i ] X5: ; ! Cannot see why e7:1 here. Thj Vx:0 <=> Vow _ Cns:+ i ( %>: ) X5: ; where Vx in (e o a) ; ! goah'tiX5jd:go0điid ! Cf. alt 2 below... !€ goah'tiX5jd !€ go000đi0id ! Two possible strategies for fixing bug 56, the oahpaheaddjiid bug: ! ------------------------------------------------------------------ ! 1. lexical strategy: XX instead of X5, and marking in the lexicon ! PrsPtc then do not go to GOAHTI, but to OAHPAHEADDJI, and point to ! :XX GODII- ; ! weak gr dipth simpl !"Optional Diphthong Simplification in Strong-Grade i-Stems before Suffixes Beginning with j:" ! Vx:0 => Vow _ Cns:+ [ i | e7:e ] XX: ; !XX is just an example of how we might do it... ! where Vx in (e o a) ; ! goah'tiX5jd:go0điid ! pro: do not need to define G3. ! pro: Opens for lexical variation: moarsi, etc, PrsPtc, Comp must be marked XX or redirected ! con: have to build a set of new lexica, identical to GOAHTI, except for the ref to GODII. ! con: gives a lexical solution to a basically morphophonological problem ! 2. morphophonological strategy, the smj way. ! Define a G3 context, and integrate it into a further rule !"Optional Diphthong Simplification in Strong-Grade i-Stems before Suffixes Beginning with j:" ! Vx:0 => Vow _ G3 [ i | e7:e ] X5: ; !G3 to be defined... ! where Vx in (e o a) ; ! goah'tiX5jd:go0điid ! pro: linguistically the right approach, says intuition ! pro: no need for confusing new lexica ! con: defining G3 is a mess, at best ! We have optional dipht. simpl. only in G3-nouns and no cons grad. ! Going for 2: !"G12 Compulsatory Diphthong Simplification in i-Stems before Suffixes Beginning with j:" ! Vx:0 <=> Vow _ G12 [ i | e7:e ] X5: ; ! where Vx in (e o a) ; ! goah'tiX5jd:go0điid !"G3 Facultative Diphthong Simplification in i-Stems before Suffixes Beginning with j:" ! Vx:0 => Vow _ G3 [ i | e7:e ] X5: ; ! where Vx in (e o a) ; ! goah'tiX5jd:go0điid !Using this one while waiting for the G3 rule to be fixed. !"Diphthong Simplification in i-Stems before Suffixes Beginning with j:" ! Vx:0 => Vow _ Cns [':Cns|Cns] Cns:+ [ i | e7:e ] X5: ; ! where Vx in (e o a) ; ! goah'tiX5jd:go0điid KEEPIF(DIPHSIMPL) "Diphthong Simplification in Actor Nouns from i- and a-Stems (Optional)" ! vieh'ki > vihkkiin Vx:0 => Vow _ Cns:+ [ i: | a: ] ¤: ( %>: ) X5: ; ! viehkkiin where Vx in (e o a) ; ! Eastern, substandard ELSE ! Use this variant in normative contexts: "Diphthong Simplification in Actor Nouns from i- and a-Stems" ! vieh'ki > vihkkiin Vx:0 <=> Vow _ Cns:+ [ i: | a: ] ¤: ( %>: ) X5: ; ! Western and norm where Vx in (e o a) ; ENDKEEP !€ vieh'ki¤X5jn !€ vi0hkki00in "Diphthong Simplification in Certain Forms from u-Stems" Vx:0 <=> Vow _ Cns:+ [ u: | o7: ] ( %>: ) [ X2: | X3: | (DelCns:) X6: | Q8: | Y5: | Y6: ] ; where Vx in (e o a) ; KEEPIF(DIPHSIMPL) "Diphthong Simplification in Actor Nouns from u-Verbs (Optional)" Vx:0 => Vow _ Cns:+ u: ¤: ( %>: ) [ X3: | X5: ] ; ! Eastern, substandard where Vx in (e o a) ; ELSE ! Use this variant in normative contexts: "Diphthong Simplification in Actor Nouns from u-Verbs" Vx:0 <=> Vow _ Cns:+ u: ¤: ( %>: ) X3: ; ! Western and norm where Vx in (e o a) ; ENDKEEP "Diphthong Simplification in Diminutives etc. from Trisyllabics with 2nd Syll u/o" Vx:0 <=> Vow _ Cns:+ [ u: | o7: ] ( %>: ) ( NonMPDummy:) Cns:+ ( %>: ) X2: ; where Vx in (e o a) ; "Diphthong Simplification: Conditional in -alla- etc. Derivatives from Trisyllabics" Vx:0 <=> Vow _ Cns:+ u:o ( %>: ) ( NonMPDummy:) Cns:+ X1: ; where Vx in (e o a) ; "Diphthong Simplification in IllSg and Certain Conjug. Forms of a-Stems" Vx:0 <=> Vow _ Cns:+ a: ( %>: ) [ X3: | Q2: | Q3: | Y7: ] ; where Vx in (e o a) ; "Diphthong Simplification in Certain Conjugational Forms of i- and u-Stems" Vx:0 <=> Vow _ Cns:+ HighVow: ( %>: ) [ Q2: | Q3: | Q7: | Y7: ] ; where Vx in (e o a) ; "Diphthong Simplification: Optional before Deverbal -alla-, -adda-, -us" Vx:0 => Vow _ Cns:+ o7: ( %>: ) ( NonMPDummy:) Cns:+ X1: ; where Vx in (e o a) ; "Diphthong Simplification in u-Stems before Conditional Marker" Vx:0 => Vow _ Cns:+ u: ( %>: ) Q5: ; where Vx in (a e o) ; "Dipthtong simplification in the attribute forms of AD-adjectives" Vx:0 <=> Vow _ Cns:+ Vow ( %>: ) [ Y9: | X9: ] ; where Vx in (a e o) ; "Diphthong Simplification: Optional before Deverbal st+alla-, st+adda-" Vx:0 => Vow _ Cns:+ u:o ( %>: ) W5: ; where Vx in (e o a) ; ! Vowel Rising ! ------------ "Vowel Rising" Vx:Vy <=> Vow: (Cns:*) ( %^: ) Cns:+ _ ( CntrCns:0 ) ( %>: ) ( Dummy:+ - W9: ) j:i ; where Vx in (e o) Vy in (i u) matched ; ! There is a conflict between e7:i here and e7:á in the Metaphony ! chapter. (e7 is "e with dot below" from the dictionary). ! We need to either restrict the context in the M chapter, or ! in the Vowel Rising part. For now, we try out without the e7:i. ! Needed: Documentation on all the rules in the M chapter. ! where Vx in (e o e7) ! Vy in (i u i ) ! The original version had Vx = e,o for Vowel Rising and ! Vx = i,e7,u for General metaphony. Thus, the new feature is ! e7 for Vowel rising. !€ reŋ'koX2jd !€ reŋ0ku0id !€ basteX2j !€ basti0i !€ asibealle7X2j !€ asibealli0ide ! Note. This will fail for now. ! -bealle +sg.ill = -beallái, +pl. ill = -beliide, !€ LineX2j !€ Lini0i ! should be Linii, exactly, fine, and is as well. ! Metaphony ! --------- "General Metaphony" Vx:Vy <=> Cns _ Cns:* ( %>: ) [ X2: | X6: ] ; where Vx in (i e7 u) Vy in (á á o) matched ; !€ áhččiX2n !€ áhččá0n !"e7 Rising" ! Awaiting discussion. ! e7:i <=> Vow: Cns:+ _ ( CntrCns:0 ) ( Dummy:+ - W9: - X2: - X6: ) j:i ; "Metaphony in Derivatives -us, -alla, -adda 1" u:o => Cns _ ( %>: ) ( NonMPDummy2:) Cns:+ X1: ; ! Here I have a new set, NonMPDummy2, without X7 and X8, to avoid clash with ! the allegro shortening rule. Check for alternative solutions !!! "Metaphony in Derivatives -us, -alla, -adda 2" Vx:á <=> Cns _ ( %>: ) ( NonMPDummy2:) [ Cns: - j: ] (Cns) X1: ; where Vx in (i e7) ; "Illative Singular Metaphony" Vx:Vy <=> Cns ( %>: ) _ ( %>: ) X3: ; where Vx in (i a e7) Vy in (á i á ) matched ; "Illative Singular Metaphony2" !because of STAHTA: stahtai, commented out beacause -ai illative is only for iija-words Vx:Vy <=> Cns _ ( %>: ) W9: ; where Vx in (i e7) Vy in (á á ) matched ; ! Commented in again for the sake of Sijte. !€ stahtaX3j !€ stahti0i !€ Sij9teW9j !€ Sijte0i "Metaphony in Px Contexts of Actor Nouns of i-, a-, and u-Verbs" Vx:Vy <=> Cns _ ¤: ( %>: ) [ X2: | X6: ] ; where Vx in (i a u) Vy in (á á o) matched ; "Metaphony in Illative of Actor Nouns of i-, a-, and u-Verbs" Vx:Vy <=> Cns _ ¤: ( %>: ) X3: ; where Vx in (i a ) Vy in (á á ) matched ; !"Metaphony in Actor Nouns of i-, a-, and u-Verbs" ! Vx:Vy <=> Cns _ ¤: ( %>: ) [ X2: | X3: | X6: ] ; ! where Vx in (i a ) !u) ! Vy in (á á ) !o) ! matched ; "Metaphony in MUSH-Derivatives" i:á => Cns _ ( %>: ) m: u: š: š: ; "Potential Marker Metaphony" Vx:Vy <=> Cns _ ( %>: ) Q7: ; where Vx in (i u) Vy in (e o) matched ; ! Stem vowel alternations ! ----------------------- "Stem Vowel in Contracted Stems" ! fijdnisY5t:fiidnát Vx:Vy <=> Cns _ ( CntrCns:) ( %>: ) Y5: ; where Vx in (i a u) Vy in (á á o) matched ; !€ fijdnisY5t !€ fiidná00t "Stem Vowel before Frequentative d" ! oažžut:oččodit Vx:Vy <=> Cns _ ( %>: ) Q8: ; ! coahkut:coagostit where Vx in (u) Vy in (o) matched ; ! Note that this rule also triggers other contexts than frequentative -d- !€ oažžuQ8dit !€ o0ččo0dit !€ coah'kuX8stit !€ coa00go0stit !"Stem Vowel before Frequentative l" ! i:a <=> Cns _ Q9: ; ! This rule is not implemented, there is no Q9 in the sme-lex.txt file (yet). !"Stem Vowel before Reflexive s" ! Vx:Vy <=> Cns _ Y8: ; ! where Vx in (i u) ! Vy in (á a) ! matched ; ! This rule is not implemented, there is no Y8 in the sme-lex.txt file (yet). "Indicative Present Singular 3rd Final Vowel" ! jearraQ1:jearrá Vx:á <=> Cns _ ( %>: ) Q1: ; ! boahtiQ1:boahtá where Vx in (a i) ; !€ jearraQ1 !€ jearrá0 !€ boah'tiQ1 !€ boah0tá0 "Imperative Present 1st/3rd Person and Connegative Final Vowel" Vx:o <=> Cns _ ( %>: ) Q3: ; ! jearraQ3n:jerron where Vx in (a i u) ; ! boahtiQ3n:bohton ! čuorvuQ3n:čurvon !€ jearraQ3n !€ je0rro0n !€ jearraQ2t !€ je0rre0t !€ boah'tiQ3n !€ bo0h0to0n !€ čuor'vuQ3n !€ ču0r0vo0n "Stem Vowel Shift a:e, i:e, u:o" ! jearra:jerre Vx:Vy <=> Cns _ ( %>: ) Q2: ; ! boahtit:bohtet where Vx in (a i u) ! čuorvu:čurvo Vy in (e e o) matched ; ! This is for Prs Du1, Pl3, and 2nd Person Imperative !€ jearraQ2 !€ je0rre0 !€ boah'tiQ2t !€ bo0h0te0t !€ čuor'vuQ2 !€ ču0r0vo0 "Stem Vowel shift i:á" i:á <=> Cns _ ( %>: ) [ Q4: | Q5: ] ; !1st and 2nd Person Singular Present/before Conditional Marker Cns:+ _ ( %>: ) W5: ; !Dim-freq verbs Cns:+ _ ( %>: ) W6: ; !Sub-freq verbs !€ boah'tiQ4n !€ boa00đá0n !€ boah'tiQ5lin !€ boa00đá0lin "Stem Vowel shift u:o" u:o <=> Cns:+ _ ( %>: ) [ W5: | W7: ] ; !Dim-freq verbs, geaffut:geafostallat + stuvvat. "Stem Vowel shift i:e" i:e <=> Cns:+ _ ( %>: ) W7: ; !!Dim-freq verbs, geaffut:geafostallat + stuvvat. Cns:+ _ ( %>: ) W8: ; ! Deverbal nouns -meahttun !"Stem Vowel before Conditional Marker" ! boahtit:boađálin ! i:á <=> Cns _ Q5: ; !Collapsed into one rule, see above ! Vx:Vy <=> Cns _ Q5: ; ! The u:o alternation ! where Vx in (u i) ! removed because it gave ! Vy in (o á) ! *čuorvvolin, but Nickel ! matched ; ! gives čuorvvulin. "Stem Vowel in Dual 2nd Person Imperative" ! jearra:jearri a:i <=> Cns _ ( %>: ) Y1: ; !€ jearraY1 !€ jearri0 "Stem Vowel in Present Participials from i-Verbs" a:i <=> Cns _ ¤: ( %>: ) [ X1: | X4: | X5: | X7: | X8: | (Hyph) [ # | #: ] | Hyph | s | n ] ; ! a:i <=> Cns _ ¤: ( %>: ) [ X1: | X4: | X5: | X7: | X8: | (Hyph) #: ] ; ! The problem with the old one was that it didn't trigger for Essive and for ! the Px forms of the nominative. ! The new one isn't perfect, the s and n are rather quasi for Ess and Px. "Stem Vowel in Dual and Pl 1st Person Imperative" ! jearra:jearru Vx:u <=> Cns _ ( %>: ) Y2: ; ! boah'ti:boahttu where Vx in (a i) ; !€ jearraY2 !€ jearru0 !€ boah'tiY2 !€ boahttu0 "Stem Vowel in Plural 2nd Person Imperative" ! Vx:Vy <=> Cns _ ( %>: ) Q2: ; ! boah'ti:bohtet Vx:Vy <=> Cns _ ( %>: ) Y3: ; ! čuor'vu:čurvot where Vx in (a i u) Vy in (e e o) matched ; ! Q2 context replaced by other rule for both imperative and other categories !€ jearraQ2t !€ je0rre0t !! boah'tiY4t ! It seems it should be Q3. ... both?! !€ boah'tiQ3t !€ bo0h0to0t !! čuor'vuY4t ! Q2, it seems. !€ čuor'vuQ2t !€ ču0r0vo0t "Stem Vowel in Imperative Plural 1st Person and Passive Stem" Vx:o <=> Cns _ ( %>: ) [ Y4: | Y7: ] ( j u v v o | j u n | j u p m ) ; where Vx in (a i u) ; ! Y7 Testing passive forms !€ jearraY7t !€ je0rro0t !€ boah'tiY7t !€ bo0htto0t !€ čuor'vuY7t !€ ču0r0vo0t !€ jearraY7juvvot !€ je0rro0juvvot ! Conflict: a:{o,u} <=> Cns _ Y7: {j: / unspecified} !"Stem Vowel in Past Tense of Derived Passive Stem" ! Vx:u <=> Cns _ Y7: (Y6:) j: ; ! where Vx in (a i) ; ! This battlefield will be cleaned up, eventually, but we leave it as-is ! for a while, so that Thomas and others can have a look at the ruins. ! New oldrule "Stem Vowel in Past Tense of Derived Passive Stem" Vx:u <=> Cns _ ( %>: ) Y7: (Y6:) j: [[ # | #: ]|m|d|g] ; where Vx in (a i) ; ! for jerrui. !€ jearraY7j !€ je0rru0i !€ dah'kaY7j !€ dahkku0i !€ loikaY7j !€ loiku0i ! Old rule ! "Stem Vowel in Past Tense of Derived Passive Stem" ! Vx:u <=> Cns _ Y7: Y6: j: ; ! where Vx in (a i) ; ! Put Y6 into paranthesis, otherwise o:u stem vowel shift doesn't work in Passive+Ind+Prt+Sg3 of ! Impersonals and intransitives. (TO) ! But we don't have passives of impersonals and intransitives, do we ?! So, what was ! the reason for changing the rule? (TT) yes we have passives of impersonals and intransitives:-) (TO): !arvit+V+IV+Pass+Ind+Prt+Sg3 !arvojuvvui !arvui ! Stem vowel deletion in compounding ! --------------------------------------------- "á-stem vowel deletion" á:0 <=> Cns: _ W3: [ # | #: ] ; "a-stem vowel deletion" a:0 <=> Cns: _ W3: [ # | #: ] ; e: _ b: b: o: ( %>: ) [X2:|X3:|X6:]; !Diphthong Simplification in -eabbo comparatives "i-stem vowel deletion" i:0 <=> Cns: _ W3: [ # | #: ] ; "u-stem vowel deletion" u:0 <=> Cns: _ W3: [ # | #: ] ; "e-stem vowel deletion" e:0 <=> Cns: _ W3: [ # | #: ] ; "o-stem vowel deletion" o:0 <=> Cns: _ W3: [ # | #: ] ; ! Special word final consonants ! ----------------------------- "Word Final Deletion of n8 m8 g8 h8" ! beatnag8:beana Cx:0 <=> Vow _ ( %>: ) [ X4: | X8: | X6: Cns: | X9: | W4: ] ; ! luopmán8:luomi where Cx in (DelCns) matched ; ! giellum8:gielu !€ beatnag8X4 !€ bea0na00 !€ luopmin8X4 !€ luo0mi00 !€ giellum8X4 !€ gie0lu00 "Word Final Neutralization of g8, h8, m8" Cx:Cy <=> Vow _ ( %>: ) [ X2: | X7: ] (Hyph) [ # | #: ] ; where Cx in (g8 h8 m8) Cy in (t t n ) matched ; !€ !€ ! Appendix: Deleted rules that still need attention ! ------------------------------------------------- !"Postvocalic j Surfacing" ! j:i <=> Vow: ( :0 ) ( %>: ) ( Dummy: ) _ [ Cns | ': | ( :0 - Y5: - Y6:)+ (Hyph) [# | #:] ] ; !more and more dummies... !! j:i <=> Vow: ( :0 ) ( Dummy: ) _ [ Cns | ': | ( :0 - Y5: )+ (Hyph) #: ] ; !more dummies... !! j:i <=> Vow: ( :0 ) ( Dummy: ) _ [ Cns | ': | ( :0 - Y5: ) (Hyph) #: ] ; !1.138 !! j:i <=> Vow: (:0) (Dummy:) _ [(Dummy:* - Y5: -Y6: ) [ (Hyph) #: | Cns: ((Hyph) #:) ]]; ! bad bec. Ess almmájin *almmáiin almmájX4in !! j:i <= _ [Cns | ( DummyNotY5: ) #: ] ; ! <============ 1.1 version: !!!j:i <=> Vow: (:0) (Dummy:) _ [Cns | ': | (:0 - Y5:) [ Hyph #: | #:]] ; !! Changed because:we get almmáj- and not almmái- !! Postvocalic j surfaces as i !! Is this what we want?? without right context??? postvoc j:i <=> Vow: ( :0 ) (Dummy: ) _ ; ! "Word Final Consonant Neutralization 1" !smirezit : smires, Troandin-bisma ! Cx:Cy <=> Vow: (( CntrCns:) Dummy:+) _ ( %>: ) ( :0 - Y5: - Y6: )( :0 - Y5: - Y6: ) [ Hyph | [ # | #: ] ] ; ! ! Cx:Cy <=> Vow: (( CntrCns:) Dummy:+) _ ( :0 - Y5: )( :0 - Y5: ) [ Hyph | #: ] ; !1.138 ! ! Cx:Cy <=> Vow: (( CntrCns:) Dummy:+) _ ( Dummy: ) ( :0 - Y5: ) [ Hyph | #: ] ; ! ! Cx:Cy <=> Vow: (( CntrCns:) Dummy:+) _ ( Dummy: - Y5: - Y6: ) [ :0 - Y5: - Y6:] [ Hyph | #: ] ; ! 1.155 ! ! Cx:Cy <=> Vow: (( CntrCns:) Dummy:+) _ [ ( Dummy:* - Y5 ) [ (Hyph) #: ]] ; ! ! Cx:Cy <=> Vow: (( CntrCns:) Dummy:+) _ [ ( Dummy:* - Y5: ) [ (Hyph) #: ]] ; ! ! Cx:Cy <=> Vow: (( CntrCns:) Dummy:+) _ [ ( Dummy:* - Y5: - Y6: ) [ (Hyph) #: ]] ; ! not vejolaš ! where Cx in (z ž m h p g b d) ! Cy in (s š n t t t t t) ! matched ; !"Word Final Consonant Neutralization 2" ! This is no longer a mystery. For once, we ask for the #: symbol in the ! left context. Why we do, is a mystery, but at least now the example works. ! We should thus consider whether the leftmost # is needed in the rule. !€ #hálijd# !€ 0háliit0 !$ hálijd# !$ háliid0 !"Stem Final Consonant Neutralization" !háliidit : háliit ! Cx:Cy <=> Vow: Cns* (Vow*) Vow: ( CntrCns:) _ #: ; ! where Cx in (z ž m h p d g b) ! Cy in (s š n t t t t t) ! matched ; ! ! Old version: ! Cx:Cy <=> Vow: (( CntrCns:) Dummy:+) _ ( :0 - Y5: ) [ Hyph | #: ] ; ! This one works for all except háliit. ! ! Tentative version: ! Cx:Cy <=> Vow: ( CntrCns:) (Dummy:+) _ ( :0 - Y5: ) [ Hyph | #: ] ; ! The nested left context was changed because of háliidit - háliit. ! NOTE: there are 454 jd# that come out incorrectly with this version. ! Overview over j translations. ! j:j # | Cns _ ! this is default ! j:i Vow _ Cns ! Most common exception ! j:0 Vow _ [ Y5: | Y6: j: ] ! Contracted stems ! j:j Vow _ Vow ! if followed by vow-initial suff ! *) footnote 1 ! Older, removed versions: ! Vx:Vy <=> [ Vow Vow | á ] Cns:+ Cns: _ StemCns:* X7: ; ! This one gave vuovden# as only possible form of vuovdin in compounds. ! Vx:Vy <=> [ Vow Vow | á ] Cns:+ Cns _ StemCns:* [ X7: | X8: ] ; ! the problem with this one was that it was not restricted to compounds !! ! The X7: | X8: line was commented out since no lexicon in sme-lex.txt ! pointed from X8 to R. This move should be checked for compounds later. !"Vowel Shortening in Compounds after Long 1st Syllable 2" ! Vx:Vy <=> [ (Hyph) #: | Cns ] Vz Cns+ Cns: Cns+ _ StemCns:* [ X7: | X8: ] ; ! where Vz in (a i o u) ; ! where Vx in (u i) ! Vy in (o e) ! matched ; ! XXX Remainder! This rule is commented out since including it ! spoils the i:á alternation of verbs (boahtit:boađán, for some reason. ! This should be looked into.