Li huanian Linguis ic Ac
This Regula ion shall en e in o o ce on he wen ie h day ollowing ha o i s publica ion in he O icial Jou nal o he Eu opean Union.
Con en s linked o Tu in e e ence
ISSN 2669-218X (online o elec onically)
Copy igh ed by E aldas Š age is in he yea 2025. Published by he Ins i u e o he Li huanian Language. This is
an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License, which
un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal au ho and sou ce a e
allows c edi ed. Published by he Li huanian Language Ins i u e. This a icle is open access, dis ibu ed unde
he e ms o a C ea i e Commons A ibu ion-Sha eAlike License ha allows unlimi ed use, dis ibu ion, and
ep oduc ion o he con en on any medium, including au ho s and sou ce.
Recei ed by: Gau a on 11 Ap il 2025 Accep ed: Accep ed by: 2025-06-02.
1
The main poin s o he epo a e as ollows: (a) The
composi ion o he Li huanian Pa liamen ; (b) The composi ion o he Li huanian Pa liamen ; (c) The composi ion o he Li huanian Pa liamen ; (d) The composi ion o he Li huanian Pa liamen .
E ec s o Consonan Voicing on Vowel Fo man Dis ibu ion in S anda d
Li huanian: A Fac o ial Analysis
This Regula ion shall en e in o o ce on he wen ie h day ollowing ha o i s publica ion in he O icial Jou nal o he Eu opean Union.
Ins i u e o Li huanian Language
The El. The Commission shall be empowe ed o adop delega ed ac s in acco dance wi h A icle 26 o supplemen his Regula ion by laying down de ailed ules o he applica ion o his Regula ion.
ORCID ID: h ps://o cid.o g/0009-0002-5624-7275 Resea ch a eas include expe imen al
phone ics, he p osody and ocalism o Bela usian languages.
This Regula ion shall en e in o o ce on he wen ie h day ollowing ha o i s publica ion in he O icial Jou nal o he Eu opean Union.
------------------------------------------------------------------------------------------------------------------
ANOTATION This a icle is a con inua ion o he expe imen al dynamic s udies o he phenomena o ocalism in he
Li huanian language (posi ion o i s a ia ion). The ocus is on ac ions ha di ec ly a ec he o ma i e
s uc u e o oices, in e ela ionships and hie a chies. In addi ion o he empi ical da a, samples o sounds wi h
loud explosi e sounds we e used o analyze he o e all e ec o he loudness ac ion on he F1 and F2 ou pu s in
he spec um. Fo his pu pose, a ac o da a analysis (ANOVA) was ca ied ou in o de o de e mine he
hie a chical posi ion and po en ial o he o ing ac o and i s abili y o combine wi h o he ac o s. The esul s
e ealed ha he weak posi ions o he ac o in ques ion in hie a chical chains indica e a ai ly low le el o i s
au onomy. I can be seen ha i s ela i e powe is co ela ed wi h he a icula ed owel sounds, and he loude
he oice, he b igh e he di e ences in he o ming o he sca dum ac ion. In addi ion, al hough i does no
in oduce g ea e di e en ial powe , his ac o expands he a icula o y-spec al space o compe ing o ces
o exe g ea e o ce (e.g., by bending he uppe do sal ins in on o he posi ion be ween he den al
p o usions).
ESMINIAI WORDS: oice o man s, common Li huanian language, ac o analysis, hie a chy o ac o s,
posi ional a ian s o oices.
This is a e y impo an poin .
This s udy aims o con inue expe imen al esea ch on he dynamics o owel o man dis ibu ion in S anda d
Li huanian. The p ima y ocus is on he in e ac ion and hie a chy o ac o s ha di ec ly in luence o ma i e
2
placemen in he spec um o Li huanian owels. A e CVC combina ions wi h oiced plosi e consonan s we e added o
he da ase , he di e en ial powe o he oicing ac o was analyzed. A ac o ial analysis ( wo-way ANOVA) was
pe o med o de e mine wha posi ion in he hie a chy he oicing ac o can ake. Addi ionally, he ques ion o ac o
coo dina ion in he ield o o man dis ibu ion was aised. The esul s sugges ha he oicing ac o occupies weak
posi ions in he hie a chy and exhibi s low-le el au onomy. To some ex en , i s ela i e powe co ela es wi h he
backness o he owel, as i shi s owa ds he back posi ion o he ongue, di e ences implied by he consonan - oicing
ac o on he o ma i e dis ibu ion inc ease. Al hough i has no signi ican impac , consonan oicing plays a ole in
expanding he spec al space, allowing o he ac o s o exe g ea e in luence ( o example, i c ea es condi ions o
high-back owels o be mo e on ed in a den al s op en i onmen ). KEYWORDS: owel o man s, S anda d Li huanian,
ac o ial analysis, ac o hie a chy, owel allophones. This is he i s ime ha he e m has been used in Li huanian.
The Commission shall be empowe ed o adop delega ed ac s in acco dance wi h A icle 26 o supplemen his Regula ion by laying down de ailed ules o he applica ion o his Regula ion.
The i s a emp s o analyze in mo e de ail he dynamics o he a ia ion o he ocalism o he
common Li huanian language by expe imen al means, mo e p ecisely, he ac o s de e mining he
phone ic s uc u e o he owels, essen ially ga e posi i e esul s (Š age is 2024; Š age is,
U bana ičienė 2024). Al hough he limi ed scope o empi ical ma e ial and he small numbe o
ac o s included in he analysis ha e so a p e en ed a - eaching conclusions, i can be no ed
ha he dynamics o he plo o quan i a i e and quali a i e cha ac e is ics o o es is no
chao ic (see simila concep s in o eign esea ch, e.g. Lindblom, 1994: 40 […] 55; Roesle , 2018: 535 […]
540; mo e om a heo e ical pe spec i e, see Lindblom, 1990: 65 […] 100). I can be seen ha he
ac o s a e linked by hie a chical ela ionships, and he spec al cha ac e is ics o posi ional
ballo a ian s can be deduced om he co ela ions de e mined by s a is ical calcula ions. Fo
example, he mo e open he oice is, he mo e i s equency spec um depends on he p osodic
kuo užpakalesnis, uo yškiau o man inė s uk ū a signalizuoja a ikuliacinį g e a jo esančio
ac o (in e ms o he posi ion o he ph ase c oss), he ype o explosi e coun e pa , e c. In
addi ion, he esul s e ealed ha he p ac ice o isola ing he ocal co ds om he acous ic
signals o he spoken language can dis o he ends conside ed no mal. Unde minimum
condi ions, he in luence o he subjec i e dic iona y e b is o en inc eased (possible hype ac i i y; o his ac o see Johnson and o he s 1993: 701[…]714; Lee and o he s 2019:
1568-1579; om he pe spec i e o he Bal ic languages, see In his a icle, we'll look a some o
he mos common misconcep ions and misconcep ions abou language, such as he di e ences
be ween long and sho owels, and we'll look a some o he mos common misconcep ions abou
language, such as he di e ences be ween long and sho owels. Fi s o all, he di ec ion o
2024).
he analysis i sel could in he long un expand and supplemen he discou se o expe imen al
Li huanian ocalism esea ch and p o ide a basis o compa a i e wo k wi h analogue o eign
language ma e ial (Hillenb and and k . 2001: 754755; Roesle , Yung 2018: 535540). On he o he hand, i
he inc ease in he numbe o ac o s and he possibili y o inc easing he empi ical da a
esul ed in e en clea e and mo e conc e e esul s, he p edic abili y and modelling o he
cha ac e is ics o posi ional o ing op ions (applicable aspec ) would inc ease. In o he wo ds,
om he poin o iew o applied speech syn hesis, i is use ul o imp o e he model ha
classi ies he cha ac e is ics o posi ional oice a ian s so ha he acous ic s uc u e o
ją iesiogiai eikiančių ak o ių są yšių pe spek y os. Papildoma al e na y a p ognozuo i
3
speech sounds can con ibu e o he imp o emen o he na u alness o he
compu e -gene a ed human oice. All his seems o be su icien incen i e o con inue he in es iga ions ini ia ed.
The p oblem o his expe imen al s udy a ises om he need o inc ease he analysis o he
quali a i e spec a o ac o s di ec ly a ec ing he common Li huanian language oices, numbe s.
I is unde s ood ha he mo e ac o s a e included in he model, he mo e comp ehensi e and
uni e sal he esul s should be. Since his is essen ially he con inua ion o he wo k begun
(Schwage 2024; Schwage , U bana ičienė 2024), he ocal ac o o he explosi e explosions is added
o he ea lie s ages o he s udy ( he dic iona y, he phonological quan i ies, he a icula ed
explosions o he ype and he ph ase c ossing). In he expe imen al s udies o he Li huanian
language, he possible impac o his ac o on he dis ibu ion o speech o ms seems o ha e no
ye been de ailed. Indi ec connec ions can only be seen in indi idual wo ks and mo e so when alking
abou he coa icula ion o language sounds (Amb aze ičius, Leskauskai ė 2014: 202225; Ja osla ienė
and o he s 2019: 62167; om he pe spec i e o o he languages, see Kewley-Po 1982). Fo eign
esea che s poin ou ha , om a compa a i e poin o iew, he lowe ing o he i s o man is
cha ac e is ic o he o es cas among he lowes o es. The dis ibu ion o he on ow owels
is mo e a ec ed by his e ec (F1 dec eases by an a e age o ~100 Hz), while he middle ow and
back ow owels a e much weake (~20 Hz; see S e ens, House 1963: 111128; Hillenb and and k . 2001:
754755; also Ad iaans 2018: 2125). T ue, he g ea e oo p in in he spec um is le by he p e- ocal,
no he pos - ocal, posi ions. Resea che s assume ha his co ela ion is mos likely a spec al
e lec ion o he a ia ion in he posi ion o he h oa s ( he h oa is lowe ed below he ibs when
sc a ching). In gene al, hese con igu a ions appea o be ela ed o he dynamics o he change in
subglo al and sup aglo al p essu e equi ed o (ne) s imula e he phona ion o he ocal olds
(Amb aze ičius, Leskauskai ė 2014: 171). In o de o main ain he phona ion o as long as possible in
he explosion phase o he explosi e p obes, i is necessa y o educe he sup aglo al p essu e
(inc ease he olume o he sup aglo al ocal ac ), so he lowe ing o he h oa s seems clea ly
mo i a ed om his poin o iew (S e ens 1998: 465-485). In addi ion, i he con igu a ion o he ocal
ac we e o be de ailed acco ding o he logic o pe u ba ion heo y, i should be no ed ha he
ocal ac no only expands bu also expands a he h oa (and his is ano he a o able condi ion
o he sup aglo al p essu e o dec ease). The i s o m can occu in wo cases: when he oice is
labialized (because he i s esonan modes o pu iness a e loca ed a he lips) o when he ocal
ac sp eads o he h oa ( his is he localized mode o his node; see Chiba, Kajiyama 1941; Ken ,
Read 2002: 28[…]32).
Thus, he pu pose o his a icle is o de e mine by expe imen al means he powe o he
ela i e explosi e oice ac o o a ec he gene al dis ibu ion o Li huanian language oice
o ma ions. I is ue ha he e is li le doub ha he in luence o his ac o will di e
signi ican ly om ha ound in simila s udies o o he languages om he poin o iew in
ques ion. I is a phone ic uni e sali y, de e mined solely by he human ana omy and he
phonological easons o sound gene a ion (i.e. he condi ions necessa y o phona ion o
4
explode o he sound e ec s), a he han a linguis ically speci ic exp ession (S e ens 1998:
465-485). The e o e, as al eady men ioned, he expansion o ac o analysis i sel ( he addi ion
1
o ano he ac ion) is mo e ele an o his s udy, a e which i should be cla i ied whe he
hei hie a chical ela ions ha e changed wi h he inc ease in he numbe o ac ions. Fo
example, i would be use ul o ind ou whe he he coa seness ac o becomes a se ious
obs acle o di e en ia ing he o man s uc u e o he balls acco ding o hose ac o s
whose in luence, based on ea lie da a, is mo e closely ela ed o he di e ences o he i s
o man . The o e all dynamic pic u e o he cha ac e is ics o quali a i e o ing is impo an
no only in e ms o he powe o he ac o s hemsel es, bu also in e ms o hei di ec ions
o ac ion and compa ibili y, and all hese objec i es de e mine he b oade objec i es o he s udy:
1.Expand and supplemen he empi ical ma e ial o he s udy (Š age is 2024; Š age is,
U bana ičienė 2024) wi h da a classi ied acco ding o he o ing ac ion o he o e s (i.e.
eco d and include in he s udy all iden ical posi ions wi h he complaining explosi e
o e s).
2. o compa e by s a is ical means he i e ac o s […] dic ion, phonological quan i ies,
ph ase c ossing, p onuncia ion o owels and […] ela i e powe o owels o a ec he
o ma i e s uc u e o owels, as well as o de ail he di ec ions o hei ac ion and
compa ibili y.
3. supplemen and, i necessa y, e ise he hie a chical ac o chains (i.e. ex end he
model i sel ) es ablished a an ea lie s age o he s udy, o de e mine he posi ion o
he o ing ac o in hem.
A b ie o e iew o he esea ch, empi ical ma e ial and me hod
As his s udy is a con inua ion o he wo k s a ed, he same me hodological guidelines (Š age is
2024; Š age is, U bana ičienė 2024) a e in p inciple ollowed. The mos ecen wo k on he subjec
has al eady been discussed in mo e de ail, so i is only possible o b ie ly e e he e o wha has
eme ged om he la es esul s o ac o analysis (Schwage 2024). As is known, a ious
linguis ic and ex alinguis ic ac o s in luence he quan i a i e and quali a i e cha ac e is ics
o owels ( om a gene al pe spec i e on he in luence o a ious ac o s on he spec al
cha ac e is ics o owels in Li huanian). In he ligh o he o egoing conside a ions, he
Commission concludes ha he con es ed measu e cons i u es S a e aid wi hin he meaning o
A icle 107 (1) o he T ea y. See also o he language s udies: Johnson e al. 1993: 701[…]714;
Hillenb and e al. 2001; Tsukada 2009). The equency spec um o he ocal o ma ions (how he
oice will be a icula ed) depends on he pi ch and sequence o he oices, phono ac ic and p osaic
language ac ions, indi idual a icula ions o he speake , e c. Howe e , a mo e de ailed acous ic
pic u e o he oices can be c ea ed when all hese ac o s a e iewed om an in e ac i e
pe spec i e, when objec i e c i e ia a e used.
1
Recall ha his ac o analysis (s a is ical me hods) aims o cla i y he ela ionship and hie a chy o ac ions
ha de e mine he a ia ion o language sounds.
5
he compa a i e weigh o each o hem is assessed. The da a hus analysed indica e ha he
phonological quan i y occupies he highes hie a chical posi ions in e ms o he in luence on F1
and F2 p onuncia ion in he mos gene al case (Schwage , 2014). Howe e , his conclusion loses i s
2
powe o p oo when he a icula o y subg oups o he oices a e analyzed sepa a ely. When he
back ow oices si a he epicen e o he analysis, he phonological leng h o he owels is
pushed ou o he dominan posi ion by he explosi e ac ion o he owels. This means ha ,
wi hou g ouping he da a, i is possible o dis inguish he mos likely g e a oices [ɑː] and [uː]
om he mos likely g e a oices [p], [k] and [ ]. Mo e speci ically, he g ea es quali a i e change
in he spec um o he pos e io in e io [ɑːɑ] is de e mined by he a icula ion o he explosi e
on o he jaw (p[ɑː]p di e s signi ican ly om he k[ɑː]k and [ɑː] a ian s due o he F2
lowe ing on a e age by 200 Hz). The quali a i e spec um o he pos e io high-pi ched [uː ] di e s
mos om o he posi ional a ian s in ha i is su ounded by he ongue's p e-ocula den al [ ]
(on a e age he same 200 Hz inc eases F2; his phenomenon is also conside ed phone ic uni e sali y;
see Hillenb and and k . 2001). The e is a s ong s a is ical basis o se e al o he co ela ions: he
openness o he oices and he leng h o he ph ases ( he mo e open he oice, he mo e
dependen he o ma i e s uc u e is on i s posi ion in he ph ase), he ise o he p e- oice and
he phonological quan i y ( he highe he p e- oice and he highe he ise, he mo e clea ly he
o ma i e dis inguishes be ween he sho and long a ian s), he sha pness o he a icula ion
and he weigh o he dic iona y ( he mo e dis an he a icula ion is om he cen e o he
acous ic spec um). Cases o ope a ion o one-sided ac o s mus also be no ed in he esidual
su plus alue. Appa en ly, he labializa ion o open owels be ween he owel sounds is o ganically
compa ible wi h he di e ences in he implici [ɑː ] openness o he ph ase c oss. Wha 's mo e,
he ela ionship be ween he la e and he phonological quan i y in his case, because, as he da a
shows, hei in luence canno g ow alongside each o he […] as he i s in luence inc eases, he
second one ends o dec ease a he same ime (which means ha hei di ec ions o ac ion a e
no coo dina ed).
In o de o be able o compa e he la es da a wi h he p e ious ones, a s ic selec ion o empi ical da a mus be
ollowed. The i s hing o say is abou he dic iona y. I o he dic iona ies we e eco ded a his s age o he
s udy, he la e ac o would o cou se need o be elimina ed om he analysis (as he da a would no be di ec ly
compa able). Fo his eason, he same 38-yea -old (D1) and 20-yea -old (D2) men who can be conside ed as sui able
ep esen a i es o he common Li huanian language subs anda d a e in i ed o eco d addi ional ma e ial
(Miliūnai ė 2018: 53). In he analysis, we will choose he igonome ic CVC- ype compounds wi h he so-called "co ne
owels" [iː ɪ], [uː ] and [uː ɑ ], co esponding o he posi ion o he ongue a hes om he neu al (Ti ze: 162 […]
163). Al oge he , he explosi e consonan s [p], [ ], [k], [b], [d], [] and he sho consonan s [ɪ], [] and [] and he long
consonan s [iː], [uː] and [ɑː] o m 36 compounds wi h which he sho iadic ph ases (e.g., […] a k KAK aip[…];
[…] a k BŪB aip[…]) a e o med. Addi ionally, by a ying hem wi h o he ac o s o be analyzed, i should be
emphasized he e ha he s a is ical calcula ion in mind is based on he ela i e ad an age o he phonological
quan i y o e he emaining h ee ac o s (ph ases, syllables, explosions and dic iona ies). As he numbe o
2
ac o s o be analyzed inc eases, hei hie a chical a angemen may na u ally change.

6
In he case o a sample posi ion o a ph ase in ela ion o a c oss, 108 clea en ies we e
p epa ed, which bo h dic iona ies we e asked o ead wen y imes ( i e a a ime). A o al o
4320 cases we e eco ded. The eco dings we e made in he Phone ics Labo a o y o Vilnius
Uni e si y wi h p o essional equipmen , keeping he usual equi emen s o expe imen s o his
na u e. All eco ded connec ions a e scanned, anno a ed and p ocessed wi h he sound analysis
sk ip u au oma iškai išgau i y imui eikalingi duomenys – cen inių balsių dalių pi mųjų ijų
p og am P aa . This is a special case o he F1, F2, and F3 o ming alues (he cais) wi h which he
wo- ac o dispe sion analysis (TWO-WAY ANOVA, Rie el, an Hou 2005) was u he calcula ed.
Fi s , his s a is ical model e alua es he powe o he p o-choice o ing ac o o
di e en ia e da a in he gene al case […], i.e. when he empi ical da a is no g ouped by o he
ac o s in any way ( he esul s a e illus a ed in h ee-dimensional g aphs). A e ha , ha ing
c ea ed he condi ions o di ec compe i ion (when compe ing ac o s a e o a ed in places),
he same model calcula ed he hie a chical chain o ac ions o each a icula ed ype o ballo .
Papildomomis g a inėmis p iemonėmis šįka analizuo as eiksnių sude inamumas. Šiuo
a eju skaičiuo a ne ik, kiek ka ų eiksniai p anoks a ieni ki us (p ocen ine iš aiška), be
also eco d hose cases whe e, despi e he supe io i y o one o hem, he da a a e
success ully sepa a ed by bo h di ec ly compe ing ac o s. i he c i ical p = 0,05 h eshold is
no exceeded in bo h cases o compa ison, he e ec s o bo h ac o s on he equency
spec um a e conside ed o be compa ible. The di ec ions o ac ion a e also illus a ed wi h
addi ional g aphic means (by selec ing he neu al e e ence poin o he ac ion). These
g aphical da a a e p esen ed in o de o p o ide maximum de ail on he dynamic o ma i e
shi s and o link hem mo e clea ly o a icula o y changes in o es ( o all o he de ails o he
su ey me hodology, see S age is 2024; S age is, U bana ičienė 2024).
3. Resul s o he s udy. E ec s o low back pain on low back pain
[ɑː/a] o he dis ibu ion o o mulas
I is op imal o s a he ac o analysis wi h he ques ion o wha s a is ical p obabili y in he
gene al da a se (ou o a o al o 1440 cases) i is possible o dis inguish be ween he quali a i e
[ɑː] a ian s ound be ween he sc a ches and he explosi e explosions. Unde s andably, i he
da a a e no g ouped in mo e de ail, he homogenei y ( iabili y) o he samples compa ed will
su e and he eliabili y o he s a is ical esul s will dec ease. Howe e , as a as can be
judged om he o e all g aphical layou o he da a (see diag am 1) and he es ima ed p obabili y
es ima e (p = 1,72765E-12), he e seems o be su icien eason o belie e ha he p e- o e ac o
con ibu es signi ican ly o he di e en ia ion o he quali a i e spec um o he back- o-back
o es. Al hough i is possible o see he di e ences be ween he wo o ms in he
h ee-dimensional g aphical space, he mean alues clea ly indica e ha he posi ional a ian s o
his ype o sound a e dis inguished only by he i s o m (be ween F1 = 714 Hz, F2 = 1298 Hz,
be ween F1 = 634 Hz, F2 = 1283 Hz; he di e ence is ~ 80 Hz). Essen ially, he phone ic uni e sali y
men ioned in he in oduc o y pa o he a icle ( he illing o he h oa wi h he phona ion o
exci e → F1 decline) is es ablished. Howe e , i should be added ha he esul s o simila s udies
7
in o he languages o back oices sugges ed a much smalle and in p inciple s a is ically
insigni ican e ec o his ac o ( he di e ences o he i s o man eached only 15 o 20 Hz;
Hillenb and e al. 2001). Diag am numbe one. Lowe back [ɑː ] dis ibu ion o he i s wo
o ma ions acco ding o he ac ion o he p ebalsial owel (p = 1,72765E-12)
3
In addi ion, despi e he alleged end in s a is ical calcula ions, ano he de ail o he dis ibu ion
o g aphic da a caugh he eye. Va ia ion e o s o une en o ma ions a e obse ed, which do
no e lec he meanings o he mean. The ed ellipse appea s o limi [ɑː] o he F2 posi ion in he
spec um. I is logically assumed ha he adi ionally mo e in ense a icula ion o he explosi e
su oca ing ongues (Picke 1999: 125) in hese cases is co ela ed wi h a lowe lexibili y o he
ho izon al ongue posi ion, and he sp ead o he glo al a ea ( he common F1 p og ession) in
samples wi h sc a chy ongue open he way o somewha g ea e F2 changes in he ongue.
Wha is he di ec ion o ac ion o he dus -dus ac ion and, mo e impo an ly, how i co ela es
wi h he co esponding ec o s o o he ac ions can be seen in diag am 2. The las poin (F1 and F2
alues) illus a es he la ges eco ded quali a i e di e ences in he oices o he g oup in
ques ion o each o he ac ions. Fo example, acco ding o he dic iona y ac o , he highes
s a is ical es ima e (p = 1,6E-14) was he long-s emmed [ɑː] samples wi h he sho -s emmed [d]
samples. The o ming a e ages o his pa icula case a e shown in he diag am (D1 F1 = 614 Hz, F2
= 1414 Hz: D2 F1 = 672, F2 = 1173 Hz). Following he same p inciple, add analogous o he ac ion poin s.
Such analy ical access o da a c ea es a isual imp ession o he limi s o a ia ion o spec al
cha ac e is ics ha posi ional [ɑː ] a ian s can achie e o ac ions wi h maximum o ce
3
Ta p dusliųjų p iebalsių – audonoji elipsė, a p ska džiųjų – mėlynoji (bend asis duomenų masy as).
8
(de i ing he highes s a is ical p obabili y o di e ence).
Diag am numbe wo. The e ec o he po en ial ac o s (spec al space) is a condi ional
e e ence poin chosen om he i s wo o man s o he neu al beam (F1 = 500 Hz, F2 = 1500
Hz; he black do on he diag am), he e ec o he ac o s on he quali y o he beam spec um
can be assessed h ough he p ism o he a icula (non) ol age (F1 and F2 shi om he
di ec ion o he cen e o he acous ic spec um). In his iew, he coas al en i onmen ac o
s ands ou mo e clea ly, di e en ia ing he spec um o he g ea es ex en (p = 1E-28)
essen ially only acco ding o he o dinal and axis (because o he b igh labializa ion in he join s
wi h he loops o he coas s), while all o he ac o s ac in he same di ec ion o
decen aliza ion. I he en i e s udy we e limi ed o his g aphene ma e ial alone, i could be
concluded ha in hese pai s some o hei membe s a e simply less a icula ed (less ense). I is
na u al ha he di ec ion o cen aliza ion is mo e displaced han he long a ian s o he sho
a ian s, bu i is no expec ed ha hei s a is ically la ges di e ence (p = 2.1E-22) is ound in
he posi ional posi ion in he sample wi h he sca e ed gummy bea s (D2, mean: […][[…]][…] F1 = 591
Hz, F2 = 1332 Hz; […][[…]][…] F1 = 667 Hz; F2 = 1166 Hz). Acco dingly, u he away om he cen e o he spec um (p = 1E-14) be ween he sc a ches, and no
a p dusliųjų p iebalsių a sidū ę balsiai (D2; poki inė pozicija; idu kiai: k[ɑː]k F1 = 629 Hz,
F2 = 1326 Hz; ɡ[ɑː]ɡ F1 = 667 Hz, F2 = 1164 Hz). Tokio pa ies pobūdžio akus inė ski is
(p = 1,6E-14) is also be ween dic o ies (poki inė pozicija, he long d[ɑː]d; mean: D1 F1 = 614
9
F2 is 1414 Hz. F1 = 672 Hz, F2 = 1173 Hz), and be ween he posi ions o he ph ase in e ms o he
in e sec ion (p = 7E-18; D2, long k[ɑː]k; mean: in e sec F1 = 782 Hz, F2 = 1228 Hz; poki iniai F1 = 629
Hz, F2 = 1326 Hz). I is ue ha he spec al exp ession and he s a is ical weigh o all hese
di e ences a e no co ela ed. The s a is ical powe o he phonological quan i ies and he
a icula o y ype o he consonan s is g ea e han ha o all he o he s, bu his ad an age
is no always accompanied by he g ea es g aphical dis ance be ween he poin s. Again, his
g aphene plo is jus an illus a ion o he mos s iking di e ences. I should be bo ne in mind
ha some imes he acous ic indica o s di e sligh ly o no a all acco ding o all hese
ac o s, so he in o ma i eness o hese da a is no absolu e. Each balsio phone ic ealiza ion is
one o ano he esul o he in e ac ion o he ac ions o he VIS, so he added alue o his
diag am is only he highligh ing o he po en ials o hei di ec ions o ac ion. The la e should
cla i y he possibili ies o compa ibili y o he ac o s a ec ing he ocal o ma i e
s uc u e. The dominan posi ion o he o e s is also con i med by he analysis o di ec
ac o compe i ion. In he case o he iangle, he inc ease in da a does no weaken he e ec
o he [p] and [b] labializa ions on he open-backed low-pi ched owels. When he numbe o pai s
di ec ly compa able in his aspec inc eased om 16 o 44, he pe cen age exp ession o he
ad an age o his ac o o e he o he s emained he highes (79 pe cen ; see diag am 3). This
means ha in a leas ou cases ou o i e, by analyzing he mixed da a, i is ai ly eliable o
say which o he explosi e esonances (ei he be ween he lungs o be ween he lungs o lungs)
con ained he oices o he subg oup. This is also clea ly demons a ed by he a e ages o he
o e all da a se ([p]/[b] F1 = 689 Hz, F2 = 1178 Hz, [ ]/[d] F1 = 675 Hz, F2 = 1337 Hz; [k]/[] F1 = 675 Hz, F2 =
1357 Hz; gian in e ac ion p obabili y p = 1,8E-122). The whole se o indica o s shows ha , in he
mos common case, he ocal old clea ly occupies he highes posi ion in e ms o he in luence
on he o ma i e s uc u e o he lowe back oices. The hie a chical posi ions o he
emaining ac ions a e much mo e p oblema ic. Ob iously, he inc eased olume o da a has
inc eased he powe o he dic iona y ac ion. Appa en ly, he join s o he balls wi h he scaly
ibs (compa ed o he dus s) a e a mo e a o able medium o indi idual a icula ion. Pe haps i
is no coincidence ha he di ec ions o he la e wo ac o s a e e y simila (see diag am 2).
When only ou ac o s we e compa ed (Schwage 2024), i was qui e di icul o dis inguish one
dic iona y om he o he in he i s wo o ma ions (abou 23 pe cen ), and now his
p obabili y is wice as high (55 pe cen ). The gene al a e age (wi hou he spike D1 F1 = 736 Hz F2 =
1326 Hz; D2 F1 = 692 F2 = 1269 Hz; wi h he spike D1 F1 = 682 Hz F2 = 1340 Hz; D2 F1 = 672 F2 = 1239 Hz)
assumes ha he indi idual spike expanded due o wo easons: he s onge e ec o he
spike on he D1 glo al memb ane (F1 dec eases) and he mo e in ensi e ac ion o he D2 in hese
a eas (D2 dec eases). The weakening o he c oss posi ion o he ph ase is mos likely a
consequence o he inc eased o ing powe o he p o- and dic iona y ac ions. The c i e ion o
phonological quan i y emains he weakes in his whole con ex . The p obabili y o sepa a ing
sho [] om long [ɑː] is he lowes (38 pe cen ), al hough o he sake o ce ain y i should be
said ha he lag o his e b om he c oss o he ph ase and he sca ci y o e bs does no
exceed he s anda d e o limi s.
16
Fu he mo e, i can be concluded ha he compa ibili y o he ac ion and i s ac ual powe
a e no di ec ly co ela ed. To be e unde s and his connec ion, we need o del e e en
deepe in o he a icula o y and spec al dynamics o oices (in o he wo ds, mo e o he
le el o mic o-a icula ion). One o he bigges obs acles o ge ing he F1 and F2 a ian s is
he explosi e coils. The mos o ganic in his espec is he combina ion o p o e b and ph ase
c osses ( he compa ibili y o his pai is […] 75 pe cen ), and he mos di icul o dis inguish
be ween Gomu explosi es and non-Gomu explosi es is he da a no g ouped by dic iona y (55
pe cen ). Howe e , he o e all compa ibili y indices o he phonological quan i ies and
dic iona y ac o s occupying he highes hie a chical posi ions a e simila (46% and 40%
espec i ely). The ph ase ki is occupies he middle posi ion (52 pe cen ) on his indica o .
5. E ec o he p e-balanced owel on he dis ibu ion o he uppe
pos e io [uː ] o ma ions
Finally, he spec al cha ac e is ics o he uppe pos e io [uː ] acco ding o he connec ions
wi h he explosi e sound p essu e-dus seem o occupy an in e media e posi ion be ween he
wo p e iously discussed sounds. Pe sky a, which di ides he gene al mass o da a in o wo
g oups acco ding o his c i e ion, has a no iceable s a is ical basis (p = 1.95E-07), bu he
a e ages (be ween su oca ion […] F1 = 372, F2 = 885 Hz; be ween sc a ches […] F1 = 387, F2 = 1019
Hz) do no show any a icula ela ionship wi h he glo aline egion (F1 is essen ially s able). The
s a is ical di e ence weigh is ep esen ed by he second o man (di e ence ~ 100 Hz), which
assumes he con igu a ional dynamics o he an e io o al ca i y. I is clea ly seen om he
gene al e o s ha he sca i ied owels (as in he case o he pos e io lowe -limbed owels)
become an addi ional s imulus o he loose ho izon al ongue mo emen s. In addi ion, i he
ellip ical o e lap in h ee-dimensional space is no me ely a coincidence o a consequence o he
non-homogenei y o empi ical da a (al hough his p obabili y should heo e ically be educed by
he la ge amoun o da a), i should be a gued he opposi e o wha he heo e ical discou se
assumes […] he lowe F1 alues o his g oup may mo e o en each he p oximi y o he
su oca ing (and no he explosi e!) Thus, al hough he esul o he di e en ia ed exp ession o he da a is a o able, i s in e p e a ion is s ill qui e p oblema ic.
The di ec ions o ac ion o he ac o s [uː ] in he spec al space a e p obably he mos mixed. I
is no enough o ocus on a (de)cen aliza ion ec o and i is necessa y o no e whe e he e is a
g ea e sp ead o da a. Since he pi ches o he wo axes a e he same (e e y 100 Hz), i is easy o
see ha he posi ional a ian s o he oices di e mo e in e ms o o dinal (F2). Acco ding o
he size o ec o s and hei s a is ical weigh , he place o a icula ion o owels and he
phonological quan i y ac o s s and ou om he o he s. Thei di ec ions o ac ion, a leas
based on his g aphene ma e ial, ob iously do no coincide. The i s o hese ec o s illus a es
he phone ic uni e sali y, no o men ion he b igh F2 ele a ion o o eign esea che s nex o
he explosi e ongue's on al den al p o usions. The g ea es con as o his kind (p = 1E-60) is
eco ded in he opposi e posi ion be ween he sc a ches o he ibs (D2, sho owels […] mean:
[…][[…]][…] F1 = 397 Hz, F2 = 1022 H; b[]b F1 = 413, F2 =
943 Hz; o d[u]d F1 is equal o 403 Hz, F2 is equal o 1394 Hz.

17
9 he diag am. Pos e io high-pi ched [uː ] dis ibu ion o he i s wo o ma ions acco ding o
he ac ion o he p e-ballo o es (p = 1.95E-07)
5
10 he diag am. Di ec ions o ac ion o po en ial ac o s (spec al space)
5
Ta p dusliųjų p iebalsių – audonoji elipsė, a p ska džiųjų – mėlynoji (bend asis duomenų masy as).
18
The phonological quan i y is qui e s ongly pushing he unca ed a ian s o hese sounds close o he
epicen e o he spec um (D2, opposi e posi ion, explosi e sc a ch- o- oo h en i onmen […] mean: d[]d F1 =
403 Hz, F2 = 1394 Hz; d[u]d F1 = 329 Hz, F2 = 712-45 Hz; p = 4E). Howe e , i is unusual ha he ph asing o he same
ec o is no main ained by he c oss ac o and seems o gi e way o he in luence o he adjec i es (D2,
he sho , explosi e sound o he den al adjec i es […] a e age: adjec i es F1 = 403 Hz, F2 = 1394 Hz; F1 is 459
Hz and F2 is 1293 Hz. I can be no iced ha D1 ends o cen alize he lowe limbs in he mos common case,
poki iniai F1 = 431 Hz, F2 = 1290 Hz; p = 3E-20). Tendencijos b uožų įgyja
indi idualiosios a ikuliacijos ski umai. Palyginus a i inkamus isų ijų balsių ek o ius,
while D2 ends o cen alize he uppe limbs. The la e is illus a ed by he g aph o his sec ion (posi i e
posi ion, sho oices, a mosphe e o su oca ing explosi e up u es […] mean: D1 F1 = 370 Hz, F2 = 671 Hz; D2
F1 = 399 Hz, F2 = 883 Hz; p = 5E-19). Meanwhile, he spec al slide o he non-sha ing acco ding o he sca ci y
c i e ia mos e ec i ely es i ies o he non-sha ing o he ac o i sel . The expec ed F1 dec ease is
ixed nex o he su oca ing oices, and he s a is ical signi icance o he di e ence i sel is
subs an ially educed by he b igh F2 inc ease (D1, posi ional posi ion, sho oices […] a e age: [] F1 = 345
Hz, F2 = 973 Hz; d[]d F1 = 372 Hz, F2 = 1268 Hz). I is wo h no ing ha in he mos p onounced opposi ions (i.e.
hose wi h he g ea es s a is ical signi icance) acco ding o he ph ase c oss, he ac ion o sca ci y and
quan i y was ep esen ed by he den al edges, so he in luence o he edges and he ec o di ec ion i sel
is no excluded. I should be emembe ed ha all di ec ions o ac ion o hese ac o s a e only o ien a ional
[…] maximum o hei in luence on he balsio [uː ] o man s uc u e illus a ion. The ac o s a e cons an ly
linked by hie a chical ela ionships, so he ec o s shown in his g aph a e mos likely s ongly adjus ed by
hei in e ac ion mechanism ( hey canno in p inciple ac in isola ion, independen ly o each o he ). The 11 h
diag am. E icien ela i e powe in hei di ec compe i i e condi ions (F1 and F2 di e en ia ion)
19
In di ec compe i ion wi h each o he , all he ac o s unde conside a ion qui e clea ly posi ion
hemsel es in ela ion o each o he (see diag am 11). A i s glance, i would seem ha he
scena io seen in he co esponding [iː ɪ] sec ion is epea ed, i.e. by supplemen ing he da a mass,
he sca ci y ac o does no acqui e he s a us o au onomy and pe cei ed powe , bu expands
he a icula o y-spec al space o o he ac o s o ac wi h g ea e o ce. Howe e , his
lowes hie a chical posi ion does no allow him o hink so. Acco ding o his c i e ion, he
weakes is he ph ase ki is, which con ibu ed he leas o he F1 and F2 dis ibu ions and
when only ou ac o s we e analyzed (changes o 18% o 20%; Schwage 2024). I emains unclea
whe he he powe o he phone ic uni e sals ( he den al esonance en i onmen […] [uː […]] F2
heigh ) is p esen in his dynamic plo ( he a e age o he o al esonance en i onmen […] F1 =
376 Hz, F2 = 803 Hz; he o al esonance en i onmen […] F1 = 383 Hz, F2 = 756 Hz; he den al
esonance en i onmen […] F1 = 375 Hz, F2 = 7512 Hz; p = 2.63-835 Hz) o no . A leas he meanings o
he meanings es i y ha he new ac o became an addi ional incen i e o enhance he e ec
o he supp ession o he uppe back ee h be ween he den al ee h. Compa ed o he
p e ious su ey, he posi ion o he quan i a i e ac o inc eases (57 pe cen o 71 pe cen ).
Such a change makes i e en mo e plausible o a gue ha he in luence o phonological leng h on
he placemen o he i s wo o ms is co ela ed wi h he sliding o he owels owa d he
on ow and upwa d. I is belie ed ha e en highe F2 (in he mos common case 956 Hz o 1012
Hz) be ween he den al c acks expanded he space o he spec al di e en ia ion o hese
sounds in e ms o quan i y. The dic iona y posi ion emains s able. Al hough, o cou se, his
ac o canno be equa ed wi h he leng h o he ocal co ds o he a icula o y ype o he
explosi e, hei combined in luence on he quali a i e spec um emains qui e signi ican (46%).
The lowes a ing o he o ing powe index in his con ex . All p e ious indica ions would be
ha such a ela i ely high a ing (35 pe cen ) has no a ional basis. Appa en ly, as al eady
men ioned, he solu ion o his ques ion equi es a mo e de ailed analysis o mic oa icula ion
phenomena. I is now possible o say oughly ha , pe haps ac ing in he same di ec ion as he
o he s (p ima ily wi h he p eposi ion ac o and he phonological quan i y), his ac o
main ains a ce ain posi ion, and i s in luence is co ela ed wi h he backwa dness o he
owels […] he u he away om he p eposi ion, he mo e no iceable he e ec o he
poo ness-slowness dis ibu ion F1 and F2. The place is le → he phonological quan i y → he
dic iona ies → he sca ci y → he ph ase ki is. While he dynamics o he quali a i e spec um
p idė i, kad hie a chinis isų šių eiksnių išsidės ymas a odo aip: p iebalsių a ikuliacijos
o pos e io uppe - igh owels is he mos di icul o analyze, he in e ac ion compa ibili y
g aph illus a es i well (see diag am 12). The e is no doub ha he [uː ] spec um and
a icula ion a e he mos p oblema ic means o ha monizing ac ions. Only in e y a e cases
does he pe cen age o compa ibili y exceed he 50 pe cen h eshold. Obse e he po en ial o
coo dina e you powe be ween he o ces o balance and dic a o ship (67 pe cen ) and he
o ces o balance and quan i y (58 pe cen ). All o he pai s emain below he hal -pe cen
h eshold. Fo his eason, he g aphic space looks e y emp y, and he da k zone co e s a small
pa o he ada .
20
I we look a he common a e ages, he sequence o ac o s acco ding o he compa ibili y
c i e ia is as ollows: he place o a icula ion o he a gumen s (51 pe cen ) → he dic iona ies (36
pe cen ) → he se e i y (30 pe cen ) → he quan i y (28 pe cen ) → he p onuncia ion o he ph ase
(24 pe cen ). In e es ingly, he e ha e been cases whe e compa ibili y was no possible in p inciple
( he in e ac ion be ween he numbe and he c oss o he ph ase is ze o in his espec ).
Appa en ly, he a icula ion o he speci ics o his ype o oice balances he in e ac i e
connec ions by selec i ely enhancing he powe o indi idual ac o s. Theo e ically and p ac ically,
he co ela ion be ween he ela i e powe o he epea ed p e- oice ac ion and he o de o he
sp ogs amųjų p iebalsių a ikuliacinį ipą), ku ią pas a ieji duomenys ik da aiškiau pa i ina.
Vis dėl o, a ei yje į isus pas ebė us polinkius eikė ų paž elg i giliau.
oices is mo e ele an ( he mo e consonan he oice, he b igh e he o man s uc u e sepa a es he 12 diag ams nex o i ). Compa ibili y o ac o s (spec al space)
O he , o a a con en , by weigh
1.The analyzed es ma e ial allows he ollowing hie a chical chains o ac ion ( om
s onges o weakes ) o be o med acco ding o he s a is ical weigh ing o he e ec s on
he ocal o ma ion s uc u e:
[ɑː] place o a icula ion o he consonan s → dic iona ies → ph ases ki is →
sca ci y → phonological quan i y;
[iː ɪ] […] phonological quan i y […] dic iona y […] place o a icula ion o he
consonan s […] ph ase ki is […] sca ci y;
[uː ] […] place o a icula ion o he p eposi ions → phonological quan i y →
dic iona y → sca ci y → ph ase ki is.
21
The inclusion o a new oice ac o in he analysis did no subs an ially change he co ela ions
ound in he ea lie s ages o he s udy. The in luence o he ype o exploding owels on he F1
and F2 dis ibu ion emained co ela ed wi h he olume o owels ( he lowe he olume, he
g ea e he in luence), he phonological quan i ies […] wi h olume and ele a ion ( he close he
[i] ype a icula ion, he b igh e he di e ences be ween he sho and long owel spec a),
he c oss o he ph ase […] wi h he openness o owels ( he mo e open he oice, he g ea e
he in luence o he p osodic e b). The ela i e powe o he owel ac ion o he owels, o
example, depends on he a icula ion slope o he owels owa ds he back ow ( he b igh e i is, he mo e no iceable he di e ences in he o ming o his ac ion).
The weakness o he sca ci y ac o in posi ions in hie a chical chains implies a low le el o i s
independence. This could be one o he easons o he a he i egula di e ences in he
placemen o F1 and F2 acco ding o his ac o . Acco ding o o he languages, he phone ic
p onuncia ion o […] be o e he owels F1 su ounding he lowe case owels […] o he common
Li huanian language is no con i med. This hypo hesis is pa ially suppo ed only by he indica o s
o he lowe back oices. I is mo e easonable o belie e ha , by supplemen ing he da a se and
no in oducing a mo e signi ican in luence, he sca ci y ac o expands he
a icula o y-spec al space o compe ing ac ions o exe a g ea e o ce, o example, o
aise he uppe pos e io balls in on o he posi ion be ween he den al explosions. Howe e , i
does no ule ou he possibili y ha he e may be di e ences in mic o-a icula o y sounds
ela ed o he ac o in ques ion, which a e p e en ed om being disce ned by he excessi e impac o o he ac o s on he quali a i e spec um.
Addi ional ac o compa ibili y analysis e ealed ha he mos a o able medium o
coo dina ing he e ec s o all o hem is he spec al space o he lowe pos e io sphinc e . In
o he wo ds, he mo e open he oice, he easie i is o econcile he e ec s o he ac o s
shaping hei o ma i e s uc u e. This should mean, among o he hings, ha he space o F1
and F2 a ia ion o open oices is he wides .
LITERATŪRA
Ad iaans F ans 2018: E ec s o consonan al con ex on he lea nabili y o owel ca ego ies
om in an -di ec ed speech. – The Jou nal o he Acous ical Socie y o Ame ica 144(1),
20–25. DOI: h ps://doi.o g/10.1121/1.5045192.
Amb aze ičius Ry is, Leskauskai ė As a 2014: P iebalsių akus inės ypa ybės: pala alizacija i
balsingumas [Acous ic P ope ies o Consonan s: Pala aliza ion and Voicing], Kaunas:
Technologija.
Chiba Tsu omu, Kajiyama Masa o 1941: The Vowel: I s Na u e and S uc u e, Tokyo:
Kaiseikan.
Čei ane Sol eiga, Ind ičāne Inese 2012: La iešu alodas oksneņu aks u ojums pēc lokusa
ienādojumiem [Cha ac e iza ion o La ian Language Sounds Using Locus Equa-
ions]. – Bal is ica 47(1), 37–50. DOI: h p://dx.doi.o g/10.15388/bal is ica.47.1.2132.

22
Hillenb and James, Cla k Michael, Nea ey Te ance 2001: E ec s o consonan en i onmen
on owel o man pa e ns. – The Jou nal o he Acous ical Socie y o Ame ica 109(2),
748–763. DOI: h ps://doi.o g/10.1121/1.1337959.
Ja osla ienė Ju gi a 2014: Spec al Cha ac e is ics o he L ihuanian Vowels: Some
P elimina y Resul s o a New Expe imen al Resea ch. – Linguis ica Le ica 22, 68–84.
Ja osla ienė Ju gi a 2015: Lie u ių kalbos umpųjų i ilgųjų balsių kiekybės i kokybės
e alonai [S anda ds o Quan i y and Quali y o Sho and Long Vowels in he Li huanian
Language]. – Bend inė kalba 88, [1–17].
Ja osla ienė Ju gi a, G igo je s Ju is, U bana ičienė Joli a, Ind ičāne Inese 2019: Bal ų kalbų
ga synas XXI a. p adžioje: balsių i ga sų są eikos ins umen inis y imas [Sounds o he
Bal ic Languages in he Ea ly 21s Cen u y: An Expe imen al S udy o Vowels and
Coa icula ion], Vilnius: Lie u ių kalbos ins i u as. DOI: h ps://doi.o g/10.35321/e-
pub.1.bal u-ga synas.
Johnson Kei h, Lade oged Pe e , Lindau Mona 1993: Indi idual di e ence in owel
p oduc ion. – The Jou nal o Acous ical Socie y o Ame ica 94(2), 701–714.
Kaukėnienė Lidija 2002: Lie u ių bend inės kalbos neki čiuo ų balsių spek ai [The Spec a
o Uns essed Vowels in S anda d Li huanian]. – Kalbo y a 51(1), 41–59.
Kazlauskienė As a 2018: Bend inės lie u ių kalbos one ikos i onologijos pag indai [Basics o
Gene al Phone ics and Phonology o he Li huanian Language], Kaunas: Vy au o
Didžiojo uni e si e as
Ken Raymond, Read Cha les 2002: The Acous ic analysis o speech, Albany: Thomson
Lea ning.
Kewley-Po Diane 1982: Measu emen o o man ansi ions in na u ally p oduced s op
consonan - owel syllables. – The Jou nal o he Acous ical Socie y o Ame ica 72(2),
379–389.
Kudi ka Robe as 2009: Lie u ių bend inės kalbos ki čiuo ų ilgųjų i umpųjų balsių
o man inė s uk ū a [Fo man S uc u e o he Accen ed Long and Sho Vowels in
Li huanian S anda d Language]. – Respec us Philologicus 16(21), 141–152.
Ledicho a Edi a 2020ɑ: Labializuo ų bend inės lie u ių kalbos balsių ekspe imen inis y imas
[The Expe imen al Resea ch o Quali a i e Cha ac e is ics o Labialised Vowels in
S anda d Li huanian]. − Bend inė kalba 93, [1−33].
Ledicho a Edi a 2020b: Lie u ių bend inės kalbos okalizmo a ian ai: ekspe imen inis kokybės
ypa ybių y imas: dak a o dise acija, Kaunas: Vy au o Didžiojo uni e si e as, Lie u ių
kalbos ins i u as.
Lee Yoonjeong, Kea ing Pa icia, K eiman Jody 2019: Acous ic oice a ia ion wi hin and
be ween speake s. – The Jou nal o Acous ical Socie y o Ame ica 146(3), 1568–1579.
Lindblom Bjö n 1990: Models o phone ic a ia ion and selec ion. – Pe ilus 11, 65–100.
23
Lindblom Bjö n, Moon Seung-Jae 1994: In e ac ion be ween du a ion, con ex , and speaking
s yle in English s essed owels. – The Jou nal o he Acous ical Socie y o Ame ica 96(1),
40–55.
Miliūnai ė Ri a 2018: Kalbos no mos i jų sa i eguliacija in e ne o bend uomenėje [Language
No ms and Thei Sel - egula ion in he In e ne Communi y], Vilnius: Lie u ių kalbos
ins i u as.
Pake ys An anas 1982: Lie u ių bend inės kalbos p ozodija [P osody o Li huanian S anda d
Language], Vilnius: Mokslas.
Pake ys An anas 2003 [1986]: Lie u ių bend inės kalbos one ika [Phone ics o he Li huanian
Common Language], 3-iasis pa ais. i papild. leidimas, Vilnius: Enciklopedija.
Picke James M. 1999: The Acous ics o Speech Communica ion. Fundamen als, Speech
Pe cep ion Theo y, and Technology, Needham Heigh s: Allyn & Bacon.
Rie eld Toni, an Hou Roland 2005: S a is ics in Language Resea ch: Analysis o Va iance,
Be lin, New Yo k: Mou on de G uy e .
Roesle Lindsay, Yung Song Jae 2018: Acous ic cha ac e is ics o ense and lax owels ac oss
sen ence posi ion in clea speech. – The Jou nal o he Acous ical Socie y o Ame ica
144(6), 535–540.
S e ens Kenne h N. 1998: Acous ic Phone ics, Camb idge, MA: The MIT P ess.
S e ens Kenne h N., House A hu 1963: Pe u ba ion o owel a icula ions by consonan al
con ex : An acous ical s udy. – Jou nal o Speech and Hea ing Resea ch 6(2), 111–128.
DOI: h ps://doi.o g/10.1044/jsh .0602.111.
Š age is E aldas 2024: Fo man inę balsių s uk ū ą lemiančių eiksnių yšiai: ž algomasis
bend inės lie u ių kalbos ilgųjų i umpųjų balsių y imas [In e ac ion Be ween Fac o s
De e mining he Fo man S uc u e o Long and Sho Vowels in S anda d Li huanian:
A Pilo S udy]. – Kalbo y a 77, 61–84. DOI:
h ps://doi.o g/10.15388/Kalbo y a.2024.77.3.
Š age is E aldas, U bana ičienė Joli a 2024: Lie u ių kalbos balsių o man inę s uk ū ą i
ukmę lemiančių eiksnių są eika bei hie a chija [Phone ic Dis ibu ion o Vowel Quali y
and Quan i y in S anda d Li huanian: A Fac o ial Analysis], Vilnius: Lie u ių kalbos
ins i u as. DOI: h ps://doi.o g/10.35321/e-pub.109.balsiu- eiksniai-hie a chija.
Ti ze Ingo R. 2000: P inciples o Voice p oduc ion, Iowa Ci y: Na ional Cen e o Voice and
Speech.
Tsukada Kimiko 2009: An acous ic compa ison o owel leng h con as s in A abic, Japanese
and Thai: Du a ional and spec al da a. – In e na ional Jou nal o Asian Language
P ocessing 19(4), 127–138.
U bana ičienė Joli a, Ind ičāne Inese, Ja osla ienė Ju gi a, G igo je s Ju is 2019: Bal ų kalbų
ga synas XXI a. p adžioje 2: P iebalsių ins umen inis y imas [Sounds o he Bal ic
24
Languages in he Ea ly 21s Cen u y 2: An Expe imen al S udy o Consonan s], Vilnius:
Lie u ių kalbos ins i u as. DOI: h ps://doi.o g/10.35321/e-pub.2.bal u-ga synas.
Vale ia V. Vai ke ičiū ė (1961): Vocal and bilingual phonemes o he Li huanian li e a y language
[Vowel and Diph hong Phonemes o he Li huanian Li e a y Language]. […]
Ques ions 4, 19 and 46 o he Li huanian Language Su ey.
E ec s o Consonan Voicing on Vowel Fo man Dis ibu ion in S anda d
Li huanian: A Fac o ial Analysis
Summa y
I is well known ha om he phone ic poin o iew, any speech sound is a ma e o g ea
dynamici y. The owel sys em in a language is no excep ion. In mo e gene al e ms, he spec al
cha ac e is ics o he owel depend on a g ea numbe o bo h linguis ic and ex alinguis ic
ac o s. This is he eason why a ac o ial analysis mus be conduc ed o gain a be e
unde s anding o he dynamics o speech. The i s a emp o analyse he dynamics o he
phone ic owel ealisa ion in S anda d Li huanian has yielded posi i e esul s. Al hough, o some
eason, a - eaching conclusions canno be d awn ye , i can ne e heless be obse ed ha he
dynamic plane o quan i a i e and quali a i e owel ea u es is no en i ely chao ic. I can be
seen ha all ela ed ac o s a e linked by hie a chical ela ionships.
This s udy aimed o de e mine wha posi ion in he hie a chy o ac o s (i.e. be ween consonan al
en i onmen , phonological leng h, p osodic ocus and speake ) di ec ly in luences he o ma i e
dis ibu ion in he spec um o Li huanian owels (so-called co ne owels [i], [u], [u], and [u] occupied
by he consonan oicing ac o ). A ac o ial analysis ( wo-way ANOVA) was pe o med o his
pu pose. The esul s e ealed qui e weak posi ions o he oicing ac o in he hie a chy and i s
low-le el au onomy. I s ela i e powe on he F1[…]F2 dis ibu ion co ela es wi h he backness o
he owel, as i shi s owa ds he back posi ion o he ongue, di e ences implied by he
consonan oicing ac o on he o man s can be no iced mo e eliably. Al hough ha ing no
signi ican impac , he consonan oicing plays some ole in he expansion o he spec al space o
o he ac o s o ac wi h g ea e powe ( o example, i c ea es condi ions o high-back owels o
be mo e on ed in a den al s op en i onmen ). The e may indeed be mic o-a icula o y di e ences
ela ed o he ac o in ques ion, bu qui e ob iously hei de ec ion is hinde ed by he excessi e
in luence o co-ac ing ac o s. Fu he analysis is o be conduc ed o ge his dynamic pic u e in
mo e de ail.