Chap e 7
Re-posi ioning he human ansla o as
he u u e expe o GenAI ansla ion
in he ansla ion class oom: The esul s
o a collabo a i e s udy
Senem Öne Bulu
Yıldız Technical Uni e si y, Tu key
Wi h he launch o Cha GPT in 2022, which allows he p oduc ion o ansla ion
h ough a p omp -based dialogue be ween human and gene a i e a i icial in elli-
gence (GenAI), ansla ion ac i i y has been eshaped in o a genuinely collabo a-
i e endea ou . The na u e and po en ial challenges o his eme ging o m o ans-
la ion p ac ice call o immedia e in es iga ion. To his end, his chap e examines
he esul s o a esea ch expe imen in which g adua e-le el s uden s collabo a ed
wi h a ansla o educa o o explo e he dynamics o p omp enginee ing as an
eme ging job ole in he ield o GenAI ansla ion. The esul s showed ha he
expe imen helped he s uden s diagnose no only he s eng hs and weaknesses
o GenAI bu also hei own, acknowledging he need o human in e en ion, edi -
ing and e alua ion, o u n aw GenAI ou pu in o a unc ioning ansla ion. The
s uden s also e-e alua ed hei human added alue and sel -concep as educa ed
and expe ienced human ansla o s, highligh ing he sense o empowe men and
esponsibili y hey el as human decision-make s in he p omp enginee ing p o-
cess while add essing a eas o sel -imp o emen such as de eloping skills o com-
munica e wi h GenAI, which a e dis inc om hose equi ed o communica e wi h
a human being. The esul s also indica ed ha he es ablished, holis ic skill se o
a ained p o essional human ansla o was ega ded a p e equisi e o e ec i e
p omp enginee ing. This needs o be conside ed by ansla o educa o s who now
ha e he ask o educa ing s uden s so ha hey can cau iously bu con iden ly
collabo a e wi h GenAI.
Senem Öne Bulu . 2026. Re-posi ioning he human ansla o as he u u e expe o
GenAI ansla ion in he ansla ion class oom: The esul s o a collabo a i e s udy.
In JC Pene , Joss Moo kens & Masa u Yamada (eds.), Teaching ansla ion in he age o
gene a i e AI: New pa adigm, new lea ning?, 125–145. Be lin: Language Science P ess.
DOI: 10.5281/zenodo.17641076
Senem Öne Bulu
1 In oduc ion
The impac o machine ansla ion (MT) on he ansla ion p o ession and ans-
la o educa ion has been p o ound in he las decade, as has he esea ch in o
his impac . The li e a u e has ocused p ominen ly on he ense ela ionship be-
ween he human ansla o and he machine/ echnology, which Picke ing (2008)
concep ualised as a “dance o agency”. Olohan (2011: 354) d ew a en ion o he
“decen ing” o he human agen in his con ex . O’B ien (2012: 119–120) u he
cha ac e ised ansla ion as “human-compu e in e ac ion”, de ining wha was
happening hen as a “shi in pa adigm” and poin ing ou he po en ial o “dia-
logue” be ween human ansla o s and echnology.
Ruokonen & Koskinen (2017: 321) u he elabo a ed on he human-machine
dynamic, wi h a ocus on he emo ional dimension inhe en in his ela ionship.
They ound ha he human agen s we e bo h “willing” and “ eluc an ” in e ms
o assigning agency o he machine and ha he ela ionship be ween he hu-
man and he machine is oo “complex” o be educed o “simpli ied man e sus
machine concep ions”. In a simila ein, Cadwell e al. (2018: 312) obse ed ha
human ansla o s held bo h posi i e and nega i e pe cep ions o MT. Posi i e
pe cep ions ela ed o educed e o and inc eased inspi a ion while he nega i e
we e linked o conce ns abou he nega i e in luence o MT on human c ea i i y.
The au ho s also a gued ha he ‘dance o agency’ be ween he human and he
machine was a symbio ic one, as hey mu ually eed o and shape each o he ;
he “dis inc ion be ween he human and he ma e ial agen ” is blu ed (Cadwell
e al. 2018: 303).
The p ac ical consequences o he pe asi e in eg a ion o MT echnologies
in o he global ansla ion indus y and hei economic impac ha e also been
explo ed. Moo kens (2017), o ins ance, no ed he nega i e consequences o he
eelance employmen model, which has been inc easingly adop ed in he global
ansla ion indus y, in e ms o p ices and isola ion. Acco ding o him, he o-
cus should be on how he human ansla o ac ually wo ks wi h MT. Simila ly,
Viei a (2018: 15) ound ha , a he han eeling an “in insic nega i i y o he
echnology”, he pe cei ed h ea om echnology among ansla o s appea s o
s em om business p ac ices in he indus y.
Ano he ocus in he ela ed li e a u e has been on he need o ede ine he
p o ile o human ansla o s. Sakamo o (2019: 68–69) unde lined he necessi y o
he “ ede ini ion o he concep o ‘ ansla o ’, and consequen ly ‘ ansla ion’ pe
se” on he g ounds ha “ he no ion o ‘ ansla ion’ is now being challenged by
he g ow h o echnologies”. The au ho also no ed ha a en ion should be paid
126
7 Re-posi ioning he human ansla o
o he unce ain ies ela ed o he legal equi emen s in MT use, p icing o pos -
edi ing, and pos -edi o s’ p o iles and skills (Sakamo o 2019: 63–66). Based on a
comp ehensi e analysis o he eme ging needs and equi emen s in he ansla-
ion sec o , which was conduc ed wi h he aim o add essing he “dis ance be-
ween eaching and p o essional p ac ice” in he con ex o MT compe ences,
Gaspa i e al. (2015: 333–334) epo ed “a s ong need o an imp o emen in
quali y assessmen me hods, ools and aining”, unde lining he “g owing up-
ake o MT and he pe cei ed inc ease o i s p e alence in u u e wo k lows”.
The e has been ex ensi e and subs an ial esea ch in o embedding in ans-
la o educa ion he new skills, abili ies and compe encies ha human ansla-
o s should possess in he age o MT, as well as in o he concomi an need o
e-p o ile ansla o educa ion. Fo ins ance, Dohe y & Kenny (2014: 296–297)
designed an SMT syllabus, which o eg ounds he “empowe men ” o he hu-
man ansla o by enabling him/he o in e ene in and add alue o he SMT
wo k lows when acing legal, e hical and echnical obs acles (Kenny & Dohe y
2014: 288). This “empowe ing” app oach has been u he ex ended by Mellinge
(2017: 284), who p oposes a holis ic, “c oss-module o c oss-cu icula in eg a-
ion” o pos -edi ing and MT “ o a oid compa men alisa ion o a ious compe-
ences and skillse s”. Mellinge also sugges s he inclusion o “con olled au ho -
ing, e minology managemen , engine uning, and pos -edi ing” in ansla ion
p ac ice cou ses, as hese a e “ ep esen a i e o many o he skills men ioned
in ecen wo ks on machine ansla ion pedagogy” (Mellinge 2017: 284). Pym
(2019) d ew a en ion o he iden i ica ion o “au oma ion- esis an skills” and
hei in eg a ion in o ansla o educa ion.
The measu es o be aken by ansla o educa o s o empowe s uden s and
help hem become “awa e o hei use ulness in o de o maximise hei agency as
ansla o s” has also been he ocus in a s udy by Moo kens (2018: 375–376), who
designed a ansla ion e alua ion exe cise o enable s uden s o “demys i y NMT
ou pu ”. Based on a econcep ualisa ion o ansla o compe ence in he age o
MT, Öne Bulu (2019: 3) designed a lea ning p ac ice o help s uden s “ aise hei
awa eness o hei p o essional sel -concep as human ansla o s” and sugges ed
he conside a ion o human ansla o compe ence and human ansla ion me a
compe ence. Ni zke e al. (2019: 248) p oposed a no el “pos -edi ing compe ence
model”, which included he co e compe ences o isk assessmen compe ence,
s a egic compe ence, consul ing compe ence and se ice compe ence. Öne &
Öne Bulu (2021: 100) in es iga ed ansla ion s uden s’ pe cei ed di icul ies
and bene i s in he con ex o “pos -edi ing o ien ed neu al machine ansla ion
e o anno a ion and quali y e alua ion”.
127
Senem Öne Bulu
Explo ing he “dynamics o he human-machine dance in he ansla ion class-
oom”, Öne Bulu & Alimen (2023: 377) designed a lea ning expe imen o enable
he s uden s o ca y ou eme gen MT- ela ed asks o pos -edi ing, p e-edi ing
and e o anno a ion. The au ho s epo ed ha he expe imen helped he ma-
jo i y o pa icipa ing s uden s “ aise awa eness o hei sel -concep as human
agen s and o he human added alue hey can c ea e while dancing wi h he
machine” (Öne Bulu & Alimen 2023: 387).
These examples o pas s udies ocusing on he impac o MT on he ansla ion
p o ession and ansla o educa ion show ha ansla ion s udies (TS) esea ch
has indeed come a long way o come o e ms wi h and adap o he pa adigm
shi (O’B ien 2012) induced by he inc easingly all-pe asi e echnologisa ion
o ansla ion, especially since he launch o NMT in 2016.
Ye , he yea 2022 p esen ed bo h ansla o s and ansla o educa o s wi h an-
o he , unp eceden ed challenge. Wi h he launch o Cha GPT in 2022 (OpenAI),
which allows o he p oduc ion o ansla ion h ough a p omp -based dialogue
be ween human and gene a i e a i ical in elligence (GenAI), he ea ly cha ac e -
isa ions o ansla ion as human-compu e in e ac ion (O’B ien 2012) and dance
o agency (Olohan 2011) ha e become eali ies in he ulles sense. This poses new
ques ions and opens new a eas o esea ch o he TS communi y, especially o
educa o s and esea che s o ansla o educa ion.
In hei abo e-ci ed esea ch, Öne Bulu & Alimen (2023: 389) no ed ha , in
he ace o he apid ad ancemen s in MT echnologies and he unce ain ies
conce ning hei in eg a ion in o educa ion, he e-posi ioning o ansla o ed-
uca o s is as necessa y as ha o human ansla o s. I has also been a gued ha
such e-posi ioning in ol es an “awa eness o he ac ha he ansla o educa-
o has he p ima y ask o diagnosing and p omo ing eme gen a eas o human
added alue c ea ion in he MT age”, and ha his can only be achie ed by e-
s uc u ing he ansla ion class oom as a “pla o m o collabo a i e lea ning,
whe e all in ol ed can wo k oge he o disco e wha emains and will emain
human in he MT age” (Öne Bulu & Alimen 2023: 389).
In he ace o he pa adigm shi induced by he in oduc ion o Cha GPT and
o he GenAI ools, ansla o educa o s should once again wo k in collabo a-
ion wi h s uden s o explo e he dynamics o in eg a ing hese ools in o hei
wo k low. Mo i a ed by his e y need, he p esen chap e p esen s he esul s
o a collabo a i e esea ch expe imen designed o explo e he dynamics o he
human-GenAI dialogue and ind he pa hways o be ollowed in inco po a ing
he insigh s gained in o ansla o educa ion.
The emainde o his chap e is s uc u ed as ollows. Fi s , he unp eceden ed
p ac ical and heo e ical challenges posed by GenAI echnology in e ms o he
128
7 Re-posi ioning he human ansla o
e y aison d’ê e o he human ansla ion a e explo ed. Second, in o ma ion
on he design and implemen a ion o he expe imen is gi en. Thi d, he da a
ob ained om he expe imen is analysed. Fou h, a discussion o he esul s is
p esen ed oge he wi h conclusions and sugges ions o u u e esea ch.
2 Human added alue and he human ansla o ’s
sel -concep in he age o GenAI: Re- isi ing he
in e p e i e ask and esponsibili y o he human
ansla o
Cha GPT is a GenAI ool powe ed by a la ge language model (LLM) and has
no been de eloped speci ically o in e lingual ansla ion. Howe e , he e -
olu ion i has b ough o ansla ion is o g ea signi icance as he ool akes
p omp s om human agen s and hence allows hem o p oduce a ansla ion by
p o iding he ool wi h con ex ual in o ma ion and guidance, i.e. by pe o ming
p omp enginee ing (Yamada 2023 and Chap e 5 o his book). This was ce ainly
no he case wi h MT, which makes GenAI echnology a eal b eak h ough ha
now equi es ansla o educa o s o ask and y o answe new ques ions abou
human- echnology in e ac ion.
In my opinion, in his endea ou we should no a emp o ein en he wheel.
Ins ead, we should e isi he ele an exis ing insigh s in o MT, especially hose
ha deal wi h he in eg a ion o MT in o ansla o educa ion, in a way ha
empowe s human ansla o s o c ea e human added alue, as ou lined in he
p e ious sec ion (see In oduc ion). This should also in ol e ying o answe
he challenging ques ion posed by GenAI: wha is human in ansla ion? This
is because, unlike MT, GenAI c ea es he illusion o ha ing an ac ual dialogue
wi h a non-human agen which p e ends o unde s and, in e p e and p oduce a
linguis ic message.
This new o m o challenge should u ge us o e isi he in e p e i e ask and
esponsibili y o he human ansla o . Acco dingly, he app oach p esen ed he e
is pe o ming a backwa d eading. Ra he han gi ing p io i y o he in es iga-
ion o he pe o mance, po en ial and/o limi a ions o GenAI, wha he p esen
s udy p oposes is he ins umen alisa ion o he human-GenAI dialogue in o de
o y o econside wha is/will emain human in ansla ion in a new, b igh e
ligh and ind ways o inco po a e i in o ansla o educa ion in he age o GenAI.
In a seminal con ibu ion by Massey & Ki aly (2019) i led “The Fu u e o T ans-
la o Educa ion: A Dialogue”, Ki aly s a es ha “ anscoding ( he mechanical e-
placemen o linguis ic uni s om a lis wi h co esponding uni s om a pa allel
129
Senem Öne Bulu
lis ) is no a he hea o ansla ion a all” (16, emphasis mine) and ha , al hough
in he u u e he ansla ion p o ession will be “di e en ”, i is “ce ainly no go-
ing o disappea – unless a some poin in ou e olu ion we no longe need o
in e p e ex s” (16, emphasis mine).
I is no coincidence ha Ki aly highligh s he ac o in e p e ing as he aison
d’ê e o he human ansla o , which echoes he he meneu ical app oach applied
o ansla ion by Schleie mache (1813/1977). Acco ding o He mans (2015: 101–
102), Schleie mache belie ed ha he sole op ion o a ansla o is “ o ac as
he he meneu icis does: o wo k o a ain he bes possible unde s anding o he
o eign ex which ne e heless emains o eign, and p esen o he eade […]
exac ly ha unde s anding”. Acco dingly, he meaning o a ex is no encoded in
a ex o be decoded by he ansla o , a he i should be in e p e ed by he ans-
la o o a ain he bes possible unde s anding, which cons i u es he he meneu-
ical ask o he ansla o . The ansla o is hen equi ed o p esen his unde -
s anding o he eade , which, acco ding o He man’s eading o Schleie mache ,
“exceeds he meneu ics” (He mans 2015: 99). In a simila ein, Şe ban & La isa
(2016: 295) unde line ha in Schleie mache ’s app oach, he mee ing be ween
he au ho and he eade is “media ed and o ches a ed by he ansla o ” and
he ansla o assumes “ he mos ac i e media ing ole”.
Al hough he he meneu ical app oach is o c i ical impo ance in ha i allows
he ansla o oom o ac i e media ion, a speci ic concep ualisa ion by A ojo
(1997: 18) o he “ine i able in e e ence” o he ansla o in he con ex o ans-
la ion e hics could u he he a emp o econside wha is human in ansla ion.
A ojo iews ansla ion e hics in e ms o he ela ionship be ween he “au ho-
ial powe ” o he ansla o and his/he “e hical esponsibili y” (A ojo 1997: 18).
Based on a pos mode n iew o “language and subjec ”, he au ho unde lines
ha “[as] no eading can e e aspi e o epea o p o ec someone else’s ex ,
ansla o s necessa ily ha e he igh o exe cise hei au ho ial powe ‘as long
as hei game is played up on ’” (Be man 1995: 93, ci ed in Simon 1996: 36). In
he iew o he au ho , “[s]uch ‘ igh ’ implies, howe e , an e hical esponsibili y
which pa allels ha o ‘o iginal’ au ho s” (A ojo 1997: 18).
In he wake o he eme ging pa adigm b ough abou by GenAI, his line o
hough can be ollowed by esea che s and ansla o educa o s o o eg ound
he human ansla o ’s in e p e i e ask and he accompanying esponsibili y
wi h he aim o explo ing human added alue (Massey 2021) and human ans-
la o s’ sel -concep (Ki aly 1990, 2000b, Ha o-Sole & Ki aly 2019 and Massey &
Ki aly 2019) in he age o GenAI.
D awing ca e ul a en ion o he human aspec o added alue, Massey (2021:
39) cha ac e ises human added alue as being “mani es in ansla ional decision-
130
7 Re-posi ioning he human ansla o
making and p oblem-sol ing on a concep ual le el ha anscends he su ace
lexical ealisa ions by which meaning is con eyed in sou ce and a ge ex s”.
He also obse es ha “ he added alue ha human ansla o s can and do b ing
o bea esides in hei socio-cul u al, socio- echnical 4EA (Embodied, Embed-
ded, Enac i e, Ex ended and A ec i e) cogni ion, as opposed o disembodied, de-
con ex ualised a i icial in elligence” (Massey 2021: 39). In doing so, he au ho
e isi s Venu i’s (2019) ad ocacy o he “he meneu ical model o unde s anding
ansla ion no as ‘ he ep oduc ion o ans e o an in a ian ha is con ained
in o caused by he sou ce ex ’ bu as ‘an in e p e i e ac ha ine i ably a ies
sou ce- ex o m, meaning, and e ec ’” and Pym’s (2003) “minimalis de ini ion
o ansla ion compe ence as ‘ he abili y o gene a e a se ies o mo e han one
iable a ge ex (TTI, TT2 ... TTn) o a pe inen sou ce ex (ST); he abili y o
selec only one iable TT om his se ies, quickly and wi h jus i ied con idence’”
(Massey 2021: 51).
The delinea ion by Massey (2021: 39) o he added alue o he human ansla-
o is-à- is AI s ongly cap u es he concep ualisa ion o he human ansla o
as an agen who has he he meneu ical ask and powe o in e p e ing he ex
and who bea s he esponsibili y o hei in e p e i e ac . This is some hing ha
emains qui e impossible o he “disembodied, de-con ex ualised a i icial in el-
ligence”. In he age o GenAI, we mus he e o e pay unp eceden ed a en ion
o he ansla o ’s “sel -concep ” (Ha o-Sole & Ki aly 2019), which includes “ he
image o he ansla o ’s social ole”, “ he ansla o ’s app aisal o his o he com-
pe ency o ansla ing a pa icula ex ” and “an unde s anding o esponsibili y
owa ds he o he ac o s in he ansla ion con ex o si ua ion” (261, emphasis
mine). Consequen ly, we belie e ha he cons uc s o human added alue and
o (human) ansla o ’s sel -concep should cons i u e he p i ileged oci in ex-
plo ing he human-GenAI dialogue and inco po a ing he gained insigh s in o
ansla o educa ion in a way ha helps s uden s aise hei sel -concep s (Ha o-
Sole & Ki aly 2019: 261–262) as human ansla o s whose aison d’ê e is o pe -
o m he he meneu ical ask o in e p e ing and bea ing esponsibili y. I is wi h
his ocus in mind ha he ollowing collabo a i e esea ch expe imen was con-
duc ed.
3 Design and implemen a ion o he collabo a i e
esea ch expe imen
The esea ch expe imen , he esul s o which a e epo ed in he p esen s udy,
was designed ollowing he ene s o he collabo a i e esea ch model, imple-
131
Senem Öne Bulu
men ed by Ha o-Sole & Ki aly (2019). I ollows he social cons uc i is and
eme gen is app oach o ansla o educa ion (Ki aly 2000a: 256), whe eby he
educa o - esea che wo ks collabo a i ely wi h ansla ion s uden s “ o in es-
iga e opics in he domain o T ansla ion Psychology wi h he goal o ha ing
eache - esea che s lea n abou he ansla o ’s psychological ‘sel ’ igh along
wi h hei s uden s”. Acco dingly, he esea ch expe imen p esen ed he e was
collabo a i ely pe o med by g adua e le el ansla ion s uden s, who pa ici-
pa ed in he esea ch on a olun a y basis, and a ansla o educa o (also he
au ho o he p esen chap e ). The aim was o explo e he dynamics o human-
GenAI in e ac ion as human ansla o s wo k wi h GenAI ools as p omp en-
ginee s. The s uden - esea che g oup consis ed o i e s uden s en olled in he
PhD p og amme in ansla ion s udies a Yıldız Technical Uni e si y, Tü kiye.
All had p e ious ansla ion educa ion and expe ience and nea -na i e English
p o iciency. The educa o - esea che was an associa e p o esso o ansla ion
s udies a he same uni e si y.
In he esea ch p ocess, he s uden - esea che s and he educa o - esea che
collec i ely decided on he design o he ask o p omp enginee ing in h ee
sessions held online in June 2024. A e he design was ag eed upon, he s uden -
esea che s pe o med he ask indi idually and epo ed on hei lea ning/
esea ch p ocesses h ough sel - e lexi e accoun s o hei expe iences.
The ini ial decisions made in he eco ded online discussions conce ned he
selec ion o he sou ce ex s and o he GenAI ool o be used o ansla e he
ex s h ough c a ing and, when necessa y, cu a ing p omp s (i.e. p omp engi-
nee ing; Yamada 2023) so as o p e-p ocess, ini ia e, p oduce, edi and e alua e
ansla ions ha achie ed he in ended quali y le el and unc ions. I was collec-
i ely decided o use he ee e sion o OpenAI’s Cha GPT. As o he sou ce
ex s, wo English ex s, an essay and a book desc ip ion, we e selec ed on he
g ounds ha bo h ex s we e o mixed o ms be ween exp essi e and ope a i e
ex ypes, while also displaying he ea u es o an in o ma i e ex ype (Reiss
1981: 124–125). Fu he , he ansla ion o bo h ex s demanded me iculous a en-
ion o cul u al nuances and con ex , idioma ic exp essions, one, in en , and lin-
guis ic s yle. The i s sou ce ex was an essay by Anna Quindlen, published in
he “Li e in he 30s” column in The New Yo k Times (Quindlen 1987). The second
sou ce ex was he book desc ip ion o The Time Regula ion Ins i u e, he English
ansla ion o Ahme Hamdi Tanpına ’s Tu kish no el i led Saa le i Aya lama
Ens i üsü, ansla ed by Alexande Dawe and Mau een F eely and published in
2014 (Tanpina 2014).
Nex , decisions had o be made conce ning he guidelines, namely he wo
ansla ion b ie s and he gene al p inciples o p omp enginee ing. Fi s , he
132
7 Re-posi ioning he human ansla o
educa o - esea che p o ided he s uden - esea che s wi h he exis ing (albei
limi ed, due o he ac ha i is a ela i ely new opic) li e a u e on p omp en-
ginee ing. This led o an exchange o opinions in he i s wo online sessions.
D awing on hose discussions, he educa o - esea che hen d a ed he guide-
lines o he ask and sen hem o he s uden - esea che s o e alua ion. In he
hi d session, he d a guidelines we e discussed and inalised collec i ely.
The educa o - esea che and he s uden - esea che s hen had o decide
whe he o use he inalised ansla ion b ie s as he i s p omp in he dialogue
wi h he GenAI ool. Taking in o conside a ion he esul s o he expe imen s
al eady conduc ed in o he s udies on p omp enginee ing (e.g. He 2024, Peng
e al. 2023, Yamada 2023), i was decided no o include he ansla ion b ie s in
p omp enginee ing and ins ead, o use hem as guidance o human ansla o s-
p omp enginee s only. In his way, he human ansla o s we e en us ed wi h
he esponsibili y o he ansla ion ac despi e pe o ming his ac h ough
p omp enginee ing. The ansla ion b ie s o he i s and second sou ce ex s
we e as ollows:
(1) T ansla ion B ie 1:
“The Name is Mine” is an essay by Anna Quindlen published in he “Li e
in he 30s” column in The New Yo k Times on Ma ch 4, 1987. An online
magazine websi e edi o in Tu key needs a Tu kish ansla ion o he
essay and asks you o ansla e i in o Tu kish o publish i on he
websi e. The in ended audience o he ansla ion is p ima ily adul
eade s who would be in e es ed in he wo ks o Anna Quindlen, who
w i es p ima ily on eminism and amily li e and is a Puli ze
P ize-winning columnis , jou nalis and au ho . The edi o asks you o
ansla e he second and nin h pa ag aphs o he essay as a es .
(2) T ansla ion B ie 2:
The sou ce ex is he book desc ip ion o he no el The Time Regula ion
Ins i u e, which is he English ansla ion o he p ominen Tu kish
no elis and poe Ahme Hamdi Tanpına ’s Tu kish no el i led Saa le i
Aya lama Ens i üsü (1961), ansla ed by Alexande Dawe and Mau een
F eely and published by Penguin Classics in 2014. A Tu kish li e a y
schola is asking you o ansla e he English book desc ip ion in o
Tu kish in o de o examine how he English ansla ion o he book (The
Time Regula ion Ins i u e) is ma ke ed o he English-speaking audience
by Penguin Classics.
133
Senem Öne Bulu
5 Discussion and conclusion
On he basis o he abo e analysis o he pa icipan s’ answe s o he ques ions,
i can be claimed ha he expe imen p o ided he s uden - esea che s wi h he
oppo uni y o wo k wi h GenAI o p oduce a unc ioning a ge ex acco ding
o a gi en ansla ion b ie and, in doing so, o e alua e ce ain aspec s o he
human-GenAI dance h ough an expe imen -based diagnosis o he s eng hs
and weaknesses o GenAI.
In he accoun s o he pa icipan s, he main s eng h associa ed wi h GenAI
was, a he expec edly, speed. Apa om his, pa icipan s placed special em-
phasis on he ex analysis abili ies o GenAI, which we e conside ed o be a
signi ican ac o ha can acili a e and imp o e he ansla ion p ocess. No able
weaknesses we e GenAI’s inabili y o p oduce a s ylis ically adequa e and luen
ansla ion, i s need o con inuous p omp ing om a human ansla o (i.e., i s
ailu e o implemen ex -speci ic ansla ion s a egies wi hou guidance om
a human ansla o ), i s po en ial o gene a e biased, cul u ally insensi i e and
ac ually inaccu a e ou pu which can only be ixed by he ca e ul moni o ing
and e alua ion o a human ansla o . Las bu no leas , GenAI’s lack o ue-li e
expe ience was also men ioned.
Fu he , pa icipan s’ accoun s also highligh ha , du ing he cou se o he ex-
pe imen , he s uden - esea che s e-e alua ed hei human added alue and sel -
concep as ained and expe ienced ansla o s. Despi e hei amilia i y wi h,
and expe ience o , he dynamics o human-machine in e ac ion du ing ansla-
ion, hey ecognised he necessi y o engaging in a new o m o collabo a ion
wi h GenAI. Indeed, he pa icipan s’ accoun s o hei own s eng hs and weak-
nesses we e closely ela ed o he s eng hs and weaknesses hey had iden i ied
o LLMs. The need o a p o essional human ansla o o in e ene (edi and
e alua e) he aw ou pu o GenAI ools o u n i in o a unc ioning ansla-
ion was emphasised by all pa icipan s. This echoes he indings o ecen s ud-
ies on he necessi y o a human ansla o ’s in e en ion due o he limi a ions
o GenAI (e.g. Alimen 2023, Öne Bulu & Alimen 2023). Thus, he p ocess by
which he s uden s adjus ed he LLM ou pu h ough p omp enginee ing mi -
o s he he meneu ic ole o he human ansla o , who in e p e s he ‘meaning’
o a ex and e ames i app op ia ely o he a ge audience. This unde sco es
ha GenAI-assis ed ansla ion is no me ely mechanical bu equi es deep cul-
u al adjus men s and human judgmen .
Pa icipan s also highligh ed he sense o empowe men and au ho i y hey
el as decision-make s, who mus make decisions conce ning ansla ion s a e-
gies and p omp he GenAI ool acco dingly. This sense o esponsibili y b ough
140
7 Re-posi ioning he human ansla o
hem powe as he ones esponsible o he inal ansla ion ou pu . This aligns
wi h he e hical and he meneu ic oles o he human ansla o . Howe e , some
o he pa icipan s also add essed a eas o sel -imp o emen such as de eloping
skills o communica e wi h GenAI, which is dis inc om communica ing wi h a
human being, and designing and cu a ing e ec i e p omp s o guide AI h ough
he ansla ion p ocess wi hou becoming impa ien and s essed. This gi es
hin s abou he possible emo ional challenges human ansla o s can ace while
wo king wi h GenAI, e idencing he need o “in oducing an emo ional in el-
ligence dimension in o ansla o aining” (Pene & Fe nandez-Pa a 2023:349).
This is some hing I belie e should be add essed in u u e esea ch on he in eg a-
ion o GenAI in o ansla o educa ion.
The analysis o he esul s also p o ided insigh s in o o he ac o s ha need
o be conside ed. All s uden - esea che s acknowledged he signi icance o hei
p e ious educa ion and expe ience in pe o ming he eme ging ask o p omp
enginee ing. They also ag eed ha he holis ic se o skills o a ained p o es-
sional human ansla o is a p e equisi e o e ec i e p omp enginee ing. This
had al eady been sugges ed in a ecen s udy on he ole o human compe en-
cies in he “ ech-d i en language se ices indus y” (Öne & Bengi 2024:92). Ye
mos o he pa icipan s no ed ha , in addi ion o his p e equisi e, ansla ion
s uden s should lea n o communica e e ic i ely wi h GenAI (i.e., p omp i ), o
c i ically e alua e and edi i s ou pu and o de elop an awa eness o he ac
ha human ansla o s a e s ill he ul ima e au ho i y wi h he igh o he inal
say by i ue o he us and esponsibili y in es ed in hem. This pa icula pe -
spec i e unde sco es he impo ance o acknowledging ansla o s’ ac i e ole
in econs uc ing cul u al alue and meaning wi hin he in e ac i e ansla ion
p ocesses acili a ed by GenAI. Fu he , his also poin s o he signi ican po en-
ial o he meneu ics-based educa ion in os e ing a deepe unde s anding o he
human ansla o ’s ole in GenAI-assis ed ansla ion.
This, in u n, places g ea e esponsibili y on ansla ion educa o s, who, ac-
co ding o he pa icipan s, need o become dedica ed lea ne s and use s o s a e-
o - he-a GenAI echnology so ha hei ansla ion class ooms become a sa e
space whe e ansla ion s uden s lea n o communica e and dance mo e con i-
den ly wi h GenAI.
These expe ience-based insigh s, which we e gained h ough a collabo a i e
esea ch expe imen , shed ligh on he eme ging oppo uni ies, as well as chal-
lenges, o eposi ioning he human ansla o as he u u e expe in GenAI
ansla ion in he class oom. They also show a possible pa h o he ul ilmen
o he ansla o educa o - esea che ’s – also eme ging – ask o e-e alua ing
141
Senem Öne Bulu
wha is/will emain human in ansla ion and inco po a ing i in o educa ion in
he age o GenAI.
Re e ences
Alimen, Nilü e . 2023. Makine çe i isinden sohbe obo u çe i isine: Cha GPT
ile deneysel bi çalışma [F om machine ansla ion o cha bo ansla ion: An
expe imen al s udy wi h Cha GPT]. RumeliDE Dil e Edebiya A aş ı mala ı
De gisi (36). 1532–1548.
A ojo, Rosema y. 1997. Asymme ical ela ions o powe and he e hics o ans-
la ion. Tex Con ex 11(1). 5–24.
Be man, An oine. 1995. Pou une c i ique des aduc ions: John Donne. Pa is,
F ance: Gallima d.
Cadwell, Pa ick, Sha on O’B ien & Ca los S. C. Teixei a. 2018. Resis ance and
accommoda ion: Fac o s o he (non-) adop ion o machine ansla ion among
p o essional ansla o s. Pe spec i es 26(3). 301–321. DOI: 10.1080/0907676X.
2017.1337210.
Dohe y, S ephen & Do o hy Kenny. 2014. The design and e alua ion o a s a is-
ical machine ansla ion syllabus o ansla ion s uden s. The In e p e e and
T ansla o T aine 8(2). 295–315. DOI: 10.1080/1750399X.2014.937571.
Gaspa i, Fede ico, Hala Almaghou & S ephen Dohe y. 2015. A su ey o ma-
chine ansla ion compe ences: Insigh s o ansla ion echnology educa o s
and p ac i ione s. Pe spec i es 23(3). 333–358. DOI: 10.1080/0907676X.2014.
979842.
Ha o-Sole , Ma ia del Ma & Don Ki aly. 2019. Explo ing sel -e icacy belie s in
symbio ic collabo a ion wi h s uden s: An ac ion esea ch p ojec . The In e -
p e e and T ansla o T aine 13(3). 255–270. DOI: 10 . 1080 / 1750399X . 2019 .
1656405.
He, Sui. 2024. P omp ing Cha GPT o ansla ion: a compa a i e analysis o ans-
la ion b ie and pe sona p omp s. h ps://a xi .o g/abs/2403.00127.
He mans, Theo. 2015. Schleie mache and Pla o, he meneu ics and ansla ion.
In La isa Ce cel & Ad iana Se ban (eds.), F ied ich Schleie mache and he
ques ion o ansla ion, 77–106. Be lin, Ge many: De G uy e . DOI: 10.1515/
9783110375916-007.
Kenny, Do o hy & S ephen Dohe y. 2014. S a is ical machine ansla ion in he
ansla ion cu iculum: O e coming obs acles and empowe ing ansla o s.
The In e p e e and T ansla o T aine 8(2). 276–294. DOI: 10.1080/1750399X.
2014.936112.
142
7 Re-posi ioning he human ansla o
Ki aly, Don. 1990. Towa d a sys ema ic app oach o ansla ion skills ins uc ion.
Uni e si y o Illinois. (Doc o al disse a ion).
Ki aly, Don. 2000a. A social cons uc i is app oach o ansla o educa ion: Em-
powe men om heo y o p ac ice. London, UK: Rou ledge.
Ki aly, Don. 2000b. A social cons uc i is app oach o ansla o educa ion: Em-
powe men om heo y o p ac ice. S . Je ome.
Massey, Ga y. 2021. T ansla ion li e acy in addi ional language lea ning: Closing
he concep ual di ide be ween ansla ion and language educa ion. Bulle in
suisse de linguis ique appliquée 114. 31–45. DOI: 10.5169/seals-1030136.
Massey, Ga y & Don Ki aly. 2019. The u u e o ansla o educa ion: A dialogue.
Cul us: The Jou nal o In e cul u al Media ion and Communica ion (12). 15–34.
Mellinge , Ch is ophe D. 2017. T ansla o s and machine ansla ion: Knowledge
and skills gaps in ansla o pedagogy. The In e p e e and T ansla o T aine
11(4). 280–293. DOI: 10.1080/1750399X.2017.1359760.
Moo kens, Joss. 2017. Unde p essu e: T ansla ion in imes o aus e i y. Pe spec-
i es 13(1). 464–477.
Moo kens, Joss. 2018. Wha o expec om neu al machine ansla ion: A p ac-
ical in-class ansla ion e alua ion exe cise. The In e p e e and T ansla o
T aine (12). 375–387.
Ni zke, Jean, Sil ia Hansen-Schi a & Ca men Can o a. 2019. Risk managemen
and pos -edi ing compe ence. Jou nal o Specialised T ansla ion (31). 239–259.
O’B ien, Sha on. 2012. T ansla ion as human-compu e in e ac ion. T ansla ion
Spaces (1). 101–122.
Olohan, Mae e. 2011. T ansla o s and ansla ion echnology: The dance o
agency. T ansla ion S udies 4(3). 342–357. DOI: 10.1080/14781700.2011.589656.
Öne , Işın & Zeh a Begüm Bengi. 2024. Essen ial o obsole e? The ole o hu-
man compe encies in he ech-d i en language se ices indus y. ansLogos
T ansla ion S udies Jou nal 7(1).
Öne , Işın & Senem Öne Bulu . 2021. Pos -edi ing o ien ed human quali y e alua-
ion o neu al machine ansla ion in ansla o aining: A s udy on pe cei ed
di icul ies and bene i s. ansLogos T ansla ion S udies Jou nal 4(1). 100–124.
Öne Bulu , Senem. 2019. In eg a ing machine ansla ion in o ansla o ain-
ing: Towa ds ‘human ansla o compe ence’? ansLogos T ansla ion S udies
Jou nal 2(1). 1–26.
Öne Bulu , Senem & Nilü e Alimen. 2023. T ansla o educa ion as a collabo a-
i e ques o insigh s in o he e-posi ioning o he human ansla o (educa-
o ) in he age o machine ansla ion: The esul s o a lea ning expe imen .
The In e p e e and T ansla o T aine 17(3). 375–392.
143
Senem Öne Bulu
OpenAI. N.d. Tex gene a ion models. h ps://pla o m.openai.com/docs/guides/
ex -gene a ion.
Pene , JC & Ma ia Fe nandez-Pa a. 2023. Dealing wi h s uden s’ emo ions: Ex-
plo ing ai EI heo y in ansla o educa ion. The In e p e e and T ansla o
T aine 17(3). 332–352. DOI: 10.1080/1750399X.2023.2237327.
Peng, Keqin, Liang Ding, Qihuang Zhong, Li Shen, Xuebo Liu, Min Zhang,
Yuanxin Ouyang & Dacheng Tao. 2023. Towa ds making he mos o Cha GPT
o machine ansla ion. h ps://a xi .o g/abs/2303.13780.
Picke ing, And ew. 2008. New on ologies. In And ew Picke ing & Kei h Guzik
(eds.), The mangle in p ac ice: Science, socie y, and becoming, 1–16. Du ham,
USA: Duke Uni e si y P ess.
Pym, An hony. 2003. Rede ining ansla ion compe ence in an elec onic age. In
de ence o a minimalis app oach. Me a 48(4). 481–497. DOI: 10.7202/008533a .
Pym, An hony. 2019. How au oma ion h ough neu al machine ansla ion migh
change he skill se s o ansla o s: D a w i en as pa o he p ojec Lan-
guage Compe ence and Wo k. h ps://www.academia.edu/40200406/How_
au oma ion_ h ough_neu al_machine_ ansla ion_migh _change_ he_skill_
se s_o _ ansla o s.
Quindlen, Anna. 1987. Li e in he 30s. The New Yo k Times.
Reiss, Ka ha ina. 1981. Type, kind and indi iduali y o ex : Decision making in
ansla ion. Poe ics Today 2. 121–131. DOI: 10.2307/1772491.
Ruokonen, Minna & Kaisa Koskinen. 2017. Dancing wi h echnology: T ansla o s’
na a i es on he dance o human and machinic agency in ansla ion wo k.
The T ansla o 23(3). 310–323. DOI: 10.1080/13556509.2017.1301846.
Sakamo o, Akiko. 2019. Why do many ansla o s esis pos -edi ing? A socio-
logical analysis using Bou dieu’s concep s. Jou nal o Specialised T ansla ion
31. 201–216. h ps://jos ans.soap2.ch/issue31/a _sakamo o.php.
Schleie mache , F ied ich. 1813/1977. On he di e en me hods o ansla ing.
In And é Le e e e (ed.), T ansla ing li e a u e: The Ge man adi ion om
Lu he o Rosenzweig, 66–89. Assen, Ne he lands: Van Go cum. DOI: 10.1163/
9789401200295_001.
Şe ban, Ad iana & Ce cel La isa. 2016. C ea i i y and al e i y in ilm ansla ion:
A e u n o Schleie mache ’s he meneu ics. Se uya Te esa & José Mi anda Jus o
(eds.). Cham: Sp inge . 291–303.
Simon, She y. 1996. Gende in ansla ion: Cul u al iden i y and he poli ics o
ansmission. London, UK: Rou ledge.
Tanpina , Ahme Hamdi. 2014. The ime egula ion ins i u e. London, UK: Pen-
guin.
144
7 Re-posi ioning he human ansla o
Venu i, Law ence. 2019. Con a ins umen alism: A ansla ion polemic. Lincoln,
USA: Uni e si y o Neb aska P ess.
Viei a, Lucas Nunes. 2018. Au oma ion anxie y and ansla o s. T ansla ion S ud-
ies 13(1). 1–21.
Yamada, Masa u. 2023. Op imizing machine ansla ion h ough p omp engi-
nee ing: An in es iga ion in o Cha GPT’s cus omizabili y. In Masa u Yamada
& Félix do Ca mo (eds.), P oceedings o Machine T ansla ion Summi XIX, Vol.
2: Use s T ack, 195–204. Macau, China: Asia-Paci ic Associa ion o Machine
T ansla ion. h ps://aclan hology.o g/2023.m summi -use s.19.
145
