Transcript 14-16 November 2013

A parallel corpus is a useful tool for
linguistic research: a case study of
English MUST and its Lithuanian
correspondences
Audronė Šolienė
Vilnius University
14-16 November 2013
OUTLINE
• Linguistic issue under analysis:
– epistemic-evidential overlap (Palmer 2001, Pietrandrea
2005, Cornillie 2007, Mortelmans 2001, 2009)
• ParaCorp:
– a parallel corpus can serve as a means to disclose the
epi-evid overlap
• Results:
– Epi-evid use of
correspondences
MUST
and
its
LT
translational
• Discussion and some conclusive remarks
14-16 November 2013 Innovative Information Technologies for Science, Business and Education IIT-2013
EPISTEMIC MODALITY
possibility
MODALITY
•propositional
modality used by
non-ep
speakers
to
express
their
judgments and attitudes towards
the factual status of the
proposition (Palmer 2001)
epistemic  propositional
necessity
non-ep
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
EVIDENTIALITY
visual/auditory/other sensory
EVIDENTIALITY
•the category the speaker uses to
indicate
what
sources
of
direct evidence
information/kind
of evidence s/he has
to assert the truth of a proposition
(Faller
2002, Pietrandrea
2005,
indirect
evidence
Cornillie 2007)
reported
inferentialobservational & logical
(Willet 1988: 65)
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
Epistemic-evidential overlap
• Inferential evidentiality = epistemic necessity
Should we bind
his wounds? He
must be in pain.
Necessity
…
Deontic
necessity
Epistemic necessity
=
Inferential evidentiality
Quotative
evidentiality
…
Evidentiality
(van der Auwera & Plungian (1998: 86)
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
MUST: two distinct
prototypical cores
• Epistemic-inferential necessity
It must be raining.
felicitous when the speaker has inferred that it is
raining, not when he has directly observed it
(Portner 2009:171)
• Non-epistemic
(participant-external deontic) necessity
Please... you must leave.
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
FOCUS (1)
• epistemic-inferential use
of MUST and its LT
correspondences
• in LT transl (EN orig  LT transl)
Hagrid must have forgotten to tell him something you had to do…
Tikriausiai Hagridas jį pamiršo pamokyti, ką reikia dar padaryti…
• in LT orig (EN transl  LT orig)
…velnias, galva kaip medinė, atmerkęs akis pamato skersai stalo
pamestas kelnes, matyt, jau labai gražus parėjo…
hell, his head feels like wood, when he opens his eyes he sees his
pants thrown across the table, he must have come home a pretty
sight…
14-16
. November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
FOCUS (2): Aux/adv Strategy
EN vs. Slavonic
• Epistemic possibility in a Slavonic parallel corpus - a pilot
study (van der Auwera, Schalley, Nuyts 2005)
•  Translations of a Harry Potter book
• 10: Bulgarian, Croatian, Czech, Macedonian, Polish, Russian,
Serbian, Slovak, Slovene, Ukrainian
• Potential vs. use: difference in frq of aux/adv
strategy use
• Areality issue: the phenomenon of the eastward
decreasing frequencies of mod auxiliaries
• Difference in degree of grammaticalization
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
AIM
• Qualitative & quantitative parameters of
translational correspondence
• Analysis of choice of adv/verb strategies
in translation
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
METHOD & DATA
 corpus-based contrastive analysis
 parallel corpus ParaCorp LTE
•
•
•
•
bidirectional (E>LT & LT>E)
fiction (written)
time span: 1980 – 2006
No of translators:
ENLT: 15 (5 men; 10 women)
LTEN: 29 (10 men; 19 women)
The sub-corpora compiled:
1. The E-LT ParaCorp (873,511 words)
2. The LT-E ParaCorp (698,987 words)
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
ParaCorp LTE
• Sentence alignment
– Aligner LYGIA (Kaunas Vytautas Magnus University)
<senalign>
<lang value="lt"><s>Prabudusi ji visada pirmiausia pamatydavo suplyšusias
užuolaidas.</s></lang>
<lang value="en"><s>The first thing Vacė ever saw when she awoke in the
morning was the torn curtains.</s></lang>
</senalign>
<senalign>
<lang value="lt"><s>Jau sudūlėjusios nuo senumo, jos plyšdavo, vos
prisilietus.</s></lang>
<lang value="en"><s>Yellowed with age, they'd tear from the slightest
touch.</s></lang>
</senalign>
• ParaConc tool (Barlow 1995)
– Excel spreadsheet
– Multidirectional search  but not automatic
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
Ep/Non-ep MUST in Eorig
•n= 265
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
Correspondences of Ep must in
LTtransl : adv/verb strategy
tikriausiai ‘most probably’,
matyt ‘seemingly’, turbūt
‘probably’  42 %
ENorig: This must all sound so silly to you. (aux)
LTtransl: Tau turbūt tai atrodo baisiai kvaila. (adv)
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
Correspondences of Ep must in
LTtransl
ParaCorpELT
%
words/
tikriausiai ‘most probably’
particles
matyt ‘seemingly/evidently’
(ADV)
turbūt ‘probably’, greičiausiai,
veikiausiai ‘most probably’etc.
verbs
turėti ‘must/have to’
22
Ø LT
25
14-16 November 2013
18
11
22
Innovative Information Technologies for Science, Business and Education
IIT-2013
Illustration of correspondences
E - original

It must be easier to live
beautiful lives when you're
posh.
TIKRIAUSIAI
Tikriausiai aristokratams
‘most probably’ lengviau gyventi gražų
gyvenimą.
A sharp cry of pain was wrung
out of her. She must have
fallen right on the injured arm.
MATYT
‘seemingly’
That must have been horrible
for you.
TURĖTI
‘must/have to’
14-16 November 2013
LT - translation
Pasigirdo skardus,
skausmo kupinas
riktelėjimas. Matyt,
nugriuvo tiesiai ant
sužeistos rankos.
– Tau turėjo būti baisu.
Innovative Information Technologies for Science, Business and Education
IIT-2013
Correspondences of Ep must in
LTorig : adv/verb strategy
35%
9%
Verbs %
Adv/Word %
Ø%
56%
LTorig: Jie visi su tarbom, maišais, krepšiais. Matyt, grįžta iš
kaimo, pasirinkę maisto. (adv)
Etrans: They're all carrying baskets, bags, and sacks. They
must be returning from the village where they got food. (aux)
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
Correspondences of Ep must in
LTorig
ParaCorpLTE
words/ turbūt ‘probably’
particles
matyt ‘seemingly’
(ADV)
ko gero ‘probably’, greičiausiai
‘most probably’, gal ‘perhaps’ etc.
verbs
turėti ‘must/have to’
Ø LT
14-16 November 2013
%
17
12
27
9
35
Innovative Information Technologies for Science, Business and Education
IIT-2013
Ø correspondences of MUST
Ø in LT_transl
Ø in E_transl
(omitted)
(inserted)
25%
35%
SL
Hagrid, he said quietly, “I Šį vakarą, vairuodamas mašiną
think you must have
Aukštaitijos pusėn, jis pirmą kartą
made a mistake”.
pagalvojo, kad nusipirko namą
kartu su Uršulėlės mirtimi.
TL
- Hagridai, - ramiai tarė
jis, - manau, jog
suklydote.
14-16 November 2013
That night, driving the car in the
direction of Aukstaitija, for the first
time Vaitkus thought that he must
have bought the house together
with Ursula's death.
Innovative Information Technologies for Science, Business and Education
IIT-2013
Ø correspondences
Ø
LT_transl
MAY
22 %
MIGHT
20 %
COULD
37 %
PERHAPS/
MAYBE
9%
14-16 November 2013
• <s>Sophie, we both may
be in danger.</s>
• <s>Sofi,
mums
abiem
gresia pavojus.</s>
Innovative Information Technologies for Science, Business and Education
IIT-2013
Translational paradigms:
different degrees of likelihood
MUST GAL ‘maybe’, GALĖTI ‘can/may/might’ 3 %
E - original

There was no telescreen,
but there must be hidden
microphones.
A few must not have reached
their homes yet.
14-16 November 2013
LT - translation
GALĖTI
Teleekrano nebuvo, bet
‘can/may/might’ galėjo būti paslėptų
mikrofonų.
GAL
‘maybe’
Gal dar ne visi grįžę.
Innovative Information Technologies for Science, Business and Education
IIT-2013
Translational paradigms: different
degrees of likelihood
SL
“It might be a long
time," said O'Brien.
Perhaps that is
because Mr. Wren
comes here every year!
Na tai tikriausiai šia
suknia vilkėjo
elektromagnetinė
stebuklinga mergelė
Maranela.
14-16 November 2013
TL
MIGHT
TIKRIAUSIAI, VEIKIAUSIAI
‘most probably’
TURĖTI ‘have to/ must’ 3 %
— Tikriausiai dar negreit,
— tarė O'Brajenas.
PERHAPS 
TIKRIAUSIAI, TURBŪT ‘most
probably’
9%
Tikriausiai todėl, kad ponas
Renas čia atvyksta
kiekvienais metais!
TIKRIAUSIAI, TURBŪT
‘most probably’ 
PERHAPS, MAYBE 8 %
So, perhaps it was the
miraculous electromagnetic
maiden, Maranela, who
wore this gown.
Innovative Information Technologies for Science, Business and Education
IIT-2013
Semantic core of evidentials inferential source: LT MATYT
• Direct perception (non-modal)
– Matau saulę.
‘I see the sun’
• Indirect perception  evidential meaning
– Iškart matyti, kad esate kūrusi muziką.
‘It is evident that you have created music’
• Parenthesis evid- ep modal meaning
– Jis, matyt, mirtinai nuobodžiavo.
‘‘Apparently, he was bored to death’
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
Correspondences of MATYT
MATYT
LT_orig
E_transl
LT_transl
E_orig
n=62
%
n=90
%
seem/appear/look
20
32
11
12
must
7
12
22
24
apparently/obviously/evidently
14
23
29
32
most likely/probably/no doubt
quite possibly/perhaps
guess/think/suppose
Ø
3
5
2
2
1
2
5
6
1
2
9
10
15
24
12
13
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
Inferential sources & MATYT (1)
• Evidence inferred from observed results
• …langelį uždengia blyški paklodė. Matyt, prausis.
• …the window was covered by a pale sheet. He was
obviously going to bathe.
• …kiemas buvo pilnas žvėrelių. Šitame kelio ruože,
matyt, buvo įrengta požeminė perėja žvėreliams.
• …the yard was full of animals. It seemed that there
was an underground pedestrian walkway on this
side of the highway.
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
Inferential sources & MATYT (2)
• Evidence inferred by logical reasoning
• Pačiam vienam būtų visai prastai. Matyt,
ir Vytautas negali likti vienas.
• If I were left all by myself, it would be
much worse. Quite possibly, Vytautas
can't stay alone either.
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
Summing up
• Choice of V/ADV strategy
– Dominance of mod ADVs in LT fiction
•  correspondence in SL & TL
• Overlap of ep necessity and possibility in LT
– language-specific conceptualization of modal
epistemic-inferential necessity (& possibility)
– different nuances of evidence  different degrees of
certainty
(Boye & Harder 2009)
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
References
•
•
•
•
•
•
•
•
•
•
•
Aikhenvald, Alexandra Y. 2004. Evidentiality. Oxford: Oxford University Press.
Boye, Kasper and Peter Harder. 2009. Evidentiality: Linguistic categories and grammaticalization.
Functions of Language 16 (1), 9-43.
Cornillie, Bert. 2007. Epistemic Modality and Evidentiality in Spanish (semi-) auxiliaries. A Cognitivefunctional Approach. Berlin - New York: Mouton de Gruyter.
Cornillie, Bert. 2009. Evidentiality and epistemic modality: on the close relationship of two different
categories. Functions of Language. 16 (1), 44-32.
de Haan, Ferdinand. 1999. Evidentiality and Epistemic Modality: Setting Boundaries. Southwest Journal
of Linguistics 18, 83-101.
Mortelmans, Tanja. 2000. On the ‘evidential’ nature of the ‘epistemic’ use of the German modals müssen
and sollen. Belgian Journal of Linguistics 14, 131-148.
Palmer, Frank R. 2001. Mood and modality. Cambridge: Cambridge University Press.
Pietrandrea, Paola. 2005. Epistemic Modality. Functional Properties and the Italian System. Amsterdam –
Philadelphia: John Benjamins.
Plungian, Vladimir A. 2001. The place of evidentiality within the universal grammatical space. Journal of
pragmatics 33, 349-357.
Simon-Vandenbergen, Anne-Marie and Karin Aijmer. 2007. The Semantic Field of Modal Certainty. A
Corpus-Based Study of English Adverbs. Berlin & New York: Mouton de Gruyter.
Squartini, Mario. 2004. Disentangling evidentiality and epistemic modality in Romance. Lingua 114 (7),
873-889
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
Explanation: change of meaning
• (Johansson 2001: 238) ’The general
background for zero correspondences is
the weakened meaning of SEEM.’
• (Aijmer
2000):
SURELY

Grammaticalization+Subjectification:
’unidirectional development from the
literal meaning of certainty  emphasis
 a degree of uncertainty.’
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
Probability seem in E, SE & NO
The findings are in line with results in SE & NO
• The frequency of zero correspondences of
SEEM in the Norwegian text is about 16%
(Johansson (2001: 238)
• Ø correspondences of Swedish particles
– VISST (‘?seems’) in English is about
16.9% (Aijmer 1996: 411);
– VÄL (‘probably’/‘perhaps’) is 38% (Aijmer
1996: 415);
14-16 November 2013
Innovative Information Technologies for Science, Business and Education
IIT-2013
Ep possibility+ necessity:
STRATEGIES
AUX / V
ADV
n
f/ 1000
n
f/1000
EN
2019
4,15
860
1,77
LT
174
0,59
897
3,02
LL
14-16 November 2013
+1044,11
-125,68
Innovative Information Technologies for Science, Business and Education
IIT-2013