Content uploaded by Andrea Götz
Author content
All content in this area was uploaded by Andrea Götz on Nov 16, 2017
Content may be subject to copyright.
Translating doubt: e case of the Hungarian
discourse marker vajon
Introduction
is chapter focuses on the translation of the Hungarian discourse marker
(DM) vajon [I wonder] in translations and interpretations of English
European Parliamentary speeches, with a special attention to the issues
of its source forms. rough the examination of vajon’s translation and
interpretation data, the chapter seeks to probe the issues of DM transla-
tion is a broader sense.
is marker occurs only in interrogative clauses and has no direct
counterpart in English. Although vajon in traditional grammar is described
as an interrogative particle with modal properties (see Keszler ), it
can, nevertheless, be roughly equated with the cognitive verb wonder, with
the important dierence that, in Hungarian, vajon appears in the clause
that wonder introduces, in English. e material of this study is collected
from contexts in which an English clause was translated into Hungarian
using vajon in the target context.
Translating DMs is usually considered to be problematic, frequently
leading to DMs being omitted in the target text. However, the opposite
side of this issue, addition, that is, DMs being added to the target text, has
received less attention. Additions are understood as forms which are added
to the target text, that is, for an added target form the source containing a
corresponding form. e role of added DMs in interpreter-mediated dis-
course has been addressed in a relevance-theoretic framework (Blakemore
and Gallai ). However, it remains to be seen if added DMs in trans-
lation function or come about in a similar way. Ultimately, the chapter
investigates what triggers and what explains the use of vajon in translations
from English into Hungarian.
First, the chapter provides a brief overview of issues relating to dis-
course marker translation and the source forms of added element in the
target text, that is, zero forms (the rst section). e third section deals
with the functions of vajon. Following this, the fourth section introduces
the aims, methods and corpus of this study, and the h section explores
the results. Finally, the conclusions of the study are presented, alongside
some concluding remarks.
Do discourse markers “translate”? Norms
of language and translation
For discourse markers (DMs), we nd a plethora of terms and a variety
of roles in the literature (for an overview see Blakemore ), they are
described as discourse connectives (see Schourup ), discourse mark-
ers (Schirin , Blakemore ), and discourse particles (Aijmer et al.
), to name a few. However, DMs are usually accepted to function in one
of the following ways: they can () express logical relations between units of
discourse, creating cohesion (Fraser ); or () work on dierent “planes
of discourse” thus maintaining its coherence, relating not only discourse
units to each other but also the attitudes and personal perspectives of the
speakers’ to the discourse itself (Schirin ); or ) guide the process of
pragmatic interpretation in utterance comprehension (Blakemore ).
Within the framework of Relevance eory (Sperber and Wilson
), DMs are thought to contribute to inferential processes (Blakemore
). is contribution to the pragmatic content of an utterance or to
the inferential process of comprehension does not necessarily have to be
expressed through DMs, as this, depending on the particular language or
the context, may not even be possible.
DMs are widely assumed to pose a particularly dicult challenge for
translators, as Blakemore (: ) notes “[…] the translation of these
Translating doubt
expressions [discourse markers] is notoriously difcult”. Others outright
conclude that “discourse particles [equivalent to DMs in this study] are
not expected to reach a high degree of intertranslatability” (Aijmer et al.
: ). Against this background, it might seem counter-intuitive to
study discourse markers in their translation data.
As noted above, languages can dier signicantly in terms of pragmatic
marking. e importance of cross-linguistic pragmatic dierences – and
their eect on translation – is not unknown to contrastive linguistics, or
indeed, translation studies. is dierence, for example, in relation to
Hungarian and English, has been pointed out by Hervey (), and could
result in dierent linguistic devices fullling the same pragmatic function,
for example, a pragmatic function conveyed in Hungarian through word
order or stress can potentially be realized in English through the use of DMs.
Not surprisingly, the translation of DMs usually produces a high
number of “zero-correspondences” (Aijmer ), “zero-translations” or
“zero-forms” (Degand ). ese terms describe source DMs which in
the target text are “le untranslated” (Aijmer : ). In this chapter,
these instances are referred to as zero forms. Non-translation is indeed
assumed to be so widespread that some researchers state: “Omission of
discourse markers seems to be a rather general phenomenon” (Aijmer :
). Other than cross-linguistic dierences, the lack of a target form in the
target text could be attributed to a translation strategy or the fact that a
DM in the target text would be redundant due to sucient information
being already present, derived from other linguistic cues. Contexts can be
clear enough “to allow for non-use or non-translation” of DMs (Traugott
: ). Zero forms are thus normally understood as “non-translations”,
however, the lack of a target form does not necessarily mean that the source
form has not been translated, or it was purposefully omitted.
However, we should not conclude that the presence or absence of
DMs do not inuence meaning. Although DMs make no contribution to
the propositional content of an utterance, on the other hand, in the infer-
ential process they “encode procedures for the recovery of implicatures or
constraints on pragmatic inference” (Blakemore and Gallai ).
As a result, omission, deletion, that is, the lack of a corresponding form
in the target text, does not necessarily equal a lack of translation. Although
a DM might not be explicitly present, it could have still inuenced the com-
prehension of the given utterance which can be reected in its translation.
But beyond diverging cross-linguistic norms of pragmatic marking
and the peculiarities of individual translation situations, there are transla-
tion specic phenomena which could impact the DMs present in target
texts. e controversial topic of so-called translation universals further
complicate the topic of translated DMs (Baker , ; Laviosa ),
as based on an interpretation of the explicitation hypothesis (Blum-Kulka
), we could expect target texts to feature more or more explicit DMs
than their sources. Although translation universals, laws and tendencies
remain a hotly debated topic, the notion of language norms inuencing
translation is uncontroversial.
Such language norms can entail diering norms of textual cohesion
and conjunction use between languages (see Becher ), and the language
contact facilitated by translation could even lead to changes in the norms
of the target language (Bisiada ). Adhering to these cohesive norms
of the target language could potentially result in an abundance of zero
forms compared to the source, but, crucially, without adversely aecting
the cohesion of the target text. is issue is further complicated by genre
norms which additionally shape the language and structure of texts, as well
as leaving their mark on the translations by determining what translation
strategies translators choose (Károly : ).
In conclusion, two points should be noted: () zero forms can be
as informative about translation as other target forms; and () the role
language and genre norms play in translation should be factored into the
contrastive study of translations.
Vajon and rhetorical questions
Traditionally treated as an interrogative modal particle (Keszler )
occurring in questions, vajon is also described as a discourse marker fre-
quently present in rhetorical questions with discourse-specic functions
(Schirm ).
Translating doubt
Since vajon fulls no grammatical role, its presence or absence does not
make a sentence grammatical, it is mainly understood to have a pragmatic
function: to modify the modal value of the utterance (Keszler ), to
express the speaker’s attitude, with the specic aim to raise doubt about
the validity of the proposition expressed, or to show the speaker is contem-
plating a question, wondering about its possible answers (Schirm ). As
vajon does not inuence the well-formedness of a sentence, it is an optional
choice for speakers and translators alike.
What makes vajon-questions dierent from standard interrogatives is
that although the presence of vajon in a question indicates the speaker is
curious about the answer, the hearer is nevertheless not implored to answer
(Gyuris : ). In other words, vajon-questions are not used to elicit
information – albeit they can be answered.
It is thus not surprising that we nd vajon in abundance in rhetorical
questions, as rhetorical questions function in a similar manner, as rhetorical
questions do not seek to elicit information, but rather to trigger an eect
from the hearer, a mental response, while showing “the speaker’s commit-
ment to its implied and inferrable answer” (Ilie : ). Polar rhetorical
questions can also function as indirect statements to the opposite eect, as
they “are interpreted as an assertion of the opposite polarity” (Han ).
Rhetorical questions play a role in verbal aggression and (im)politeness
strategies (Frank , Schaer ), and depending on the discourse
context, as retorts (Schaer ). e eect of these questions can fur-
ther vary with status of the discourse participant asking them (Ilie ).
Although all rhetorical questions work along the same principle,
they can full dierent discourse-specic functions based on their dis-
course context (Ilie : ). Rhetorical questions in a political context,
for example, can be used to manipulate opinions (Ilie ), or to insult
opponents (Ilie ). is is line with Schirm’s () ndings about the
functions of vajon-questions regarding their role in verbal aggression and
(im)politeness strategies in Hungarian parliamentary speeches.
Nevertheless, vajon neither turns every question into a rhetorical one,
nor is it a compulsory feature of Hungarian rhetorical questions. However,
vajon does seem to profoundly inuence utterance comprehension. e
divergence in the interpretation of vajon-questions as rhetorical or con-
templative questions can be explained as follows: the hearer realizes the
speaker has not posed the question to elicit information, but rather to lead
the hearer to an answer implied by the speaker – if, however, the answer is
not evident to the hearer, vajon-questions will not be interpreted as rhe-
torical (Gyuris : ), as the question itself cannot guide the hearer
to an implied answer.
Research design
Research questions and hypotheses
As discussed previously, translating DMs is not straightforward, however,
additions and zero forms could shed some light on this process. Since DMs
do make a contribution to utterance comprehension, and target language
norms might heavily inuence the use DMs, as well as translation-specic
or cognitive phenomena arising from the act mediating language, added
DMs should be studied as they can be informative about both source and
target texts, and translation itself.
Consequently, this study seeks to answer the following questions: ()
What is the marker vajon used for to translate? () Are these occurrences
“translations” or “non-translations”? () If we can we identify correspond-
ing source forms, what are these? () Do translations and interpretations
show dierent source forms?
Based on the results and assumptions found in the literature, the fol-
lowing hypotheses were formed: () most target vajon forms will have cor-
responding source forms; () the translation and interpretation data will
show strong links between particular source and target forms. Issues not
addressed by the hypotheses will be discussed in the concluding remarks.
Corpus and methods
e translation and (simultaneous) interpretation data were collected from
European Parliament (EP) speeches. EP speeches are available online on
Translating doubt
the website of the EP, including audio and video les, as well as the written
version and translation of the speeches. However, a substantial translation
backlog and missing recordings of older speeches restrict the range of
available material. EP speeches are transcribed, corrected, and published
together with their translations. Verbatim transcripts are not available,
and the normalized speech transcripts dier from what was actually said.
For the study, contexts containing vajon as a target form in the trans-
lations were specically searched for. e interpretation data came from
the same speeches with the relevant contexts transcribed and analysed.
Unfortunately, it is not possible to search the speeches directly, only the
normalized transcripts and translations.
e translation corpus contains authentic English source texts and
their translations in Hungarian. It consists of occurrences of vajon in
the Hungarian translations, and their English source contexts. e inter-
pretation data is analysed in seventy-ve contexts, in the original speeches
in English and their simultaneous interpretation in Hungarian.
e present study thus applies a contrastive analysis to translation data.
e so-called translation method (Aijmer and Simon-Vandenbergen ),
an extended form of contrastive analysis, developed to study pragmatic mark-
ers in translation data. By studying the connection between source and target
forms, the method aims to shed light on the functions of pragmatic markers.
Target forms are thought to reect and clarify the function of source forms.
e present study explores the connections of target forms containing vajon
and forms present in their source contexts that can be related to vajon. It
remains to be seen whether the translation method’s assumption holds true
for markers like vajon, which have no direct counterpart, therefore it is not
obvious what single form or structure they could reect on.
In the following, I investigate discrepancies between the transcripts
and the spoken material and discuss some potential inuences that could
have shaped their translations.
e eect of transcription
Properties of spoken language are found in all spoken material. Although
some MEPs read out loud pre-prepared speeches, even in these we sometimes
nd these characteristics. Many speeches due to these would not be suit-
able for serving as source texts for translation. Although it is not clear what
editing process EP speech transcriptions go through (they are proofread
according to the EP website), some dierences between the language of
the spoken and transcribed speeches are evident, with the latter being more
formal. is means that translations will start from a more formal basis
and presumably remain consistent with it in formulating the target texts,
which in and of itself can be a reason why interpretation and translation
data profoundly dier, specically for the marker under examination in
this chapter, and in general.
e transcriptions show a clear preference for a more formal register:
in the present corpus expressions characteristic of spoken language were
edited or omitted, for example, I think, actually, you know. More substan-
tial edits include examples of independent interrogative sentences in the
spoken speeches being turned into subordinate interrogative clauses in the
transcripts. ese included clauses which contained vajon in the Hungarian
translations. In cases such as these, the spoken and written material may
have given very dierent signals to translators and interpreters, further
compounding the dierences of the two sets of data.
Other edits would not necessarily interfere with the translation or
interpretation of the DM vajon, though they still show a formal preference.
Such modications include substituting if for whether.
e transcripts of the seventy-ve authentic English EP speeches
interpreted into Hungarian were examined and contrasted with the lan-
guage of the spoken version of the speeches. Of the seventy-ve, the lan-
guage of thirty-nine speeches was unaltered in the transcript regarding
the relevant contexts, with thirty-six showing modications to variant
degrees.
In only eleven did these changes aect the contexts containing the
linguistic forms under investigation. In only eight of these were more
substantial corrections found, in these whether was either omitted from
the transcript although it was present in the speech (), or it was added
to the transcript (). In one case wonder was changed to wonder if, in
another it was changed to wonder whether in the transcript, thus subordi-
nating the question succeeding it. In the remaining cases () if or wonder
Translating doubt
if occurring in the speeches were replaced by whether or wonder whether
in the transcript, which can be seen as an eort to achieve a more formal
register.
Results and discussion
In the following, I present the translation and interpretation data of vajon
on a corpus of EP speeches, with authentic English texts translated and
interpreted into Hungarian. As many dierent structures contained vajon
(see Table ), their classication needs to be addressed.
e marker vajon can appear in interrogative sentences, as well as
subordinate interrogative clauses, which can be made evident through the
presence of a subordinator or be implicitly clear from syntax. A distinction
is therefore made between occurrences where such syntactic properties were
explicit and implicit. Instances of vajon where no explicit linguistic form
of subordination is present are labelled “free”. Vajon appearing together in
a clause with the polar particle –e are denoted as vajon … –e. Vajon and
vajon … –e could also be present in verbal phrases expressing cognition.
Dierent target structures were connected to dierent source forms, which
makes it necessary to establish separate categories for target and source
structures such as these.
Translation data of vajon: English source texts translated into Hungarian
From the circa search results, in which vajon was found in the Hungarian
translation of English texts, were selected. e English source texts fea-
tured the following forms: zero form (), whether (), wonder whether
(), wonder if (), wonder (), and whether or not (). is shows that in
most cases, target vajon forms represent additions without any identiable
corresponding linguistic form in the English source context. Table shows
the prevalence of source forms.
Table : Source forms in the translation data
Source forms in the corpus Number of
occurrences of all occurrences ()
zero form .
whether and whether or not .
wonder whether and wonder if .
wonder .
As Table demonstrates, explicitly subordinated clauses form the
second most populous group () shortly aer zero forms, either containing
whether or if. e Hungarian translations (see Table ) contained vajon
and vajon … –e forms, or expressions featuring either. ese expressions
included tűnődik [to wonder] (), kíváncsi [to be curious] (), szeretné tudni
[would like to know] (), felmerül [to occur to someone] (), all appearing in
the rst person singular, reecting the speaker’s point of view. ese twelve
contexts make up . per cent of the total number of contexts.
Table : Target forms in the translation data
Target forms in the corpus Number of
occurrences of all occurrences ()
vajon (free) .
vajon … –e (free) .
vajon in a subordinate clause .
vajon …–e in a subordinate
clause .
Of all occurrences of vajon (), were free, two appeared in sub-
ordinate clauses, and a further three in subordinate clauses as part of an
expression mentioned above. From the occurrences of vajon …–e,
were free, were found in subordinate clauses, and a further nine in
subordinate clauses featuring one of the expressions above. As shown
Translating doubt
in Table , this comes to a total of free vajon and free vajon … –e
forms, with far more vajon … –e forms being present in subordinate clauses
than vajon forms. In Table the number of occurrences in subordinate
clauses is combined with those containing the aforementioned expressions.
Table demonstrates that vajon was more likely to occur in main clauses
than vajon … –e, and vajon … –e was far more likely than vajon to be present
in subordinate clauses, but both were frequent in their free form.
In the following, I examine the most frequent target and source forms
(see Tables and ). e most numerous source form of Hungarian trans-
lations containing vajon, came from source contexts featuring zero forms,
whether, and wonder whether/if. As seen in Table , in most target contexts
for zero forms, we nd vajon and, to a lesser degree, vajon … –e. In the few
subordinate clauses (), vajon … –e appears more frequently in the target
texts, though these target forms are very low. No zero forms were translated
with cognitive verbs in subordinate clauses.
Table : Hungarian target forms of English zero forms
Target forms of zero forms Number of
occurrences of all zero forms ()
vajon (free) .
vajon … –e (free) .
vajon in a subordinate clause .
vajon … –e in a subordinate clause .
Table shows the Hungarian target forms of contexts containing
whether. In the target contexts, we nd vajon … –e in either subordinate
or main clauses, in a combined number of occurrences. No free target
vajon forms are present, however, in one case a source context was translated
with vajon in the target text as being present in a subordinate clause. And
although this context does not contain the polar particle –e, it nevertheless
represents a yes/no question. e single source context containing whether
or not is treated as a variant of whether here.
Table : Hungarian target forms of whether
Target forms of whether Number of
occurrences of all zero forms ()
vajon (free)
vajon … –e (free) .
vajon in a subordinate clause .
vajon … –e in a subordinate clause .
Target forms of whether point to the importance of syntactic features.
As mentioned before, a subordinator does not necessarily have to be pre-
sent in subordinate clauses, and the polar particle –e is not compulsory for
polar questions. Looking at it more closely, beyond the apparent linguistic
forms, free vajon … –e forms are also found in subordinate clauses. e one
occurrence of vajon in a subordinate clause represents a polar question,
without the particle –e making it explicit.
Although the translations of wonder whether and wonder if show some
variation, their low occurrence makes it dicult to substantiate any claim
that these forms trigger dierent translations. In the transcriptions if and
whether are at times substituted, which suggests they were treated as inter-
changeable. Since these forms share some of the same syntactic properties,
to combine the translation data could oer more insights than discussing
them separately. Accordingly, Table shows the target forms for wonder
whether and wonder if.
Table : Hungarian target forms of wonder whether/if
Target forms of wonder whether/if Number of
occurrences
of all zero forms ()
vajon (free) .
vajon … –e (free)
vajon … –e in a subordinate clause .
vajon part of an expression, in a
subordinate clause .
vajon … –e part of an expression,
in a subordinate clause .
Translating doubt
In one translation of a source context containing wonder whether, we
nd vajon. In this case, the English subordinate interrogative clause was
transformed into an interrogative sentence in the Hungarian translation,
as shown in Example ().
() (S/E) I wonder whether he could give us a timeframe on this issue […]
(T/H) Vajona soros elnök úr fel tudna vázolni egy ütemtervet e
kérdést illetően […]
[Vajon could the President-in-Oce draw up a time frame regarding
this question]
e last remaining source form type not yet discussed is wonder, with
four occurrences. In two cases, we nd vajon in corresponding target con-
texts, in two other, vajon appears together with cognitive verbs in subordi
-
nate clauses. It should be noted that the source contexts that were translated
with these cognitive verbs contained wonder. As we have seen (Table ),
not all wonder whether/if forms were translated in this fashion, however,
most of them were. is points to the issue of translatorial choice, as it
aects translation data. Translators’ decisions could be swayed by factors
such as the register and style of texts, not to mention institutional conven-
tions, possibly guidelines (for example the EU’s “clear writing” initiative).
When analysing translation data, such translation related considerations
should be acknowledged.
In the Hungarian target contexts, we have observed a variety of transla-
tion solutions. Table presents a summary of the Hungarian translations of
the English source contexts. We see striking dierences between the target
and source forms of individual linguistic expressions.
e results of this corpus make it clear that greater attention has to be
paid to the properties of the individual source and target contexts before
we can establish connection between them. If we do not consider the syn-
tactic properties of both source and target contexts (whether or not the
polar particle –e is present, or if it is a subordinate clause, or if it contains
a complex expression), subtle but not inconsequential dierences cannot
be detected. However, the small data set of the corpus and linguistic and
well as extra-linguistic factors (norms, guidelines, translators’ choices)
allow for only cautious conclusions to be drawn.
In the translation data, vajon shows a stronger correlation with zero forms:
it tends to be added to the target text. Nevertheless, zero forms are the most
numerous source form type, and as such, they are frequent source forms
for target vajon …–e forms as well. Furthermore, in Hungarian translations
that correspond to English contexts containing whether, we nd almost
exclusively vajon …–e forms, save for that one occurrence of vajon which is
in a subordinate polar interrogative clause. is points to the need to extend
to grammatical features as well when studying pragmatic linguistic forms.
Interpretation data of vajon: English speeches interpreted into Hungarian
As we have seen before, there is some discrepancy between the source
texts used for translation and interpretation. ese dierences need to be
accounted for when the interpretation data of the seventy-ve selected
contexts is analysed.
English source forms, drawn from the transcribed speeches, are pre-
sented in Table . Most of the occurrences are zero forms. However, most
of these zero forms in the interpretation data, in contrast to translation
data, correspond with zero forms in Hungarian as well (see Table ).
Table : Translation data in the corpus:
English source forms and their Hungarian target forms
free free in a sub. clause in expression in a
sub. clause
vajon vajon … –e vajon vajon … –e vajon vajon … –e
zero form – –
whether – – –
wonder whether – – –
wonder if – –
wonder – – – –
whether or not – – – – –
Translating doubt
Table : English source forms in the interpretation data
Source forms in
the corpus
Number of
occurrences
of all
occurrences ()
zero form .
whether and whether or not and if .
wonder whether and wonder if .
wonder .
Since the contexts for the corpus were selected based on their trans-
lations, and that the interpretations cannot be searched, they could not
have been selected for the presence of vajon. is discrepancy between the
prevalence of vajon in translation and interpretation data points to their
dierence. As seen in Table , target forms in the interpretation data pre-
sent a greater variety than in the translation data. We nd the particle –e
in interrogative sentences and in subordinate clauses. Since translation
data was specically collected to include vajon, it did not have data on the
particle –e, occurring on its own.
Table : Hungarian target forms in the interpretation data
Target forms in
the corpus
Number of
occurrences
of all
occurrences ()
zero .
vajon .
vajon … –e .
–e .
–e in a subordinate clause .
cognitive verbs .
–e in a sub. clause with cognitive verb .
vajon …–e in a sub. clause .
vajon … –e in a sub. clause with cognitive verb .
context missing .
Among the cognitive verbs used in the Hungarian interpretations,
we nd some that were not present in the translation data, including nem
tudom [I don’t know] () and gondolkodik [to think] (). Nem tudom cor-
responds to occurrences of wonder (), wonder whether (), and wonder if
(). In this usage, this expression is informal, probably too much so to be
used in translation. Other cognitive verbs and expressions, such as “I am
curious” and “I’d like to know”, were not used by interpreters.
In Table , we see the Hungarian interpretations of contexts which in
the translation data showed no corresponding forms to target vajon forms.
Most of these occurrences represent zero forms in the interpretation data.
However, the only occurrences of vajon forms, altogether ve, are found
responding to English zero forms. is, although in a weak form, could
indicate a connection not between vajon and linguistic forms, but rather,
certain contexts.
Table : Hungarian target forms of English zero forms
Target forms of
English zero forms
Number of
occurrences
of all
occurrences ()
zero
vajon .
vajon … –e
–e
–e in a subordinate clause .
context missing
In the following, I examine the target forms of whether as shown by
Table . If and whether or not have only one occurrence each, the contexts
containing if and whether or not were interpreted with –e and –e in subor-
dinate clause in the corresponding contexts, respectively.
Table : Hungarian target forms of whether
Target forms of whether Number of occurrences of all occurrences ()
zero .
–e .
–e in a subordinate clause .
context missing .
Translating doubt
In the English speeches, we nd four occurrences for wonder whether, and
six for wonder if, due to this low number, their interpretation data is shown
combined in Table .
Table : e Hungarian target forms of wonder whether/if
Target forms of
wonder whether/if
Number of
occurrences
of all
occurrences ()
zero
–e
cognitive verbs
–e in a sub. clause with cognitive verb
e interpretations of contexts containing wonder () include the
particle –e (), vajon … –e in a subordinate clause (), and with the cogni-
tive verbs gondolkodik [think] () and nem tudom [I don’t know] () as
target forms.
Table summarizes the interpretation data found in the corpus. As we
see, zero forms represent both the most frequent source and target forms.
e data do not show strong enough correlations between source and target
forms to postulate a connection between them. In the interpretation data,
we nd a very low prevalence for vajon, which only occurs once and four
other times in the form of vajon … –e.
Table : Interpretation data in the corpus
Interpretation data of vajon zero f. whether if
wonder whether
or not
whether if
zero form – –
vajon – – – – –
vajon … –e – – – – –
–e – –
–e in a subordinate clause – – –
cognitive verbs – – – –
–e in a sub. clause with cognitive verb – – – – –
Interpretation data of vajon zero f. whether if
wonder whether
or not
whether if
vajon … –e in a sub. clause – – – – – –
vajon … –e in a sub. clause with
cognitive verb – – – – – –
context missing – – – –
However, even the low occurrence of vajon-forms could signify that
interpreters detected a discourse function in the source context and sought
to convey it.
In the translation data, a weak link between wonder and vajon, and more
obviously, between wonder and cognitive words, as well as between whether
and subordinate clauses frequently containing –e can be established. e
data from simultaneous interpretation underlines how important the par-
ticulars of the translation or interpretation situation are. Apart from the
obvious time constraint, the dierence between interpretation and trans-
lation units can contribute to the dierences between translation and
interpretation data. For the translation method to be applicable to inter-
pretation data, presumably a much more substantial data set is required.
Conclusion
In the translation data, zero form contexts () were just about outnum-
bered by contexts containing a corresponding linguistic form in the source
(). In this respect, the rst hypothesis (most target vajon forms will have
corresponding source forms) was conrmed. However, the high number of
zero forms and the small scale of the study do not allow for generalizations.
Translating doubt
As regards to the second hypothesis (the translation and interpretation
data will show strong links between particular source and target forms), I
have established strong links only between certain forms (e.g. vajon … –e;
vajon in a subordinate clause and whether, etc.), and only in translations.
As such, the second hypothesis is not conrmed, as this question requires
further research.
To the question what vajon is used for to translate between English
and Hungarian, no clear answer emerged in either data set. However, to a
certain degree, vajon structures can be linked to certain source structures,
for example, whether and if as subordinators correspond with subordinated
clauses that frequently feature –e and vajon, and more specically, correlate
with the particle –e, rather than with vajon. On the basis of this, it can be
argued that vajon is an addition, rather than a “translation”, that is, vajon
is not so much a translation of a specic source form or source forms, but
rather a tool with which the pragmatic eect of the source context can be
reproduced in the target language.
Building on this assertion, it stands to reason that the discussion should
move away from how DMs are translated to how the contexts which con-
tain them are translated. All data came from rhetorically charged contexts.
Translators could have detected this, interpreted English source contexts
as having a rhetorical function, and as a result, translated them into using
vajon, an otherwise frequent DM in Hungarian rhetorical questions. is,
however would suggest, this data set is concerned with the translation
of rhetorical questions rather than that of vajon. In this view, applying
Relevance eory, translators insert vajon, as participants in attributive
language use, conveying their interpretation of a given text or utterance
in translation (Gutt ).
Supporting the relevance theoretical view, additions in the translation
data were found to be in line with Blakemore and Gallai’s () ndings
about DM additions in interpretation data, which they understand to reect
the interpreters’ thought processes. As such, we can assume translators and
interpreters detect the strong personal commitment vajon-questions and
rhetorical questions share, and seek to communicate it, thus reecting the
speaker’s point of view. To test this, we should compare the prevalence of
vajon-questions in authentic Hungarian discourse with that in translated or
interpreted discourse. e rhetorical and subjectivized, contemplative uses
of vajon are hard to separate, as it is the speaker’s intent that gives rhetori-
cal questions their potency. Presumably, vajon’s rhetorical functions can
be derived from this usage that triggers the hearer to momentarily adopt
the speaker’s perspective, and thus steer the hearer’s inferential process so
as to yield the interpretation most relevant from the speaker’s perspective.
As interpretation situations dier greatly, it is worth mentioning that
Blakemore and Gallai’s () data came from interpreter mediated police
interviews, as data from various forms of interpretations are expected to
dier fundamentally. Although interpretation data were found to be less
suitable for the translation method, the method was somewhat successful
in that it has extended the range of expressions which can be connected to
vajon through the inclusion of this data set. In addition, the high prevalence
of the particle –e, outnumbering vajon, called attention to the particle’s
role beyond syntax. Interpretation data also helped to underline the clas-
sication of vajon in translations as an addition. However, given the small
scale of the corpus, to substantiate any claim further research is needed.
Primary source
European Parliament speeches: <http://www.europarl.europa.eu>
Bibliography
Aijmer, Karin. . “e meaning and functions of the Swedish discourse marker
alltså – evidence from translation corpora”, Catalan Journal of Linguistics , –.
Aijmer, Karin, Ad Foolen and Anne-Marie Simon-Vandenbergen. . “Pragmatic
markers in translation: methodological proposal”. In Kerstin Fischer (ed.),
Approaches to Discourse Particles. Amsterdam, Boston and London: Elsevier.
–.
Translating doubt
Aijmer, Karin, and Anne-Marie Simon-Vandenbergen. . “A model and a meth-
odology for the study of pragmatic markers: e semantic eld of expectation”,
Journal of Pragmatics (), –.
Baker, Mona. . “Corpus linguistics and translation studies: implications and
applications”. In Mona Baker, Gill Francis, Elena Tognini-Bonelli and John
Sinclair (eds), Text and Technology: In Honour of John Sinclair. Philadelphia,
PA: J. Benjamins Pub. Co. –.
Baker, Mona. . “Corpus-based translation studies: e challenges that lie ahead”.
In Juan C. Sager and H. L Somers (eds), Terminology, LSP, and translation studies
in language engineering in honour of Juan C. Sager. Amsterdam and Philadelphia,
PA: J. Benjamins Pub. Co. –.
Becher, Viktor. . “When and why do translators add connectives?: A Corpus-
Based Study”, Ta rge t (), –.
Bisiada, Mario. . “Structural eects of English–German language contact in
translation on concessive constructions in business articles”, Text & Talk (),
–.
Blakemore, Diane. . Relevance and Linguistic Meaning: e Semantics and Prag-
matics of Discourse Markers. Cambridge: Cambridge University Press.
Blakemore, Diane. . “Discourse markers”. In Laurence R. Horn and Gregory L.
Ward (eds), e Handbook of Pragmatics. Malden, MA and Oxford: Blackwell.
–.
Blakemore, Diane, and Fabrizio Gallai. . “Discourse markers in free indirect style
and interpreting”, Journal of Pragmatics , –.
Blum-Kulka, Shoshana. . “Shis of cohesion and coherence in translation”. In
Juliane House and Shoshana Blum-Kulka (eds), Interlingual and Intercultural
Communication: Discourse and Cognition in Translation and Second Language
Acquisition Studies. Tübingen: Narr. –.
Degand, Liesbeth. . “On describing polysemous discourse markers. What does
translation add to the picture”. In A.M. Simon-Vandenbergen, Stefaan Slem-
brouck, Miriam Taverniers and Mieke van Herreweghe (eds), From Will to
Well: Studies in Linguistics, Oered to Anne-Marie Simon-Vandenbergen. Gent:
Academia Press. –.
Frank, Jane. . “You call that a rhetorical question?”, Journal of Pragmatics (),
–.
Fraser, Bruce. . “What are discourse markers?”, Journal of Pragmatics (),
–.
Gutt, Ernst-August. . Translation and Relevance: Cognition and Context, nd
edn. London: Routledge.
Gyuris, Beáta. . “Megjegyzések a pragmatika tárgyáról és hasznáról”, Magyar
Nyelv (), –.
Han, Chung-hye. . “Interpreting interrogatives as rhetorical questions”, Lingua
(), –.
Hervey, Sándor G.J. . “Speech acts and illocutionary function in translation
methodology”. In Leo Hickey (ed.), e Pragmatics of Translation. Clevedon:
Multilingual Matters. –.
Ilie, Cornelia. . “Question-response argumentation in talk shows”, Journal of
Pragmatics (), –.
Ilie, Cornelia. . “Semi-institutional discourse: e case of talk shows”, Journal of
Pragmatics (), –.
Ilie, Cornelia. . “Insulting as unparliamentary practice in the British and Swed-
ish”. In Paul Bayley (ed.), Cross-Cultural Perspectives on Parliamentary Discourse.
Amsterdam and Philadelphia, PA: J. Benjamins Pub. Co. –.
Károly, Krisztina. . Szövegkoherencia a fordításban. Budapest: ELTE Eötvös Kiadó.
Keszler, Borbála (ed.). . Magyar grammatika. Budapest: Nemzeti Tankönyvkiadó.
Laviosa, Sara. . “Universals”. In Mona Baker and Gabriela Saldanha (eds), Rout-
ledge Encyclopedia of Translation Studies. London and New York: Routledge.
–.
Schaer, Deborah. . “Can rhetorical questions function as retorts?” Journal of
Pragmatics (), –.
Schirm, Anita. . “A diskurzusjelölők funkciói: a hát, az –e és a vajon elemek törté-
nete és jelenkori szinkrón státusa alapján”. PhD thesis, Szegedi Tudományegyetem.
Schourup, Lawrence. . “Discourse markers”, Lingua (–), –.
Sperber, Dan, and Deirdre Wilson. . Relevance: Communication and Cognition.
Oxford: Blackwell.
Traugott, Elizabeth Closs. . “Discussion Article: Discourse Markers, Modal Par-
ticles, and Contrastive Analysis, Synchronic and Diachronic”, Catalan Journal
of Linguistics , –.
