“You are certainly my best friend” – Translating adverbs of evidential certainty in The Picture of Dorian Gray

1 Introduction

The focus of this investigation is to examine a relatively narrow group of modal adverbs from the larger class of sentence adverbs (SAs), namely evidential adverbs of certainty, i.e., adverbs that express certainty as following from evidence. To this end, we will investigate the range of these adverbs in a corpus made up of ten versions of the translation of The Picture of Dorian Gray into Romanian. We aim at discussing the way in which the ten translations chose to render the adverbs in this class and see whether, in doing so, they conform to Berman’s (1990) Retranslation Hypothesis (RH).

2 Theoretical framework

Evidentiality represents an interesting, most likely universal, semantic category of human cognition. There seems to be an almost general consensus in the literature that evidential expressions have propositional scope. This is what usually distinguishes an evidential marker from an expression describing or designating some form of evidence.

Boye (2010, 304) argues that “for a given linguistic expression to be considered as having evidential meaning, it must be attested with a proposition-designating clause as its semantic scope,” which is not a state-of-affairs-designating clause.

Epistemic modality has been defined as “an evaluation of the chances that a certain hypothetical state of affairs under consideration (or some aspect of it) will occur, is occurring or has occurred in a possible world” (Nuyts 2001, 21), and so it seems to be strongly connected to the idea of truth and the authors’ responsibility concerning their statements.

Cornillie (2009, 47) proposes a disjunctive model, where epistemic modality and evidentiality are perceived as distinct categories. This is also our opinion as it goes along with our own analysis of SAs (Protopopescu 2012, 150), in which we agree that evidential adverbs should be analyzed separately from epistemic adverbs in spite of many authors lumping them together under the same label of epistemic adverbs. An important thing that needs mentioning in this case is the fact that these domains are not mutually exclusive. One expression may exhibit either evidential or epistemic readings. Cornillie (2009, 50) also argues that “confusion concerning the overlapping of these domains is due to the frequent association of the mode of knowing and the degree of the speaker’s commitment concerning the proposition. In his view, modes of knowing do not really imply any degree of authorial certainty, evaluation, commitment or likelihood of the proposition to be true” (Almeida 2012, 18).

By far, one of the richest classes of SAs in English, epistemic adverbs appear to have developed gradually in English, see Wierzbicka (2006, 247) and Protopopescu (2012, 137). Wierzbicka (2006, 247) also argues that the evolution and rise of this category in English “is related to the rise of verbal epistemic phrases, such as I presume, I assume, I gather, I understand, and I suppose.” Both categories are the same, and in both cases, they can be linked with the emphasis on the limitations of human knowledge, on the need to distinguish knowledge from judgment, and on differentiating between different ‘degrees of assent’. This seems to fall in line with Vișan’s (2017, 26–7) view that the parenthetical I think may also acquire the function of “an epistemic qualifier (expressing evidentiality and commitment of the speaker) at the level of discourse.”

In reviewing the literature on this topic, Protopopescu (2012, 158) notes that the larger class of SAs exhibit a low degree of syntactic and (even) semantic integration into sentence structure. Upon looking at distributional facts, we find that these adverbs occur in sentence-initial (1a) and sentence-final position (1b), usually separated from the rest of the clause by a pause, as well as in auxiliary position (1c).

These adverbs are modal operators and may be viewed as one-place operators on propositions. Pragmatically, they occur only in constatives strengthening or weakening the speaker’s commitment to the truth of the constative, as in (2a) and (2c). Their characteristic entailment is: It is Adjective that Sentence, as in (2b) and (2d).

Thus, modal (epistemic) adverbs can be grouped into several categories. Simon-Vandenbergen and Karin Aijmer (2007, 84), whose classification is illustrated below, distinguish four subcategories within this group of adverbs:

1. epistemic adverbs, which focus on the speaker’s own judgment of the reliability of truth: certainly, definitely, undoubtedly, no doubt, indeed, surely, decidedly, for sure, for certain, assuredly, and indubitably

2. evidential adverbs, which refer to certainty coming from available evidence: clearly, evidently, manifestly, obviously, patently, and plainly

3. expectation adverbs, which relate the speaker’s expectations to the state of affairs: of course, inevitably, naturally, and necessarily

4. speech act adverbs, which refer to speech acts that could potentially be used to support the speaker’s opinion or raise voices against his/her point of view: avowedly, admittedly, apparently, arguably, incontestably, incontrovertibly, indisputably, unarguably, unquestionably, and undeniably.

The fact that these adverbs may also occur in the final position with a pause does not affect their entailment. Therefore, we can additionally distinguish between the members of this larger class of modal adverbs in terms of the nature of the evidence to which the speaker is appealing or in terms of the degree of probability assigned to the truth of the sentence. Therefore, a further distinction needs to be made within the class of modal adverbs on the basis of the kinds of evidence that modal verbs lexicalize:

1. perceptual evidence: apparently, clearly, seemingly, obviously, and evidently (knowledge by acquaintance)

2. evidence that is communicated: allegedly, apparently, and reportedly (knowledge by description)

3. evidence involving inferential processes: certainly, conceivably, definitely, possibly, presumably, probably, supposedly, surely, undoubtedly, and unquestionably.

In analyzing the situation of Romanian adverbs of this type, one has to take into account the fact that there is a different degree of insertion of SAs, setting Romanian apart from English. These adverbs can be either directly merged into clause structure or they can optionally allow for the selection of the complementizer că (that). Romanian also allows for some compensation because of its apparent lack of adverbs (most of them are derived by means of zero derivation from their corresponding adjectives) with prepositional phrases (PPs, e.g., cu siguranță (with certainty), de bună seamă (for real), fără îndoială (without doubt)). In fact, there is a clear tendency toward the use of a PP in Romanian to express a modal (epistemic) adverb, according to Protopopescu (2012, 166).

In Table 1, we present an overall picture of a construction that is available to certain Romanian SAs, more precisely, in this case, evidential adverb + complementizer că .

3 Translation strategies and predictions

In analyzing the data in the corpus, we employ the following translation strategies: equivalence, PP paraphrase, substitution, omission, overtranslation, and mistranslation. The main prediction is that equivalence and PP paraphrase should be the most frequently employed strategies for the translation of Romanian evidential adverbs of certainty. However, we expect a fair amount of instances of substitution, since this is also a possibility with evidential adverbs of certainty in Romanian where more often than not the preferred structures are (se pare (că) (it appears that/apparently), se vede (că) it is clear that/clearly, se zice (că), it is said that/allegedly). Given our previous investigation of other adverb classes in translation (see Protopopescu 2022, 394), we expect equivalence and PP paraphrase to be the preferred strategies here as well.

4 The corpus

Oscar Wilde’s ‘The Picture of Dorian Gray’ is obviously a very famous novel, which boasts a rich array of Romanian translations, ten to be more exact, spread over the past century. The first translation in Romanian goes back to 1920, but unfortunately, the translator has remained anonymous, and no evidence has been found to date as to who was the author of that translation. Unlike the following nine versions that followed, the title for this version is not what the reader might expect, namely Portretul lui Dorian Gray, but Crimă și conștiință (Crime and Conscience). The following two versions were published within 4 years of each other and 20 years later than the first translation. And, unlike the anonymity of the first translator, these two versions have a pair of translators each. The 1942 version, published in Bucharest, was translated by Viorica Hangic and Pericle Martinescu, while the one from 1946, published in Iași, was translated by Tilda Holda and Daria Luca. Then, in 1967, the fourth version was due to Dumitru Mazilu, and it was published in Bucharest, with subsequent editions after 2000. Following the 1967 version, there is a 34-year gap until the next version, published in Iași în 2001, in Magda Teodorescu’s translation. In 2005, a new translation by Răzvan Taliu was published in Bucharest. The year 2008 witnessed the publication of two versions: one published in Chișinău in Radu Tătărucă’s translation and another published in Bucharest in Amuliu Mircea Proca’s translation. The 2012 version represents not only a new version but also the only one that is dedicated to children, in the translation of Andreea Șeler. This is a clear departure from author-centered translations toward a user-centered translation (see Vîlceanu 2017, 79–80). Finally, the last, or rather the latest, version to be published appears in Bucharest in 2014, with subsequent editions.

Table 2 shows the timeline of these ten versions of the Romanian translation of The Portrait of Dorian Gray, with the labels to be henceforth used in this article.

Out of all the members in the class of evidential adverbs of certainty introduced in the previous section, I identified only four in the corpus provided by the source text (ST). These adverbs are as follows: certainly, of course, apparently, and surely. Not all positions discussed and illustrated in example (1) in Section 3 are available in the ST, as can be seen in Table 3.

From the data counted so far, we conclude that in the case of certainly, there is a clear preference for the medial position in the ST; in the case of surely, the preference is for the initial position in the ST; for of course, the preference is for the parenthetical position; for apparently, with only two occurrences, the preference is for the parenthetical position as well. The apparent lack of the final position is mainly due to the fact that, in cases where the adverb occurs in the final position, it is isolated by a comma, making it parenthetical, as we will illustrate in the examples discussed in later sections of the article.

5 Analysis of the data

The aim of this section is to discuss the distribution of the four evidential adverbs of certainty found in the ST in correlation with the translation strategies introduced in Section 3 of the present article. To this end, the analysis proceeds with a look at each one of these adverbs encountered in the ST, with relevant examples to illustrate the various cases. The back translation (BT) is our own.

6 Conclusions

Looking at the overall results of our investigation into the types of translation strategies employed in the corpus, we can conclude that the predictions are mostly borne out by the data, to the extent that the total of 507 instances of equivalence actually comprises instances of both the adverb equivalent as well as the adverb + complementizer că (that) type of structure. Therefore, as expected for Romanian, given its general lack of adverbs, the PP paraphrase is actually the preferred translation strategy. Substitution and omission are relatively equal in occurrence throughout the corpus. Overtranslation and mistranslation are at the bottom of the ladder, as was to be expected. Table 8 summarizes these findings.

If we are to analyze the results for each TT in the corpus against Berman’s RH, we note that the results are in line with what is to be expected, since TT1 exhibits the largest number of omissions, leading to a semantically poorer version of the text. As the target texts progress chronologically, the number of omissions decreases, and the number of equivalences and PP paraphrases increases. If we are to compare the three most recent target texts (TT8, TT9, and TT10), we note that TT9, which is also the translation meant for children, has a significantly higher number of PP occurrences compared to the number of equivalents, which seems to fall in line with Vîlceanu’s (2017, 79–80) argument related to author centered versus user-centered translations. Table 9 summarizes these findings.

Therefore, as predicted at the beginning of the article, equivalence and PP paraphrase were the preferred translation strategies, while the rest of the strategies represented a lower number of cases.

Given the analysis of our corpus, the translation of evidential adverbs into Romanian relies heavily on PP paraphrase and equivalence, sometimes resorting to substitution by adjective and mostly impersonal reflexive verbal phrases. This is explained mostly by the status of Romanian as a partly adverbial language, Protopopescu (2012). Sometimes the PP paraphrase may also lead to instances of overtranslation.

We envisage a more extensive analysis of a larger corpus of English literary texts to further check the translation strategies for Romanian evidential adverbs of certainty for an even clearer view of the preferred patterns for Romanian.

To conclude, the theoretical premise we started from, namely that the translation of evidential adverbs of certainty from English into Romanian poses problems of equivalence arising from the parametric difference between the two languages was borne out by the data in our corpus. The article demonstrated that, given this distinction between English and Romanian, the translation of evidential adverbs of certainty was achieved mostly by means of PP paraphrase and equivalence, closely followed by substitution with corresponding adjectives and verb phrases.