Assessing the English Translation of BFSU ’ s New Motto

The motto is a potent marketing tool in today’s globalized site of higher education. Beijing Foreign Studies University (BFSU) adopted a new motto in 2011 to reflect its new self-branding as a cosmopolitan scholar-doer. Its English translation has since then triggered much discussion about quality assessment. The current study critically surveys the existing literature on translation quality assessment (TQA), in an effort to identify an appropriate framework to assess the translation of Chinese university mottos. House’s model (2015) is found the most appropriate and applied to the official translation of the BFSU motto, after being adjusted in two important aspects. One, in regard to the rise of English as a language of global communication, it is proposed that more broad-based English norms than those of English as a native language be established for the purpose of adjudicating cultural filtering. Two, the use of corpus-based contrastive pragmatics is expanded to gauge the justifiability of overt as well as covert mismatches. While the errors identified by such a modified model are better intersubjectively verifiable, it remains to see how social research can be incorporated into the system to assess the degrees different errors may impact on the perceived quality of a translation.

(second criticism).But how will one argue or counter-argue about the quality of a motto translation less subjectively, more scientifically?

A Theoretical Prequel: TQA Literature
Literature on translation quality abounds.It is agreed that a translation should meet a quality standard and that it is of paramount importance to determine whether or to what extent a translation meets the standard.But the labeling and definition of the standard varies dramatically, and the means of measuring a translation against it is a matter of ongoing debate.
In my view, a sound TQA model should address the two fundamental relations that characterize a translation: its relation with the source text and its relation with the recipient language/culture environment.In House's (2015) words, " [T]ranslation is in a double-bind relationship.Translations are texts which are doubly constrained: on the one hand to their source text and on the other hand to the (potential) recipient's communicative conditions" (p.21).Interestingly, the first relation, between TT and ST, was made the centerpiece of TQA discussion (Note 1) from the start, albeit under the most controversial concept in translation studies "equivalence."Generally speaking, understanding of equivalence has loosely followed the advances in linguistic studies ever since its inception and benefited from the increasingly interdisciplinary horizons of the translation studies from the 1980s onward.It evolved from a static correspondence between elements of language systems to a situation-dependent transference of the important aspects of the ST, from atomic analysis to holistic considerations such as "text-bound equivalence", from a more or less monolithic notion of what should be reproduced to a multi-componential and often hierarchical conceptualization of priorities.But no matter how equivalence is conceptualized, TQA under this paradigm inevitably involves a comparison between ST and TT that determines the nature and degree of likeness between the two.
The second relation, between the TT and the translation users/readers, threatened to "dethrone" equivalence as the overriding principle of translation and thus the primary yardstick of TQA, with the advent of the cultural turn of the 1980s, especially with the immergence of the functionalist approaches to translation studies.Differences of focus and terminology between these approaches notwithstanding, the functionalist theorists agree in general that translation is a human action, so it is the skopos, i.e., the function(s) of the act of translation, rather than properties of the ST, that determine the translator's decision making.This model of translation highlights the relation between the TT-target readers/context, in that "it takes target-orientation seriously and even makes the target receiver the most important determinant of translational decisions" (Nord, 1997, p. 46).The quality of translation becomes a matter of how successful the TT is in fulfilling the function it is expected to serve in relation to its readers/users in the target context.Instead of measuring TT against ST, functionalists propose gauging TT against the intended purpose it sets out to achieve.In other words, TT adequacy with regard to function(s) replaces TT-ST equivalence as the primary standard for judging translations.
It is important to note that although the "equivalence" and functional approaches are viewed as contending theories of translation, the TQA proposals raised by the two camps, except the some of the early ones from the "equivalence" side, do make allowance for the legitimacy of the central concern of each other as a (possible) criterion of translation quality.As early as 1964, Nida put forward "dynamic equivalence," explicating the importance of eliciting the same response in TT audience as the ST does in the ST audience.And the development of linguistic inquiry towards Pragmatics and text linguistics certainly shifted the emphasis of equivalence towards target readers/listeners, giving rise to many function-oriented typologies of equivalence, for instance "Pragmatic equivalence" by Koller (1989, p. 102) and "functional equivalence" by Newman (as cited in Baker & Saldanha, 2009, p. 97).More recently, Emery (2004) attempted to rework the notion of equivalence, together with fidelity, in the framework of a pragmatic theory of communication.
On the other side, the functionalists had to grapple with the concept/standard of equivalence from the start, and as much as they privileged TT and its relation with the target context, equivalence was never completely wiped out of the map.For instance, besides the "skopos rule" and the "coherence rule," which are geared towards TT's reception, Vermeer did also postulate the "fidelity rule" which stipulates that some relationship ("intertextual coherence") must remain between the ST and the TT (as cited in Nord, 1997, p. 152).Furthermore, in cases where the skopos is the same for the ST and the TT, functional constancy (as cited in Schäffner & Wiesemann, 2001, p. 15), which can be construed as a special form of equivalence, is the standard of translation and the criterion of translation quality.Finally, it is generally accepted by functionalists that although a translation is not ipso factor a simulation of the ST, an invariance from the ST (it could be in terms of ST formal structures, or ST semantic meaning, or ST function among others) is one possible or legitimate TT skopos.Therefore, a translation may in deed in some cases, or even in a majority of cases, be judged by its likeness to the ST.
We can see from the above that "equivalence" and "skopos"/functional approaches to TQA are not mutually exclusive.Their difference, at least in theory, is a matter of priority in general and it diminishes almost to the point of nonexistence in the special (or normative) case of translation aimed at the same purpose as the ST.This can best be seen in the cognitive conflation of adequacy and equivalence in equivalence exponent Juliane House's definition of "functional equivalence" (Note 2) and representative functionalist Christiane Nord's notion of "equifunction".
House calls a translation covert when the source-text function is kept intact or invariant so that the TT enjoys the status of an original in the target culture (2015, p. 56).She adopted the term "functional equivalence" in her twice revised TQA work to refer to the goal in covert translation and thus its central criterion.To her, a functionally equivalent translation is "an adequate translation…a pragmatically and semantically equivalent one" (63).The ambivalent wording of such a definition, "adequate" on the one hand and "equivalent" on the other, is a manifestation of the mergence of the two fundamental pulls that shape a translation.In addition, House's "covert translation," corresponds to what Nord calls "equifunctional translation," a sub-categrory of "instrumental translation" in Nord's translation typology.And for Nord, the objective of such cases of translation is, as the name suggests, to achieve equivalent functions for the target audience (1997, p. 51).So, again the ST demand of equivalence and target-context call for functionality interfuse.This said, however, the quality of an equifunctional translation is determined by whether or to what extent TT (linguistic) solutions meets the real-world function expected of TT (which just happens to be the same as that expected of the ST).In other words, "equivalent function" is measured as a special form of adequacy, in line with the larger conviction of the functional framework that (all types of) equivalence may be subsumed under the "general" criterion of adequacy.The expected function is in turn concretized in the translation brief.Since the specification spelled out in a translation brief are typically real-world oriented and TT solutions are linguistic solutions, direct comparison between the two is problematic.The discrepancy between (abstract) linguistic structures and concrete real-world objectives explains one of the most frequent criticisms leveled at functionalist approaches: lack of (reliable) empirical data to prove that what the theorist-evaluator thinks is adequate solutions are really adequate, or inadequate solutions really inadequate.As Hönig mused: "The problem with any TQA […] is that it tacitly implies an assessment of a supposed reader's reaction.However, the empirical basis of the reader's putative reaction is often unclear" (1998, p. 29).Despite recent efforts to generate empirical evidence by means of interviews and questionnaires in one recent study and through think-aloud protocols in another (Nord, 2012, p. 33), the reliability and sufficiency of such methodologies are still open to question.
More important to the task at hand, in terms of translation into English, the ultimate validation in the form of expected effect-actualized effect comparison looks increasingly elusive, because any effort in gauging receptors' response is infinitely complicated by the fact of English as a lingua franca and the subsequent question of how to assess and reflect the audience heterogeneity of an English translation in evaluative methodology.In other words, now that English is the global language in almost all spheres of human engagement commerce, politics, education, etc, the potential readership of an English translation is no longer restricted to native speakers.If we use the "three circles model" originally discussed by Kachru (1986), an English translation, like any other English texts, is to be read by the Inner circle (communities for which English is first or the dominant language of social interactions), Outer circle (former colonies where English has been institutionalized for intra-communal use, like HK, Singapore...) and Expanding circle of English speakers (non post-colonial communities who nonetheless uses English as a foreign language for communication with people who do not speak their language).How to scientifically represent this diverse and elusive group in empirical data is a question that functionalist-oriented researchers have not begun to address.
In contrast, "functional equivalence," like other types of equivalence proposed by equivalence theoreticians, is clearly linguistically defined, entailing close analysis and comparison of ST and TT linguistic elements.For House, "Function" being compared is really "textual function," i.e., "the application (or use) of the text in a particular context of situation."And because House predicates her TQA model (Note 3) upon Halliday's functional Grammar, according to which "context of situation" is not viewed as a separate real-world entity, but a textual world, "encapsulated in the text…through a systematic relationship between the social environment on the one hand and the functional organization of a language on the other" (as cited in House, 2015, p. 63).The "textual function" that she refers to is really the functional potentiality of the text, not its actualized function in the social environment of real-world application.Such a text-based view of function allows "functional equivalence" to restrict itself to a symmetrical ST-TT linguistic/pragmatic comparison, sidestepping the above-mentioned conundrum faced by functional TQA approaches.

Development of the Current Story: Tailoring House's TQA Model for C-E Motto Translation
BFSU's motto and its English translation share the same objective of promoting BFSU's reputation and attracting more students, so it may be characterized as either "equifunctional translation"/"covert translation."(Note 4) As discussed in section 2, in this case, the determining forces of a translation, ST on the one hand and target context on the other, converge, though the two standards of translation quality 1) ST-TT functional equivalence 2) TT "equifunction", i.e., its adequacy to deliver the same effect as that of the ST, remain distinct from each other.It's tempting to argue that still either standard is enough to gauge the success of a covert translation, because such a special case of translation is by definition one in which ST and target context "pulls" the translation in one direction.And given the discrepancy problem still befuddling the functional approach, one had better opt for a functional equivalence-based methodology.Along this line of thinking, House's TQA model for covert translation, is arguably one of the most systematically conceived and extensively tested, and therefore an optimal candidate.
According to House, the first step of TT quality assessment is analysis of both ST and TT according to the scheme in Figure 1, which yields two functional-pragmatic profiles."While register captures the connection between texts and their 'micro-context', Genre connects texts with the 'macro-context' of the linguistic and cultural community in which the text is embedded" (2015, p. 64).
Figure 1.A revised scheme for analysis and comparing original and translation texts (House, 2015, p. 127) In the second step, the two textual profiles are compared to capture mismatches.They will be justified or questioned by way of what House calls "contrastive pragmatics", i.e., understandings of how the same genre (for instance, school mottos) texualizes and operates according to different norms and conventions in the target lingual-cultural community.The reliability of her contrastive pragmatics hinges on corpus studies, which she suggests enhances our understanding of communal norms and conventions by providing "an element of empirical intersubjectivity" (2015, p. 107).In other words, while certain mismatches can be justified as results of conscious or unconscious attempts by the translator to reflect differences in communicative preferences and norms ("cultural filtering") in the target lingual-cultural context, others would be proven unwarranted, by evidence drawn from corpus-based cross-cultural research.The discredited mismatches are called "covertly erroneous errors." Finally, a quality statement is produced by taking account of both the covertly erroneous and overtly erroneous errors, the latter resulting "either from a mismatch of the denotative meanings of source and translation text elements or from a breach of the target language system," and offering a description of the relative match between ST and TT in terms of textual functional components (p.33).
We can see from the above mentioned steps that House's model offers a fundamentally error-based assessment.Two types of errors are factored into the final quality statement, but only one of them, the "covertly erroneous" can be captured by way of the scheme as shown in Figure 1.The "overtly erroneous" ones do not factor into her evaluative scheme until the very last minute.This is easily compensated, according to House, because it does not take as much expertise to spot the obvious defects in a translation.I agree with House that it is the functional-pragmatic shifts in translation that are generally hard to account for (2015, p. 33).But her assumption that all overt mismatches, i.e., mismatches in denotative meanings, are errors is worth questioning.While unmotivated breaches of the target language system are erroneous, one can well imagine a translator deleting, adding, transforming denotative meanings in order to conform to norms in the target lingual-cultural context or in line with translational norms of the genre.Along this line of thinking, not all overt mismatches should be counted as errors, in the same way not all covert mismatches are.Of course, in view of the distinction she made between versions and translations, House would consider departures from the denotative meanings of the ST as acts of "versioning", the results of which fall outside the evaluative capacity of her model.But one would start wondering what little would, given that deleting, adding, adapting, transforming is much of what translators do.After all, what we often regard as a translation is probably more often than not partly the result of translating and partly that of versioning.Therefore, I suggest that we accept versioning as a legitimate form of translation and subject overt mismatches of denotative meaning to the same process of adjudicating as undergone by the covert mismatches.This means that we will use corpus-based contrastive pragmatics as the yardstick to gauge the justifiability of overt as well as covert mismatches.Figure 2 shows the adapted error-generating processes.We have discussed in section 2 how, in the case of translation into English, ELF has complicated the process of gathering empirical data to support a functionalist "equifunction" evaluation.Functionalists have yet to come to grip with this issue and design operable methodologies.House, on the other hand, broached the issue of EFL in relation to translation into English in the following remarks.
"While translation into English as a non-mother tongue has consolidated it presence in professional settings, the assumption underpinning the debate on this type of translation is that translators should adhere to the expectations of native English readers, rather than those of readers using English as a lingua franca.Such an attitude may well change as we can now see a veritable explosion of demand for translation from and into English as a lingua franca, translation being at the heart of global economy.Globalization, which characterizes much of contemporary life, has brought about a concomitant rise in the demand for texts that are simultaneously meant for recipients in many different languages and cultures."(pp.105-106) She suggests that in face of the rise of texts translated for EFL as well as ENL users, the traditional expectation for translators to adhere to ENL norms may change.As a consequence, I suggest, the evaluation of translation into English will have to answer to the challenge of establishing a more broadly defined set of English norms than those of ENL for the purpose of adjudicating cultural filtering in a certain genre.Unfortunately, House gives no guideline as to how to establish such broad-based global English norms, other than that it should be corpus-based.In fact it is unclear whether she has given practical implementation much thought, because the focus of her exemplar analyses (and the Hamburg project 'covert translation' she often refers to) remains translation from English, for which English as a global language has given rise to a different priority.
House's position that translation evaluation is optimally driven by corpus studies is shared by a number of scholars.As early as 2001, Bowker proposed the use of specially designed "evaluation corpus" that can act as a benchmark against which trainers can assess trainee translations.Hassani (2011) took a step further to use corpus to assess and improve translations in professional settings.Although both examined the directionality of translation into English, neither touched upon the issue of EFL in discussion of corpus construction for evaluative purposes.But, as Bowker notes, clearly, the value of what comes out of a corpus is largely dependent on what texts are included in it.In our case, the questions are Do we put in mottos from the Inner circle communities alone?Do we include the Outer circle and even the Expanding circle?
It is tempting to limit the corpus to the Inner circle on the assumption that it embodies the standards the rest of the world look up to.But research on ELF has shown that the Inner circle is not the sole rule provider.It is an established fact that not only has the Outer circle developed and is still developing their own rules, the Expanding circle users of English don't depend on Inner circle norms for meaning making either."ELF's common ground inevitably contains linguistic forms that it shares with ENL (English as a native language), it also contains forms that differ from ENL, and that have arisen through contact between EFL speakers, and through the influence of ELF speakers' first languages on their English" (Jenkins, 2009, p. 201).
But if we try to represent the whole spectrum of English speakers in the corpus, we face the issue that the boundaries of the Expanding circle is so fuzzy and expansive, that the whole world is included.An exhaustive corpus would include the mottos of all universities on earth that have been represented in English.Such an enterprise is ambitious but perhaps still manageable.But it is probably counterproductive in a project aimed at translation evaluation because one can well imagine that many of these mottos in English are wanting in quality themselves.And when they are put in a corpus, the chances are they are huge in number and may end up distorting the picture of what is normative.The trick of corpus studies in an evaluative program is that the descriptive turns prescriptive.When the objective is totally descriptive, an exhaustive corpus is okay.But when it is ultimately prescriptive, what we need is actually a corpus of "good" English mottos, "good" in the sense of embodying the generic norms.But if we know how to construct such a corpus, we would already know what the norms are.We will essentially plunge in chicken-and-egg type of circular reasoning.
In face of this dilemma, I decide to take a shortcut.Instead of including in the standard bearing corpus "good" English mottos per se, I put in the English mottos of the top world universities (THE World University Rankings 2015-16).Higher education has become in its own right a site for negotiation and evolution of English as a global language.The top universities are looked up to and emulated not only in their research capacities, teaching excellence, administrative acumen, but also in the ways they establish and maintain a global image.It is reasonable to assume that the patterns embodied by the mottos of top universities will be perceived as "good" and hence standard-setting.Likewise, when constructing the Chinese corpus for comparison, the mottos of the top universities in China, are deemed appropriate candidates.Because a total of 112 first-tier universities were listed in Project 211 and Project 985 of China as of 2015, 112 is the number of constituent mottos in both the Chinese and English corpora.It is worth noting that the English corpus thus constructed is still dominated by ENL norms, since the top 112 institutions with identifiable official or widely-circulated motto are overwhelmingly from Kachru's Inner circle.The United States tops the list with 48 mottos, followed by UK with 19, and Australia and Canada each with 6.But we also see significant representation of the Outer and Expanding circle countries/regions.Totaling 33, they account for about 1/5 of the mottos collected.

Evaluating the English Translation of BFSU's Motto: Quality Statement
Because of space limit, we are skipping ST-TT profiling and comparison (7 pages long) to present its results in the form of the quality statement.It is found that the most significant mismatches occur along the dimension of Tenor, the interpersonal component which is not totally incidentally what House considers the most difficult (and interesting) part of covert translation.In particular, the subcategories of "author's personal stance" and "social role relationship" have registered more major differences than others.These differences and their linguistic manifestations are summed up in the table below (Table 1).The Chinese text can be characterized as expressing co-existing layers of meaning and doing it artfully in a positive, involved tenor.It creates a sense of belonging and togetherness in the addressees at the same time stimulating them for further enhancement.The exhortative function is achieved with minimal pressure because it is up to the addressee to decide whether to read the motto as descriptive or requestive and hence how to participate in the communicative event.The social role relationship (that between the speaker and the addressee) is symmetric or near symmetric.
The English translation is less indirect/implicit.The lack of imperative-declarative ambiguity renders it a naked directive, which, together with its inherent negative connotations, amounts to camouflaged threats.The rhetorical strategies that play such an important role in the English text are less marked in the English translation.As a result, it comes across as a kind of injunction with a strong requestive teacher-like stance.
The single most important linguistic cause for the above-mentioned covert mismatches is the presence of syntactic ambiguity in the ST and its absence in the TT.
The Chinese motto seems a straightforward imperative clause complex, especially with reference to the English gloss.But since Chinese is a null subject language, i.e., a language in which lexical subjects are optional, the BFSU motto can be as likely interpreted as a null subject declarative clause complex with the implied subject "the university" dropped.In addition, although the four constituent words of the motto are easily identified as verbs, they may also be read as nouns.This is due to the well-observed slippage from verbs (and adjectives) to nouns in Chinese grammar (Yip & Rimmington, 2015, p. 16).Given the elliptical nature of the motto which does not lock the words in grammatical functions typically fulfilled by verbs such as the predicator, the words might be taken as nouns as well.In that case, the motto is read as an exophoric ellipsis, the omitted part of which is understood from the context as either "We uphold …" or "Uphold," reinforcing the declarative-imperative ambiguity.In contrast, the English translation has used a straightforward imperative clause complex.It is true that imperative is not face-threatening by nature.On the contrary, when used among equal-footed interlocutors, it reflects and builds in-group solidarity and camaraderie.But in the context of university mottos, it reinforces the asymmetrical relationship between the authors (university leadership) and the addressees (students, other international readers, etc.).Such a considerable functional discrepancy can be only justified by the use of the cultural filter.That is to say, if English mottos are found to have routinely expressed direct exhortations.But my corpus analysis shows the contrary.The most typical form employed by English mottos is the nominal group (complex) (43/112), followed in frequency by the declarative clause (complex) (35/112), the adverbial/adjectival group (complex) (14/112), the imperative clause (complex) (9/112), verbal group (complex) (8/112), and dependent clause (3/112).Although the imperative clause is not the rarest construction, it is important to note that all other constructions express exhortations in markedly subtler manners.The declarative clause (complex) typically presents a university's objectives and/or convictions, clearly missionizing but short of issuing commands.All the other categories are elliptical and can be fleshed out as declarative or imperative, thus displaying a syntactic ambiguity buffering the requestive speech act in the same way as we see the Chinese original.
What's more, the overwhelming majority of imperatives used do not call the addressees directly to action.This is because five of the imperatives use the third person imperative, and four of these five the special let-imperative.As a result, the burden of action is shifted totally (for instance, "Sun of Justice, Brighten Us" and "Let there be light") or partially ("Let your light shine") from the addressees of the motto.Altogether, only 4 instances are found in the corpus that call the addressees directly to action in the way the translation under examination does.All in all, the general tendency of English mottos is to express imperative meanings in subtler forms.The use of imperative in BFSU's motto translation is substantially more unjustified than justified.
As discussed in section 3, covert error is just one type of errors, the other being the overt.In regard to the latter, one overt mismatch, i.e., mismatch of the referential meanings of ST and TT items, is identified.笃行 duxing (in line with what's learnt practice) ≠ Serve a great cause.Clearly, the translator has rewritten the text substantially.Such adaptation maybe justified in certain genres, for instance, advertisement translation.But an examination of all the translated mottos in our English shows that motto translation is a fairly faithful to the original's denotative meaning.None of the translated instances exhibit significant departures.Therefore, the overt mismatch, now proven unwarranted, counts as an error.

Conclusion: Towards an Integrative Future
Two errors have been identified by using the tailored House's model, one covert and the other overt.They look reminiscent of the criticisms voiced by Wang as noted at the beginning of the current paper.But unlike those criticisms, they are firmly grounded in ST-TT functional-pragmatic analysis and have gained corpus-based intersubjectivity.Furthermore, corpus studies suggest that the overt error may be considered as more erroneous than the other, because the latter registered nil comparable practice in the corpus while the former at least a minimal incidence.
While we are confident that the quality of BFSU's English translation is adversely affected by these errors, we are not certain about the degree of the negative impact, and we will be hard pressed to grade the translation on the basis of these findings alone, for instance on a scale of 1-5.As House acknowledges (1998), her functional-linguistic approach does not enable the evaluator "to pass judgments on what is a 'good' or 'bad' translation" (254), a point reiterated towards the end of her 2015 book (142).This is because, any evaluation depends on social factors, "which concern human agents as well as socio-cultural, political or ideological constraints and which tend to be far more influential than linguistic considerations or the translator herself" (142).In the case of the English translation of BFSU's motto, these factors may at the very least include the attitudes among global users of English towards 1) the importance of camouflaging the exhortative function of university mottos, and 2) the value of faithfulness in motto translation.In other words, the linguistic diagnosis as presented in this paper is only the first step, albeit fundamental, towards assessing translation quality.Along this line of thinking, social research may after all have a role to play in TQA.Although it is beyond the scope of this paper to discuss how to triangulate linguistic analysis with social research exactly, it is imaginable to use it as the basis to assign weights to different types of errors identified by linguistic comparison.

Notes
Note 1. TQA discussion quoted here does not necessarily treat the issue of translation quality per se.Much of it is taken from writings on translation in general.But because TQA theories are predicated upon a premise about the nature of translation, and because translation and the standard of equivalence are often defined in terms of each other (Emery, 2004, p. 143), TQA discussion is part of the literature of translation studies from the very beginning.
Note 2. There are different opinions about House's theoretical allegiance.Here I agree with Nord, who insisted that House had always "subscribed to an equivalence-based…concept of translation" (2012, p. 30) Note 3. House's TQA model for translation, was originally proposed in 1977, revised in 1997, and most recently updated in 2015.House's model is aimed at offering systematic quality evaluation for translations in general.And that includes two types according to her typology: the covert and the overt.For overt translation, the criterion is not functional equivalence, but an equivalence of a "second level" function.Given that she has failed to make clear what it is exactly and none of her exemplar analysis includes cover translation, her model is better viewed as for the evaluation of covert translations only.Note 4.Although House acknowledges that the distinction of overt and covert translation is a not absolute and the two be better interpreted as extreme endpoints of a cline (143), it is not clear how these concepts can be applied in gradation when it comes to a specific translation.In all her model analyses, she has categorically labeled the text in concern "covert translation," with no discussion of where it stands along the cline.Here I am studying the translation of university mottos as "covert translation", but want to make it clear that I am aware of it not being at the extreme point of "covert translation", since after all, the conventional juxtaposition of Chinese original and translation (s) in university souvenirs (image products), makes the translation's status as such pretty obvious.

Figure 2 .
Figure 2.An adapted model for error-generating processes

Table 1 .
Major covert mismatches between ST and TT