The origin of errors in translation

Updated: May 6, 2019

This communication will address the issue of error in translation, in the context of the teaching of translation practice in the Degree in Translation and Interpretation, aimed especially at the initial stages of training. Although this may seem to be a subject already widely dealt with from the perspective of translation studies, the truth is that most of the time it is limited to a purely taxonomic approach or to a simple quantitative count, remaining a simple observation of faults that do not always respond to objective or explicit parameters. My intention is to go beyond a new classification of potential errors or mistakes with a view to the evaluation of the translations produced during the period devoted to this teaching. For this reason, the analysis of the origin and classification of the error in translation (and not only for pedagogical purposes), must necessarily be preceded by a definition of it, eliminating as far as possible imprecise and subjective criteria, such as "it does not sound good" or "so it is not said". A simple classification of errors has no pedagogical value if it is not accompanied by an explanation of the cause of the error and, above all, why it can be considered an error regardless of the tastes or preferences of the reviewer.

Therefore, starting from the experience acquired in the task of correcting the translation assignments given to the students as a fundamental activity of their practical training in this matter, I will try to provide a definition of error attending not to abstract concepts such as translation units that supposedly have to be transferred in their entirety, but rather conjugating purely formal linguistic parameters, together with those that delimit the enunciation situation in which the TT is inserted and which also attends to extralinguistic parameters. Secondly, I will establish a classification of them, which is not intended to be definitive, of course, and which makes it possible to qualify their scope in relation to the text as a whole. It is in this moment when the explanation of the different causes that have provoked them intervenes, as well as the phase of the translating process in which they are produced, in order that the teacher can show the students how to prevent incorrect results.

This implies a conception of translation teaching that rejects trial and error methodology. That is to say, it does not have any pedagogical effect to present the students with a corrected assignment without providing comprehensible and clear explanations about the parameters followed for this correction, since it does not necessarily improve results or performance, but rather frequently generates a feeling of frustration ("I can't do it well", "I'll see if I'm right next time") that does not lead to the acquisition of skills or to routine practices of verification and documentation. On the contrary, it is more effective to explain to students, even the first time before correcting a translation order, what possible faults may appear in the reformulation process, i.e. the production of a TT. But this remedy is only completely effective when, after the correction of the order, the scope and type of fault is explained to them, making them aware of those that are serious, and can even invalidate the entire text, and those that are unsuitable for the declarative situation of the text: purpose, public, register, etc. If they do not acquire this awareness, or are shown the reason for the failure, they will think that they are translating according to the taste of a specific teacher, and not putting into practice a series of minimum translatological and linguistic skills, among which is the terminological search and the correct handling of the LT, which necessarily implies the grammatical and orthographic correction of the language. In short, the student must be able to assess for himself whether a text is suitable for the communicative purposes for which it is intended.

In a generic way, a translation error can be defined (the one detected in a text, as a translation) as a rupture of the rules of coherence of a TT, be they grammatical, lexical combinability, semantic congruence or in accordance with the knowledge of the world and the accumulated experience. It is these last three aspects that may apparently give rise to incomprehension and even disagreement in the eyes of those who have produced a translation that is delivered corrected, and in which they try to detect possible errors.

This definition is based on the model of the translation process I am starting from, a synthesis of two existing schemes, that of R. Muñoz and D. Gile, and gives an account of the different phases of this process, as well as the factors that condition it and that must be taken into account in order to carry it out successfully.

One issue that cannot be stressed enough is that of the translation order, since it is the one that conditions the other phases. In other words, the hypotheses of meaning are verified in relation to the situation of enunciation, and the reformulation is adapted to the intentionality of the TT, taking into account the subject matter (and degree of specialisation of the text), the textual type (e.g. instruction manual, advertising brochure, restaurant menu, etc.), the target audience or the scope of use of the text. An appropriate TT will only be achieved if all these elements are taken into account during the translation process.

Independent of the intentionality, which is determined by the translation order, are the errors that affect the correct linguistic formulation of the TT, and which can be qualified as invariant errors with respect to the translation process. For this reason I believe that we should not interpret, as some classifications such as that of D. Gouadec, for example, that there are absolute and relative errors, since in reality, all errors are absolute with regard to the production of a TT. The error is not only a concrete reformulation but the inability to perceive the rupture. The simple fact of not knowing or not being able to adapt to these conditions leads to reformulations that can hinder or impede the comprehension of the TT by the addressee, cancelling all or part of the communicative function assigned to the addressee. What is, however, relativisable is the impact that a given error causes in relation to the entire TT. Certain errors, or the accumulation of these, may invalidate the expected effect of the TT as a whole, or only segments of it. In any case its limit is given by the acceptability of the product for the client of the translation.

Thus, once the notion of error and the principles on which it is based have been established, classification can be established on the basis of the criterion of the stage at which it occurs, since only this can make it possible to determine its origin. In this way, students can clearly understand at what point the execution of the process has failed, so that they can anticipate and prevent similar behaviour in the next translation assignments they carry out.


Although experience allows me to be able to point out the errors that invariably occur one course after another, I have considered it more appropriate and above all rigorous to do so on the basis of a corpus of translation assignments carried out at the end of the course in the last two years by students in the first and fourth years of Translation and Interpretation. What did not seem relevant to me is to quantify the frequency with exact figures since I believe that it is necessary to considerably increase the TT number in order for the sample to be mathematically reliable. I will now enumerate them by frequency of appearance.

Sense TO ≠ Sense of TT

The most frequent errors detected are those that affect the process of attributing meaning to a text, and are reflected in the reformulation phase of the TT, causing breaks in coherence, terminological inadequacies, the use of literal meanings or lexical combinations not tolerated by the LT.

Coherence breakdown

a) The rupture of coherence can be caused by a hypothesis of meaning a which has not been subjected to a plausibility test. The failure originates in the construction phase of meaning, due to insufficient knowledge of the world (of the culture in which the TO is inserted) or of the subject being translated. It should be pointed out that we only refer to non-specialized texts, that is to say, intended for dissemination and for a public that is not a specialist in the field either.

A difficult, obscure, ambiguous wording, that is, the incorrect handling of the means of expression in LT that do not allow the comprehension of the sense that very probably has been correctly attributed. Its prevention therefore lies in a reinforcement of writing skills in the mother tongue.

b) Inadequate grammatical decoding of the TO, or in other words, the

ignorance of the grammatical value of certain categories, especially syntax, which lead to the attribution of a meaning that they do not possess and cannot be acquired by context or metaphorical use.

Terminological inadequacies

Rupture of semantic coherence by choosing a term whose meaning is not relevant in relation to the general sense of the whole text or of the segment in which the term is inserted.

The failure occurs at the level of the cognitive process, by the wrong choice of one of several possible meanings of a term stored in memory. Even one may not have been chosen inappropriately, but it is the only one that is possessed. In this case the origin of the failure lies in the insufficient lexical knowledge of the LO possessed by the student, which leads him to overlook the congruence test, since he is not aware of his shortcomings.

The bad choice can also be due to the non-respect for the translation order, that is to say, to the conditions of enunciation foreseen for the TT and which guide the series of taking of decisions on the general reformulation of the same one, as they are the tone, the register or, in the case to which it alludes, the terminology, selected from among the possible repertoire of the LT. If the translator is aware that he lacks sufficient repertoires, he searches for them in parallel texts and in other auxiliary tools such as the simple dictionary or the consultation of Internet pages.

Literal reformulation

With this denomination I am encompassing on the one hand the error originated by the non-recognition of idiomatic constructions (made phrases, lexicalized metaphors, paremias, etc.) in the phase of attribution of meaning and which provokes a word by word reformulation that changes the sense of the TT segment, or even causes it to lack the same. On the other hand, I include a fairly widespread trend among students that consists of the strategy of opting for such reformulation when they are not able to attribute any sense to the TO, and they consider that perhaps the recipient will be able to understand what for them is incomprehensible.

In the first case the knowledge of the LO fails once again. Although it is obvious that a speaker of the LO does not know all the repertoire of constructions either, the one he possesses (by comparison they are similar constructions) gives him sufficient clues, to know that the literal sense is not pertinent. A student who does not have this minimal repertoire will also be unable to recognize such clues, and will therefore retain the first meaning.

In the second case, the task is abandoned before it is completed, without probably having used all the resources the translator has to elucidate the meaning of a segment that presents special interpretation difficulties (and these difficulties are experienced by both a student and a seasoned translator). The student therefore considers that his comprehension abilities are limited, but that this may not be the case with those of the addressee, and hence he proceeds to rephrase word by word.

Formal incorrectness of the TT

It occurs in the reformulation phase, and its indisputable origin lies in the greater or lesser ignorance of the spelling, grammatical and lexical rules of the LT. A priori, this kind of errors are independent even of the translation condition of the text itself, i.e., they are not errors due in general to failures of a translation process, understood as the construction of a meaning from a TO with the expressive means available in the LT. These are also found in texts originally written in the LT. However, spelling and grammatical correction is the minimum of acceptability and quality required of a TT, since it is the necessary condition for the recipient to be able to make his own decoding and attribution of meaning, a task that for this reason may be difficult.

However, it is necessary to mention, precisely because of the frequency of its appearance, the existence of errors of this type originated by the inability to delimit the rules proper to each of the languages involved in the translation process, producing interferences of the LO in the LT. These errors are commonly referred to as "casts", which can be:

– Orthographic (both of spelling mistakes stricto sensu by interference or words that do not exist). Examples: renewal (le renouvellement), inequality (inégalité), onctuous (onctueuse).

– Grammars (morphology and syntax). Examples:

To conclude, it is necessary to point out that most of these errors could be solved with a thorough review, which is often overlooked or neglected. If we take into account that in many cases the student is not aware of the shortcomings in the linguistic competences of the LO and LT, it is difficult for the revision to be an effective means to avoid mistakes.

Thus, we can conclude that most of the errors that occur in the construction phase of meaning are not due, as is believed in many cases, to the proximity between the French and Spanish languages. They are the result of insufficient knowledge of the LO, and secondarily of the specific subject of the text to be translated. In the same way, the majority of those produced in the reformulation phase are due to insufficient knowledge of the LT, as well as not using auxiliary tools, such as dictionaries, internet, parallel texts, reference works, etc., to fill gaps of which they are not always aware.

For this reason, and it cannot be stressed enough, there is a need to strengthen competence in both languages, as well as to acquire the routine of consulting auxiliary means, an aspect that students systematically neglect.

Therefore, teachers have a large part of the responsibility for our students to be able to correct these errors by themselves, both in a careful revision and by designing tasks parallel to the translation whose specific purpose is an increasingly better handling of the possibilities of expression of both working languages.


© 2018 Freelance Translation Group

Birth certificate translation? Marriage certificate translation? Divorce certificate translation?
Whatever your needs for standard or official translation, the freelancers at Certified Translation UK can help, for the best price, guaranteed.