She declares that unnatural translations which convey the source text's meaning receive half a score, whereas inaccurate translations receives no score, no matter how appropriate and natural the target texts sound.

In error recognition items, one score is given for spotting the error, and another one for correcting it.

Farahzad believes that scoring the long text can be done in several ways:

A: it can be scored holistically: since the item assesses a wide, variety of competencies, the examiner may find it convenient to approach the text as the unit of translation and adopt this system, especially with a large number of students.

The examiner may, for instance, come up with the following scheme:

B: it can be subjected to objectify scoring. In this system the target text must be read two times, first to check the accuracy and appropriateness, then for cohesion and style. Albeit time-consuming, this system is more reliable.

Farahzad suggests that sentence and clause might be the units of translation. Thus each verb in the source language text marks a score. The main clause receives one score and each sub-clause another score. So the accuracy and appropriateness are checked in each sentence and clause.

Cohesion and style cannot be checked and scored at the sentence and clause level. The elements of cohesion (e.g. transitional, appropriate use of pronouns, linkages, etc.) are spread all over the text as are the elements which form the style of discourse (choice of words, grammatical structures, etc.)

If, for instance, the source text is fairly neutral, one may allot a smaller number of points to it than in other cases where the preservation of style is important.

Farahzad's method seems a holistic method and it may cause some problems in evaluation of translations. Hence, it seems that Waddington's method might complete the Farahzad's method in assessment.

Waddington (2001) indicates that almost all the contributions in Translation Quality Assessment have been descriptive or theoretical and have centered mainly on the following themes:

(i) Establishing the criteria for a "good translation" (Darbelnet 1977, Newmark 1991);

(ii) The nature of translation errors:

- Defining the nature of translation errors as opposed to language errors (House 1981, Nord 1993, Kussmaul 1995, Gouadec 1989);

- Drawing up a catalogue of possible translation errors (Gouadec 1981);

- Establishing the relative, as opposed to absolute, nature of translation errors (Williams 89, Gouadec 89, Pym 92, Kussmaul 95);

- The need to assess quality not only at the linguistic but also the pragmatic level (Sager 1989, Williams 1989, Hewson 1995, Kussmaul 1995, Nord 1996, Hatim & Mason 1997);

(iii) Basing quality assessment on text linguistic analysis (House 1981, Larose 1989);

(iv) Establishing various textual levels on a hierarchical basis and linking the importance of mistakes to these levels (Dancette 1989, Larose 1989);

(v) Assessment based on the psycholinguistic theory of "scenes and frames"

(Dancette 1989 and 1992, Bensoussan & Rosenhouse 1994, Snell-Hornby 1995).

In order to find out the kind of translation exam and the kinds of methods of correction currently in use in Faculties of Translation, Waddington sent out a questionnaire to 48 European and Canadian universities. A total of 52 teachers replied from 20 of these universities and their answers reflected the following situation:

(i) All the teachers said that they require the students to translate a text, although over half also include other complementary tests.

(ii) As far as methods of evaluating student translations were concerned, 36.5% of the teachers use a method based on error analysis, 38.5% use a holistic method, and 23% combine error analysis with a holistic appreciation.

In accordance with these findings, he considers the validity of the results obtained through applying these different types of method to the correction of translations of part of an authentic text done by students under exam conditions.

Then, Waddington introduces four methods of assessment. The first method (method A) is more known than other methods and is functional in translation classes.

Method A is taken from Hurtado (1995); it is based on error analysis and possible mistakes are grouped under the following headings:

(i) Inappropriate renderings which affect the understanding of the source text; these are divided into eight categories: contresens, faux sens, nonsens, addition, omission, unresolved extralinguistic references, loss of meaning, and inappropriate linguistic variation (register, style, dialect, etc.).

(ii) Inappropriate renderings which affect expression in the target language; these are divided into five categories: spelling, grammar, lexical items, text and style.

(iii) Inadequate renderings which affect the transmission of either the main function or secondary functions of the source text.

In each of the categories a distinction is made between serious errors (-2 points) and minor errors (-1 point). There is a fourth category which describes the plus points to be awarded for good (+1 point) or exceptionally good solutions (+2 points) to translation problems. In the case of the translation exam where this method was used, the sum of the negative points was subtracted from a total of 110 and then divided by 11 to reach a mark from 0 to 10 (which is the normal Spanish system). For example, if a student gets a total of -66 points, his result would be calculated as follows: 110-66=44/11=4 (which fails to pass; the lowest pass mark is 5).

Jamal Al-Qinai in Translation Quality Assessment: Strategies, Parameters and Procedures (2000) indicates that translation is a complex hermeneutic process in which intuition plays a crucial role in interpreting the intentions of the ST writer. Further, languages vary in their choice of lexical connotations, sentence structure and rhetorical strategies, the only tangible tools for assessment. It is prudent, therefore, to talk about the adequacy of a translation rather than the degree of equivalence. Quality is relative and absolutes of accuracy cease where the end user (i.e. client) imposes his own subjective preferences of style in TT. Standardization of quality is thus a fuzzy grey area. For instance, does accuracy and good translation mean that a shoddy poorly-written, poorly-structured ST be reproduced as a shoddy poor TT? Is it professional for a translator to act as a filter, an advocate of ST? Alternatively, should a translator produce a "straight" translation rather than a "sanitized" one? (ITI Conference 1994: 72-3).

However, Al-Qinai in his study concerns to textual/ functional (or pragmatic) compatibility (i.e. quality of linguistic conversion) rather than to the logistics of management and presentation (i.e. quality of service). He points out that the ultimate end-users are interested in the quality of the product and not the means sought to serve its creation.

Al-Qinai Sets-up a model for translation quality assessment. He writes:

The assessment of a translated text seeks to measure the degree of efficiency of the text with regard to the syntactic, semantic and pragmatic function of ST within the Cultural frame and expressive potentials of both source language and target language.

Al-Qinai states that since no two languages are identical, either in meaning or in form, the best we can hope for is an approximation given the following variables: