The ex post facto study provided statistical evidence
for what follows:
· There is a significant difference between prior language
means of
the two groups differing on the basis of their translation
means.
· There is no significant difference between prior means in
academic cultural disciplines of the two different groups.
· There is a strong positive relationship between prior
language scores and subsequent translation scores.
· Prior language scores account for 62% of subsequent
translation scores.
These findings were revealing. Statistics showed not only a
statistically significant relationship between prior language scores and
subsequent translation scores, but also a meaningful one. It is meaningful in
the sense of its magnitude and strength. We believe that this result reflects
the relationship between prior linguistic competence and
subsequent quality of translation competence acquisition.
Now, could this confirm the hypothesis establishing linguistic competence as a
prerequisite to learning translation? In other words, could it prove
that having good prior linguistic knowledge causes good learning, and having
poor prior linguistic knowledge causes poor learning9
What we can claim, as a result of this statistical study, is
the existence of a strong correlation. In spite of this, we should admit that
correlation does not establish causality (Brown, 1988, p.146; Cohen & &
Manion, 1980, p. 131). In fact, what may suggest causality are the nature and
the direction of the relationship. These should constitute the theoretical
basis upon which hypotheses are set. Indeed a sound theoretical basis is what
determines the quality of correlational research (Cohen & Manion, 1980).
In the case of the present paper, theory had already
established the nature and the direction of the relationship. The link existing
between translation and language (Schleiermacher, 1999; Humboldt, 1880;
Catford, 1965; Mounin, 1963), and hence between translation competence and
linguistic competence (Mounin, 1962, 1973; Darbelnet, 1966; Hatim & Mason,
1990; Nord, 1999; Titone, 1995) were the basis of our hypotheses. The
literature suggests that language differences are the reason for translation
existence. This answers for the direction of the relationship; language was
there before translation. Furthermore,
language is the tool of translation, which determines the
nature of the relationship. Therefore, language competence, the tool, should be
there for translation, the activity, to be performed.
Correlation, then, established the fact that prior linguistic
competence had a strong association with subsequent learning of translation.
The nature and the direction of this relationship being determined, we believe
that correlation is all what was required to confirm the hypothesis stating
that prior linguistic competence is a prerequisite to learning translation.
The statistical study proved also that no significant
difference existed between the two different groups' culture means. Various
justifications might explain this. First, the information these disciplines
include may not be of use in the process of learning translation. History,
Geography and Philosophy curricula might not have much to do with the cultures
of the countries speaking the involved languages. In other words, the specific
contents of these disciplines might not help much in the acquisition of
communicative competence or in any phase of the translation learning process.
Or specifically, they might not have much to do with the translation course
content. As a result the learners did not need to use any of that information,
so their achievement in these disciplines did not contribute to their
translation scores.
Secondly, it might also be explained by the fact that
students did not learn well the content of these branches of learning. Our
qualitative study, exploring the knowledge of freshmen, confirmed this. Most of
the students proved unable to remember or use already seen information to
answer general culture questions. Thus, it might be a question of poor learning
or inability to use learned information outside its restricted context.
It is important, at this level, to tackle the issue of
culture of the language i.e. culture in its anthropological sense (see p. 26).
It is true that this type of knowledge was not part of our field exploration,
because testing it was problematic. Nevertheless, theory establishes the
importance of culture in language competence. The relationship between language
and culture (Newmark, 1988; Lotman, 1978; Bassnett, 1991) and hence between
linguistic competence and cultural knowledge (Chastain, 1976) account for this.
It is clear, as well, that cultural knowledge is what develops linguistic
competence into communicative competence (Hatim & Mason, 1990).
It is this strong relationship that leads us to express an
additional implication of this study's results. If prior linguistic competence
leads to better learning of translation, this would be also true of cultural
knowledge. The more prior cultural knowledge, the more communicative
competence, the better translation learning
We corne now to the discussion of the qualitative study's
results. The qualitative analysis provided qualitative and quantitative
evidence for what follows:
· The linguistic level of first year translation students'
is, in general, very low.
· First year translation students, in general, possess very
poor general culture.
· Third year students' translation competence is of a
relatively low level.
On the light of the ex post facto study results, we
believe the qualitative data could be interpreted as follows. First, we could
corne out with a general image of the current knowledge level of freshmen. Of
course, this evaluation does not concern the value of the Baccalaureate degree
as such. Actually, it concerns the level of the recent holders of the degree in
this specific part of the country i.e. the current level of the
Baccalaureate degree as reflected in its holders. It is clear that the level is
quite low, whether it concerns languages or general culture.
Secondly, we gained insight into the main characteristics of
third year students' translation competence. Concerning translation into
English, the level of the best translations produced by these students does not
exceed the third level, out of five, of the scale designed by Waddington
(2001). More revealing is the fact that this scale was
designed for second year Spanish students. As to
translation from English into Arabic, it does not exceed the fourth level of
the adapted scale. We should remind the reader that very few translations
fitted into the highest levels. This means that the majority were of levels
one, two and three. It follows that the level of our third year translation
students does not reach that of Spanish second year translation students.
The meaning we are tempted to attribute to all these data is
the following. The low level of third year translation students appears to be
explained by their low linguistic level as new university students. We strongly
believe that it must have been comparable to that of current first year
students. This interpretation is further supported by the correlation
established by the ex post facto study.
Some of the reviewed literature asserted that the amount of
knowledge included in a translation course is hard to cover within four or five
years (Pym, 2002, Mossop, 2000). With the observed students' level, this amount
of knowledge is increased by basic language material. Indeed, teachers feel
obliged to adapt their course contents to the students' level (Nord, 2000;
Gouadec, 2000; Gambier, 2000). Therefore, the pace of the learning process is
significantly slowed down. At the end of the course, we assume that the general
level would be barely intermediate (i.e. a little more than basic
knowledge).
The analysed translations showed also a great deal of
interference in the students' basic knowledge of the involved languages. This
seems to suggest that three years were not sufficient for students even to,
effectively and properly, acquire basic linguistic knowledge. It could be
deduced that learning to translate from and into languages whose basic
principles are not yet mastered might hinder language learning itself. Thus,
the qualitative study provided evidence that simultaneous learning of basic
linguistic knowledge along with translation from and into these languages is
not effective, and hence inappropriate.
This conclusion supports the theory cited in the literature
review about controlled linguistic knowledge (Titone, 1995). This
author asserted that acquiring two languages without interference requires hard
cognitive and affective efforts. Thus, acquiring more than two languages
(Arabic, French and English) along with translation would certainly be of a
questionable worth.
Another issue cannot be overlooked. The study indicated a low
level in Arabic language competence, in spite of the fact that the students
received their entire academic learning in this language. This might be a sign
of either the students' poor overall linguistic knowledge, or poor knowledge of
all kinds. Anyway, this leads us to draw two conclusions. First, the fact that
the selection system (see Appendix A) does not take into consideration grades
obtained in Arabic is based upon erroneous
beliefs as to the students' knowledge of their first
language. Second, we claim that this study's conclusions about linguistic
knowledge should be generalised to French as well.
In conclusion, all what precedes suggests that students
selected on the basis of scores in Baccalaureate exams cannot attain acceptable
degree of translation competence within three, or even four, years of study.
What seems quite fair to say is that these students will not be able to
practice the profession after their four-year course. In addition, the fact
that the three students who held university degrees in English produced the
best translations further confirms our main hypothesis.
To conclude, we claim that the established students'
selection system is not appropriate to train translators within four years.
Therefore, it should be adapted to the situation.
