Technical Translation: Usability Strategies for Translating Technical Documentation

Target-Orientated Approaches to Translation
 
Toury (1995:26) in his work 
Descriptive Translation Studies and Beyond 
puts forward the notion that the position and function of translations “are 
determined first and foremost by considerations originating in the culture 
which hosts them”. Toury regards translations as “facts of target cultures” 
and in his discussion of traditional methods of examining texts he mentions 
the fact that they “were primarily concerned with the source text and with 
its inviolable ‘sanctity” whereby target text factors “while never totally ig-
nored, often counted as subsidiary especially those which would not fall 
ing target language texts. And this is precisely how the translations are 
generally viewed by readers. 
The following paragraphs deal with approaches to translation which do 
not depend on the supremacy of the source text. In other words, theories 
which acknowledge the importance of the target text in the translation 
process. 
Relevance 
Gutt (1991:22) stresses that translation should be approached from a com-
municative point of view rather than any other, for instance, a theory of 
translation. Basing his relevance theory on work by Sperber and Wilson 
(1986), Gutt proceeds from the notion that the key to human communica-
tion is our ability to draw inferences from people’s behaviour, be it verbal 
or non-verbal. Thus, what is said (or written) provides, within the context 
of relevance theory, the stimulus from which the recipient can infer what 
the sender means. This is referred to as the informative intention. 
Breaking the communicative process down into steps, Gutt maintains 
that utterances are firstly decoded and, on the basis of their linguistic prop-
erties, are assigned to meanings or things they represent. This is similar to 
de Saussure’s notion of 
signifiant
and 
signifié
or 
aliquid statt pro aliquo 
[something stands for something else] (Linke 
et al. 
1994:18, 30). Thus, 
these representations are referred to as semantic representations and they re-
fer to the mental representations which are essentially the output of the 
emphasis away from the source-orientated treatment of translations and to-
within linguistics of any kind” (Toury 1995:24). Here we see a shift in 
wards the treatment of translations as autonomous, independently function-

---

Theory in Technical Translation 35 
mind’s “language module
5
” (1991:24). He continues to say that these se-
mantic representations are assumption schemas which need to be processed 
in order to become truth-conditional and to have a propositional form. In 
other words, out of all of the possible meanings or representations associ-
ated with an utterance, the brain must select what is in fact feasible, likely 
or credible within the particular context of the utterance. Indeed, it is the 
very notion of context which helps, according to Gutt, to explain the fact 
that while linguistic expressions do, in fact, have a meaning, it is not neces-
sarily the same as that conveyed by the expression at a particular time and 
place (1991:25). 
Gutt regards context as a psychological concept which, in contrast to 
common definitions, does not refer “to some part of the external environ-
ment of the communicative partners, be it the text preceding or following 
an utterance, situational circumstances, cultural factors etc., it rather refers 
to part of their ‘assumptions about the world’ or cognitive environment” 
(1991:25). Cognitive environment does not exclude the various external 
factors but rather includes them along with information they provide and its 
“mental availability for the interpretation process” (
ibid.
). 
But we find ourselves asking the question of how readers select the cor-
rect assumptions and meanings as intended by the sender, or author. This is 
indeed essential for communicative success. Gutt asserts that communica-
tion is governed by a desire to “optimise resources”. In other words, read-
ers want to gain as much information with as little expenditure of resources 
as possible. This means that meaning and understanding will always take the 
path of least resistance, and readers will start processing information using 
those contextual assumptions which are most accessible to them. He goes 
on to say that “[a translation] should be expressed in such a manner that it 
yields the intended interpretation without putting the audience to unneces-
quire us to build in certain contextual clues for a reader to fully and cor-
5
This notion of a language module is presumably made for the sake of conven-
ience. Although we can associate language skills with certain parts of the 
brain - usually in the left hemisphere (Eysenck 2000:79; Kalat 2004:441) - there 
sary processing effort” (1991:101-102). Relevance theory might well re-
rectly understand a text. This modification and addition of information
in the text brings us towards the notion of interlingual technical writing 
mentioned elsewhere. 
cognitive system. 
is no specific “language module” to speak of in the human cognitive system; 
language is regulated by a range of processes and functions throughout the 

---

36 Technical Translation 
Thus, the assumptions and information which allow the reader to under-
stand the communicative intent must be easily accessible to the reader. Of 
course, when applied to the text we can deduce that this information can 
be composed of both contextual information in the form of knowledge of 
the world and also information presented in the text itself. Gutt says that in-
formation obtained through perception is generally given more importance 
than information based on inference and so it gives the translator greater 
flexibility to ensure that the first interpretation of the utterance which the 
reader arrives at is the correct one as intended by the sender. And perhaps 
equally as importantly, particularly in the case of communicative texts and 
manuals to ensure that the effort involved in obtaining the meaning or con-
textual effects is justified (1991:30-31). So, the responsibility ultimately falls 
to the translator to decide how best to achieve this to ensure that the con-
textual effects are “adequate to the occasion in a way the speaker could 
have foreseen” (1991:31). Such a notion of target audience expectations is 
similar to assertions by Pinchuck (1977) who claims that elements in texts 
provide “triggers” for readers to allow them to anticipate what is to follow 
next. Indeed, Gutt states this even more explicitly when he says “introduc-
tory words would guide the hearer in searching his memory for the in-
tended referent and hence considerably ease his processing load” (1991:33). 
Similarly, Gerzymisch-Arbogast (1993) states that information must be pre-
sented in varying amounts of given and new information - relative to what 
the author perceives the audience to already know - in order to achieve the 
appropriate register and flow of information. 
Hönig (1997), however, levels the criticism at relevance theory that it 
fails to take into account precisely what readers regard as a good or bad 
translation, only what they regard as relevant. It is indeed a sound point in 
that we do need to take into account what readers regard as good and bad. 
It would be conceivable to determine what readers regard as good or bad 
texts either from empirical studies involving reader reactions or usability 
studies or to analyse a broad range of style guides relating to the text type in 
question. 
Relevance theory does provide compelling support for taking the cogni-
tive abilities of readers into account to improve translation but Gutt’s opin-
ions regarding what does and does not constitute a translation is quite prob-
lematic for technical translation particularly with regard to interpretative 
and descriptive translations. 
At the heart of relevance theory is Gutt’s distinction between interpre-
tive and descriptive language use and it is this distinction which is most 
relevant in terms of technical translation. Interpretive language use, accord-
ing to Gutt, refers to an utterance which is intended to represent what 

---

Theory in Technical Translation 37 
someone else thinks or says. In the context of translation, an interpretive 
translation is one which has some form of link or relationship with the 
source text. Descriptive language use, on the other hand, is intended to be 
an accurate representation of what the writer or speaker believes to be true. 
In the context of translation, a descriptive translation is intended to function 
as an independent and autonomous target language text whose readers will 
not be aware of the existence of a source text. 
According to Gutt, only an interpretive translation is a “real” translation. 
A descriptive translation, he claims, cannot be regarded as a true translation 
because a translation only achieves relevance by communicating effectively 
AND by standing in for the original text. While Gutt maintains that a 
translation must share at least some link with the source text, the notion 
that a text which can survive on its own and which is undetectable as a 
translation cannot be regarded as a true translation is difficult to reconcile 
with the reality of translation. 
We could argue that a descriptive translation mimics the production of 
an original text but it is still a translation albeit one that could have been 
achieved using very different means. Gutt’s assertion that only interpretive 
translations are true translations seems to contradict the communicative ori-
entation of his theory if translation is a communicative act, then the transla-
tion should do the same as a standard text or communicative act, i.e. con-
vey information in order to affect a change in behaviour, attitude or 
knowledge of the receiver (cf. Gommlich 1993:177-178). 
He does try to clarify interpretive use by making two sub-categories: di-
rect and indirect translation. Almost as if he realised that the existing cate-
gories of descriptive and interpretive would effectively eliminate translations 
whose function was to blend seamlessly into the target language landscape, 
he says that indirect translation allows the translator to “elaborate or sum-
marize” (Gutt 1991:122) in order to communicate as clearly as possible 
rather than “give the receptor language audience access to the authentic 
meaning of the original” (Gutt 1991:177). This certainly bears a remarkable 
similarity to the aims, factors and features involved in a descriptive transla-
tion. 
Gutt claims that if a translator aims to produce a translation that can stand 
alone as a target language text using whatever means necessary, then it is 
not a “real” translation but rather an “adaptation”, in principle this might 
be true but there is a problem with it insofar as technical translation is con-
cerned. Some texts require more work than others. For example, a techni-
cal specification or description of a process will require less work than a 
user guide or technical brochure. So while the technical specification or 

---

38 Technical Translation 
brochure are intended to serve as freestanding original texts, they cannot be 
regarded as anything but adaptations because there has been no “adapta-
tion” beyond that which is necessary to ensure compliance with target lan-
guage norms, requirements and conventions. 
On the other hand, a user guide which needs to be modified in a num-
ber of ways, e.g. register, tenor, cognitive suitability, legal compliance etc. 
can only be regarded as an adaptation. Here the ST has been used as the ba-
sis for producing an original target language text. The problem is that most 
people would argue that the user guide is still a translation and not an adap-
tation. Perhaps the problem is really one of scales. If there are huge 
amounts of adaptation, omission and addition then it becomes an adapta-
tion. What we are then left with is the problem of how much is too much? 
Nevertheless, despite the confusing approach to what does and does not 
constitute a “real” translation, relevance theory does provide some useful 
insights into technical translation: 
a communicative approach which concentrates on the needs and expec-
tations of the target audience 
the minimax principle which, put simply, states that people do not want 
to spend more time and effort than is absolutely necessary in order to re-
trieve information from a text; this is especially true for technical texts 
such as user guides.