Microsoft Word 05 descriptions doc

5.4 Theoretical concepts within the descriptive paradigm
Within this intellectual context, there is nothing obvious or banal in wanting to describe 
translations, rather than “prescribe” what translations should ideally be. Descriptive 
approaches were setting out to discover relations and laws that were in some way 
hidden; they were working on the problems of structuralism. Further, given the terms 
and concepts that had been accumulated in many related disciplines since the Russian 
Formalists, the researchers working on translation were able to draw on a broad range of 
established categories (no one can simply describe in language what they see with the 
eye) and indeed of hypotheses about what they would find. The underlying approach 
was certainly empirical, in the sense that these researchers were going out to test their 
ideas on actual data. But it would be quite misleading to reduce the approach to 
empiricism alone (as is done, for example, in Merino and Rabadán 2004), as if there 
were no properly theoretical concepts at work. 

---

In the following sections we will briefly describe the main concepts at work 
within the descriptive paradigm. In the next chapter we will look at the larger concepts 
of norms and laws.
5.4.1 Translation shifts and their analysis 
The most obvious way to apply structuralism to translation is to see the source and 
target texts as sets of structures. We can compare the texts and see where the structures 
are different, we then have specific structures (the differences) that somehow belong to 
the field of translation. That idea is as simple to understand as it is difficult to apply.
The structural differences between translations and their sources can be 
described as “
translation shifts
,” a term found in many different theories. For 
Catford
, 
shifts are “departures from formal correspondence” (1965: 73), which sounds clear 
enough. If formal correspondence is what we find between “Friday the 13
th
” and 
“viernes y 13,” then any other rendition will be a “shift” of some kind. The range of 
possible shifts might thus include all the things that Vinay and Darbelnet (1958) found 
translations doing, or indeed anything detected by anyone within the equivalence 
paradigm. A shift might come from the translator’s decision to render function rather 
than form, or to translate a semantic value on a different linguistic level, or to create the 
correspondence at a different place in the text (using a strategy of compensation), or 
indeed to select different genre conventions. Much research can be carried out in this 
way: compare the texts, collect the differences, then try to organize the various kinds of 
shifts.
There are at least two ways of approaching this task: bottom-up analysis starts from 
the smaller units (usually terms, phrases or sentences) and works up to the larger ones 
(text, context, genre, culture); top-down analysis goes the other way, starting with the 
larger systemic factors (especially constructs such as the position of translations within 
the sociocultural system) and working down to the smaller ones (especially categories 
like translation strategies). In principle, it should make no difference which end you 
start at: all roads lead to Rome, and there are always dialectics of loops and jumps 
between levels. Yet, perhaps surprisingly, the difference between bottom-up and top-
down has a lot to do with the role of theory in description.
5.4.1.1 Bottom-up shift analysis 
The range and complexity of bottom-up analysis is most completely seen in the 
comparative model developed by